Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

It is well established that cell adhesion can trigger a cascade of signaling events which profoundly influences the biology of the cell and its microenvironment (1). Adherence of monocytes (MOs) to tissue culture plastic or extracellular matrix components, such as fibronectin, collagen, or laminin, causes an increase in steady-state messenger RNA (mRNA) concentrations of genes encoding inflammatory cytokines, including interleukin (IL)-1, IL-8, tumor necrosis factor (TNF)-, and GRO (4). Freshly isolated MOs or cells cultured nonadherently express these genes at much lower concentrations. The adhesion-induced upregulation of cytokine mRNAs reflects a rapid (within 5 min) transcriptional activation, which is presumably due to the observed increases in the activities of several transcription factors (7).

In parallel with transcription factor activation, adherence of MOs also leads to an immediate increase in the activity of tyrosine kinases (10, 11), evident by the phosphorylation of many cellular proteins, the predominant one having a molecular weight of 76 kD (10, 12). Many of the major tyrosinated proteins belong to the set responsible for cytoskeletal assembly and movement, including the tyrosine kinase syk and the cytoskeleton-associated protein paxillin (10). Although less abundant, several key signal transduction kinases are also likely to be regulated through tyrosine phosphorylation. For example, all three members of the mitogen-activated protein kinase (MAPK) family are dual phosphorylated on serine/threonine and tyrosine as a step in activation (13). A dependence on tyrosine phosphorylation for gene regulatory events, as reflected in cytokine induction, has been suggested because treatment of monocytic cells with either genistein or herbimycin A decreased the adhesion-dependent expression of IL-1 mRNA (10, 11). It is also likely that extracellular regulated kinase (ERK) and p38 MAPKs are involved in cytokine expression, as we have previously reported that inhibitors of both kinases induced degradation of otherwise stable cytokine mRNAs (14).

Numerous studies have implicated reactive oxygen intermediates (ROIs) as inducers or second messengers in activation of the transcription factors nuclear factor (NF)-B and activator protein (AP)-1, which, in turn, are responsible for the subsequent induction of various cytokine genes (15- 18). Adherence of MOs and alveolar macrophages (AMs) have been shown to result in ROI production measurable by the generation of superoxide anion, hydrogen peroxide, or chemiluminescence (18), and recently it has been reported that stimulation of the oxidative burst in macrophages was sufficient to trigger translocation of NF-B (19).

Whereas MOs are mobilized and recruited to local sites after inflammatory stimuli, tissue macrophages are present constitutively and adhesively interact with matrix components in the absence of inflammatory signals. In the lung, resident AMs clear particulate debris, microbes, and pathogens from the airway passages by adhesive interactions leading to phagocytosis (20). Furthermore, AMs are major producers of cytokines and growth factors in various airway diseases (21, 22). There is very little information regarding the role of adhesion or phagocytosis-induced signals in cytokine gene induction in AMs. In the present study, we have examined the involvement of tyrosine phosphorylation and oxidant stress, induced by adhesion or the interaction with a particulate pollutant, in NF-B activation and IL-8 expression in AMs and compared the response to MOs that previously have been more extensively studied. Our results demonstrate that both AMs and MOs respond to adhesion with rapid increases in protein tyrosine phosphorylation and oxygen radical production. Whereas MO adhesion resulted in activation of JNK, p38, and ERK/MAPKs as well as IL-8 expression, the activation of ERK/MAPK in AMs was minor in response to adhesion and absent in response to residual oil fly ash (ROFA). AMs also failed to induce IL-8 expression after adhesion or ingestion of ROFA particles. Thus, neither generalized protein tyrosine phosphorylation nor ROI production are sufficient in AMs for NF-B translocation and transcriptional activation of IL-8. In addition, it appears that AMs have a dampening mechanism that uncouples routine adhesion signals from those required to provide immediate defensive responses.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation of Human AMs and MOs

Human AMs were isolated nonadherently from randomly selected healthy volunteers 18 to 35 yr of age, as previously described (23). The lavage cells were washed three times with cold RPMI 1640 (GIBCO, Grand Island, NY) and counted. AM purity was assessed by centrifuging some cells onto microscope slides and staining with Diff Quik (GIBCO). The isolated cells were 85 to 95% AMs, less than 1% polymorphonuclear cells, and the rest were lymphocytes. The AMs were used in the experiments without further density or adherence purification.

Peripheral blood was also collected from AM donors. MOs were isolated nonadherently by centrifugation through Ficoll/Histopaque 1077 (Sigma, St. Louis, MO) followed by Percoll (Pharmacia, Piscataway, NJ) gradients using procedures described previously (4). The isolated MOs were washed with sterile endotoxin-free saline (Baxter, Deerfield, IL), resuspended in RPMI, counted, and used for the experiments described below. Isolated MOs were > 75% pure as determined by morphology and by staining for nonspecific esterase, and were > 98% viable as determined by trypan blue staining. AMs and MOs from the same donor were used for each experiment. Stringent control of the handling procedures was maintained to prevent MO or AM activation.

Culture Conditions and Cell Stimulation

After isolation, cells were washed twice more with cold RPMI before use in the experiments. MOs and AMs were cultured separately in endotoxin-free RPMI 1640 medium (1 × 10⁶ cells/ml), without serum at 37°C under 5% CO₂. Nonadherently cultured cells were incubated in polypropylene tubes (Becton Dickinson Falcon, Franklin Lakes, NJ), with periodic mixing to minimize cell-cell interactions. Adherently cultured cells were plated at 5-10 × 10⁶ cells per tissue culture plate (100 mm; Corning, Corning, NY), or fibronectin-coated plate, prepared according to procedures described elsewhere (11). For lipopolysaccharide (LPS) treatment, cells were cultured nonadherently in the presence of 1 µg/ml LPS. For ROFA (10 µg/2 × 10⁵ cells) treatment, cells were also stimulated in suspension.

Western Blot Analysis

Whole cell extracts were prepared from 4 × 10⁵ cells by standard methods (24), resolved by 8% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis, and analyzed by immunoblotting using an antiphosphotyrosine monoclonal antibody (PY20; Transduction Laboratories, La Jolla, CA). Antibodies to the activated forms of p38, JNK, and ERK/MAPKs were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The antigen-antibody complexes were visualized by using goat antimouse immunoglobulin G peroxidase conjugates, followed by the use of an enhanced chemiluminescence kit (Amersham Corp., Piscataway, NJ) according to the manufacturer's guidelines.

Oxidant Generation

Chemiluminescence assays were performed as described previously (23). Briefly, immediately after stimulation, luminol reagent was added to the cells. The generated oxidants oxidized the luminol reagent and the resultant chemiluminescence (counts per minute [CPM]) was measured continuously over 30 min. The data are expressed as integrated 30-min chemiluminescence counts. Superoxide anion production was determined by monitoring the reduction of ferricytochrome C as has been previously described (23).

Electrophoretic Mobility Shift Assay

AM nuclear and cytoplasmic extracts were made using the procedure described previously (10). Briefly, after stimulation, cells were washed twice with phosphate-buffered saline and allowed to equilibrate for 5 min in ice-cold cytoplasmic extraction buffer (CEB) consisting of 10 mM Tris-HCl (pH 7.9), 60 mM KCl, 1 mM ethylenediaminetetraacetic acid (EDTA), and 1 mM dithiothreitol. Cells were lysed on ice for 5 min in NP-40/CEB/protease inhibitors (PI) (CEB containing 0.4% NP-40, 1 mM phenylmethylsulfonyl fluoride, 50 mg/ml antipain, 1 mg/ml leupeptin, 1 mg/ ml pepstatin, 40 mg/ml bestatin, 3 mg/ml E64, 1 mM 1,10-phenanthrolene, and 100 mg/ml chymostatin). After centrifugation, the supernatant (cytoplasmic extract) was collected and frozen, and the nuclei were washed in detergent-free CEB containing all the protease inhibitors, then suspended in nuclear extraction buffer (20 mM Tris-HCl [pH 8], 0.4 mM NaCl, 1.5 mM MgCl₂, 1.5 mM EDTA, and 1 mM dithiothreitol, 25% glycerol, and the panel of protease inhibitors listed above). After a 10-min incubation on ice, the solution was clarified by centrifugation and the supernatant (nuclear extract) was collected and snap frozen on dry ice before storage at 70°C. Protein concentration was determined by bicinchoninic acid method (Pierce, Rockford, IL). Electrophoretic mobility shift assays (EMSAs) were performed using 1 µg of AM nuclear extracts and 10,000-20,000 counts per minute (cpm) of [³²P]-labeled DNA probe containing the decameric DNA sequences from the class I major histocompatibility enhancer as described previously (10).

RNA Isolation and Northern Blot Analysis

Steady-state RNA concentrations for individual RNA species were determined by hybridization of gene-specific complementary DNA (cDNA) probes to Northern blots using procedures described previously (7). Total cellular RNA was purified from 3 × 10⁶ cells per sample point by the guanidinium isothiocyanate-CsCl method (24). Purified RNA (3-5 µg/lane) was loaded into each lane of denaturing agarose gels. The blots were visualized by exposing the membranes to films at 80°C.

Nuclear Run-On Analysis

After culture, nuclei were extracted from monocytes (10 × 10⁶ per sample point), and run-on transcription was performed exactly as has been described earlier (8). Control plasmid pEMSV and plasmids encoding IL-8 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were denatured and slot blotted onto Nytran membranes as specified by the manufacturer (Schleicher & Schuell, Keene, NH). [³²P]-labeled run-on RNA was hybridized to slot-blotted cDNAs. The blots were exposed to Kodak XAR film (Eastman Kodak, Rochester, NY).

Statistical Analysis

Statistical analysis of data in Figure 1 was performed by paired t test using INSTAT (Graph Pad, San Diego, CA) statistical software. Significance is indicated by an asterisk when P < 0.05. 

View larger version (41K): [in this window] [in a new window]   Figure 1.   Superoxide ions are generated after stimulation of human monocytes and alveolar macrophages by adhesion or by ROFA particles. Donor-matched human MOs or AMs were cultured nonadherently (NAD) or adherently to plastic tissue culture plates (ADH) in media alone or in media containing pollutant ROFA particles (+ROFA). After stimulation, superoxide anion production was measured by monitoring ferricytochrome C reduction as described in MATERIALS AND METHODS. Data are expressed as mean ± SEM.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Adhesion or Stimulation with Particle Pollutants Induces Protein Tyrosine Phosphorylation in Human Peripheral Blood Monocytes and Alveolar Macrophages

In contrast to MOs, little is known about the involvement of adhesion-induced signaling in the activation of cytokine production in AMs. Therefore, the adhesion-dependent induction of immediate-early events in AMs were compared with donor-matched MOs. The cells were incubated in suspension, with and without a particle, or plated on tissue culture dishes for different periods of time. Whole cell lysates were prepared and equal amounts of protein were analyzed by antiphosphotyrosine immunoblotting. As shown in Figure 2 (left panel ), 7 min of adhesion to plastic caused a significant increase in tyrosine phosphorylation. A series of proteins, with molecular masses ranging between 35-76 kD, were phosphorylated in both adhered AMs and adhered MOs with proportionally more low molecular weight species in the MO lane. As has been described previously (11), MO attachment to plastic resulted in tyrosine phosphorylation of a major protein with an apparent molecular mass of 76 kD, which also was heavily phosphorylated in AMs. In contrast, very low concentrations of endogenously tyrosine phosphorylated proteins were observed in freshly isolated MOs and a single species was detected in fresh AMs or in cells cultured in suspension. Culturing AMs on fibronectin-coated plates for 7 min resulted in tyrosine phosphorylation of a set of proteins similar to those phosphorylated on adherence to plastic (data not shown).

View larger version (83K): [in this window] [in a new window]   Figure 2.   Induction of protein tyrosine phosphorylation by cell adhesion or ROFA particle stimulation in human monocytes and alveolar macrophages. Immediately upon cell isolation (To), or after stimulation, whole cell lysates were prepared, equivalent amounts of protein per sample were resolved by SDS-8% polyacrylamide gel electrophoresis and analyzed for phosphotyrosyl-containing proteins by Western immunoblot as described in MATERIALS AND METHODS. Molecular mass references are indicated. (A) Donor-matched human peripheral blood monocytes and alveolar macrophages were stimulated by adhesion to plastic (ADH) or by nonadherent incubation with media alone (NAD) or with media containing pollutant ROFA particles (ROFA) for the times indicated.

Exposure of MOs and AMs in suspension to ROFA particles lead to tyrosine phosphorylation in both cell types, with phosphorylated substrates found in a size range from 20-200 kD (Figure 2). The response was stronger in AMs than in MOs. Whereas several bands appeared to overlap with those observed after adherence, a number of unique bands were noted. Particularly in the AM sample, the intensity of bands less than 46 kD was greater than that seen in MOs exposed to ROFA particles. Induction of tyrosine phosphorylation after particle stimulation of either cell type was observed at 7-15 min. Thus, resident AMs and peripheral blood MOs respond to adhesion or to particle stimulation by similar rapid, and at least as strong, tyrosine phosphorylation events.

Enhanced Oxidative Activity Is Observed after Adhesion or ROFA Particle Stimulation of Human Monocytes and Alveolar Macrophages

ROIs have been implicated in the induction of tyrosine phosphorylation, NF-B activation, and chemokine expression (15). Therefore, we investigated whether our adherence conditions would induce oxygen radical formation in MOs and AMs. Both adherence to plastic and exposure to ROFA caused an increase in the release of superoxide anions. Figure 1 compares the baseline and induced superoxide production by nonadherent and adherent cells in the absence or presence of ROFA particles. Adherence alone significantly (P < 0.05) increased the release of superoxide in both cell types, although AMs under these conditions produced approximately 10-fold the amount of superoxide compared with MOs. Exposure to ROFA resulted in increased superoxide production, again with significantly more superoxide production when cells were adherent. Thus, adhesion of either MOs or AMs promoted tyrosine phosphorylation of multiple proteins (Figure 2), and the generation of reactive oxygen species (Figure 1), events which have been linked to transcription factor activation and gene induction.

Adhesion or Exposure to ROFA Particles Fails to Rapidly Activate NF-B Function in Human Alveolar Macrophages

The preceding experiments demonstrated both ROI production and increased tyrosine kinase activities in MOs, as well as AMs, after adherence or exposure to ROFA particles. It was, therefore, expected that AMs like MOs (9) would translocate NF-B into the nucleus upon adherence. We performed EMSAs with AM extracts using the class I major histocompatibility complex enhancer NF-B motif as the DNA probe. Figures 3A and 3B show the presence of low concentrations of the p50/65 and p50/50 NF-B complexes in nuclear extracts of AMs cultured for 30 min in suspension or adhered to plastic dishes, and they show no increase in NF-B binding activity upon adherence. However, stimulation of nonadherent AMs with LPS for 30 min was sufficient to trigger strong NF-B DNA-binding activity (Figure 3A). To demonstrate that the low nuclear activity of NF-B was due to the retention of this factor in the cytoplasm, the detergent deoxycholate (DOC) was added to cytoplasmic extracts before performing mobility shift assays. DOC treatment revealed the presence of high amounts of p50/65 NF-B complexes in the cytoplasm of nonadherently or adherently treated AMs. Furthermore, detergent treatment of the cytoplasmic extracts exposed a slower mobility DNA-protein complex. By supershifting this complex with subunit specific antibodies, we determined that the binding activity was due to p65/65 NF-B homodimers (data not shown). Whereas 30 min of adhesion failed to activate NF-B, culture of the adherent cells for 4 h greatly increased the activities of p50/65 and p50/50 (Figure 3B). Exposure of nonadherent AMs with ROFA particles also failed to activate NF-B DNA-binding activity (Figure 3B). Thus, in contrast to the rapid activation of NF-B in MOs that occurred within 5 min of adhesion and paralleled tyrosine phosphorylation and ROI release, after either AM adhesion or particle exposure, these cells failed to respond with NF-B activation.

View larger version (18K): [in this window] [in a new window]   Figure 3.   Adhesion of alveolar macrophages results in delayed NF-B activation, whereas ROFA particle stimulation fails to activate this transcription factor. (A) Human AMs were cultured for 30 min, either adherently (ADH) to culture dishes or nonadherently in media alone (NAD) or in media containing 1 mg/ml LPS (LPS). After stimulation, nuclear and cytoplasmic extracts were prepared and equivalent amounts of protein per sample were used to perform EMSAs as described in MATERIALS AND METHODS. To establish the presence of cytoplasmic p50/65 NF-B complexes, sodium deoxycholate (DOC) was added to some extracts before the mobility shift assay. Using the major histocompatibility complex class I enhancer NF-B DNA motif as probe, the presence of p50/65, p50/50, and p65/65 NF-B complexes were determined. The identities of the different proteins were confirmed with subunit specific antibodies (data not shown). (B) Human AMs were stimulated by adherence to plastic (ADH) or by nonadherent incubation in the absence (NAD) or presence of ROFA particles (ROFA) for the times indicated. After stimulation, nuclear extracts were prepared and EMSAs for NF-B were performed using the class I major histocompatibility complex enhancer NF-B motif as DNA probe.

Neither Adhesion nor ROFA Particle Exposure Induces Early Cytokine Expression in Human Alveolar Macrophages

The lack of NF-B induction in AMs suggested a potential impairment of those specific immediate-early gene responses that are required for gene induction in response to adhesion or ROFA exposure. To examine the kinetics of adhesion-dependent cytokine production, we used donor-matched AMs and MOs and compared IL-8 and GRO mRNA expression by Northern blot analysis. AMs adhered for 30 min to plastic (Figure 4A), or to fibronectin-coated plates (data not shown), failed to express either IL-8 or GRO at concentrations higher than those seen in nonadherent cells. In agreement with previously reported data (9), the steady-state mRNA concentrations for genes encoding chemokines IL-8 and GRO increased after a 30-min stimulation of MOs by adhesion (Figure 4A). The level of expression of both cytokines increased over the following 4 h.

View larger version (39K): [in this window] [in a new window]   Figure 4.   Adhesion of alveolar macrophages results in delayed cytokine induction; exposure to ROFA particles does not induce cytokine expression. (A) Donor-matched human MOs and AMs were cultured adherently (ADH) or nonadherently with media alone (NAD) or media containing ROFA particles (ROFA) for the times indicated. (B) AMs were stimulated for 30 min by adherence to plastic (ADH) or nonadherent stimulation with media alone (NAD) or with media containing LPS (LPS). Total cellular RNA was extracted from freshly isolated (To) cells or cells stimulated as described previously, and analyzed for IL-8 and GRO mRNA expression by Northern blot. The ethidium-stained ribosomal RNA (rRNA) bands are given as loading indicators.

Lack of IL-8 and GRO expression by AMs was also observed after interaction of these cells (in suspension culture) with ROFA particles (Figure 4A). In MOs, IL-8 and GRO mRNAs were detected after 2 h of particle stimulation. Figure 4B shows that stimulation of AMs with LPS rapidly induced strong IL-8 mRNA expression.

Thus, the rapid surge in tyrosine phosphorylation events and ROI production after either adhesion or particle exposure did not result in cytokine expression by AMs.

The Tyrosine Kinase Inhibitor Genistein Inhibits Adhesion- and ROFA-Dependent ROI Production in Monocytes

Blocking tyrosine kinase activity with genistein is known to inhibit adhesion-mediated IL-1 mRNA expression (10, 11). To study the relationship between adhesion-dependent or ROFA-induced tyrosine phosphorylation activity, ROI production, and NF-B-dependent gene activation, MOs were first treated with various concentrations of genistein before stimulation. As determined by luminol-dependent chemiluminescence and ferricytochrome C reduction assays, ROI production by MOs decreased in a dose-dependent manner after genistein treatment (Figure 5). The chemiluminescence response (Figure 5A) after the oxidation of luminol reagent was used to measure ROI (hydroxyl radicals, hydrogen peroxide, and superoxide anions) production induced by ROFA. Ferricytochrome C reduction assays were used to demonstrate inhibition of adherence-induced superoxide release (Figure 5B). These data demonstrate that ROIs were generated in both adherence and ROFA-exposed MOs by a tyrosine phosphorylation-regulated mechanism.

View larger version (14K): [in this window] [in a new window]   Figure 5.   Genistein inhibits generation of reactive oxygen species by adherence- or ROFA-stimulated human monocytes. (A) MOs were pretreated with genistein as in Figure 4, then cultured nonadherently in the absence (Unstim.) or presence of ROFA particles (ROFA) for 90 min. Generation of oxidants was determined by chemiluminescence response. (B) Human MOs were pretreated with media alone or with the indicated amounts of genistein for 20 min at 37°C then cultured nonadherently (NAD) or adherently (ADH) for 90 min in the presence of the indicated amount of drug. After stimulation, superoxide anion production was measured by monitoring the reduction of ferricytochrome C. The figures represent experiments with cells from three different donors and are expressed as mean ± SEM.

Inhibition of Tyrosine Kinase Activity with Genistein Does Not Inhibit Adhesion-Dependent Transcription in Monocytes

To directly assess whether tyrosine kinase activity was required for the transcription of adhesion-induced IL-8 mRNA in MOs, we performed nuclear run-on analyses. MOs were treated for 20 min with genistein (5-20 µM) before 20 min of stimulation by adhesion to plastic. After stimulation, nuclei were isolated and nuclear run-on analyses were performed. The results demonstrate that MO adhesion induced transcriptional activation of the IL-8 gene (Figure 6A). Genistein, even at concentrations that decreased the steady-state mRNA concentrations of this gene (Figure 6B), failed to modify the rate of transcription of IL-8.

View larger version (20K): [in this window] [in a new window]   Figure 6.   Monocyte adherence-induced transcriptional activation of IL-8 is not inhibited by genistein. (A) Human MOs were pretreated with the indicated concentrations of genistein for 20 min at 37°C then cultured nonadherently (NAD) or adherently for 20 min in the presence of the indicated amount of drug. After incubation, nuclei were isolated and run-on transcription was performed. [32P]-labeled run-on RNA from each sample was hybridized to slot blots of cDNAs for IL-8 and GAPDH as indicated. Hybridization of run-on RNA to the pEMSV cloning vector was used as a control (control). (B) After treatment of monocytes with genistein or the diluent dimethyl sulfoxide (DMSO) (1% vol/vol), as described previously, total cellular RNA was extracted and analyzed for the expression of steady-state level of IL-8 mRNA by Northern blot. The ethidium-stained 28 S rRNA band is given as a loading indicator.

Adhesion and ROFA Exposure of Human Alveolar Macrophages Minimally Activates ERK/MAPK

The absence of transcriptional activation and expression of cytokine genes after adhesion of AMs suggested that initial signal transduction events required for activation of NF-B and other transcription factors may have been missing. To assess this, AMs and MOs were examined for activation of the three MAPK members, p38, ERK, and JNK, which have been demonstrated to play critical roles in transcriptional activation of NF-B or AP-1 (25). Adhesion of MOs, which is sufficient for transcriptional activation of IL-8, resulted in high level activation of all three kinases (Figure 7). Only p38/MAPK was constitutively expressed to a significant level. Treatment of the MOs with 20 µM genistein, before adhesion, almost completely blocked activation of all three kinases. In contrast, adhesion of AMs induced minimal activation of ERK/ MAPK, JNK/MAPK was constitutively active, and a slight induction of p38/MAPK was noted. Treatment with genistein had no effect on concentrations of JNK/MAPK and modestly decreased ERK and p38/MAPK. In contrast, addition of LPS to the adhering AMs, a condition shown to result in IL-8 induction (Figure 4B), resulted in a genistein-sensitive induction of ERK/MAPK, whereas the induction of p38/MAPK was genistein resistant. ROFA particles, which failed to induce early cytokine expression, also failed to activate ERK/MAPK, whereas a genistein-resistant induction of p38/MAPK did occur. Constitutive concentrations of JNK/MAPK were unaffected by ROFA particle exposure. Thus, it appears that minimal or absent ERK/MAPK activation by either adhesion or ROFA may reflect the inability to initiate immediate-early transcriptional activation of cytokine genes in AMs.

View larger version (41K): [in this window] [in a new window]   Figure 7.   MAPK activation in response to adhesion or exposure to ROFA. Human alveolar macrophages (AM) were stimulated by adherence in the presence (LPS) or absence (ADH) of 100 µg/ ml LPS. For comparison, AMs were also treated in suspension with ROFA particles. Duplicate cultures of each were also exposed to 20 µM genistein (+). As a control, human monocytes (MO) from the same donor were adhered in the presence or absence of genistein. After 30 min, cell extracts were made and Western transfer analysis of the activation of p38, ERK, and JNK kinases were carried out. For comparison, freshly isolated MOs were also cultured nonadherently for the equivalent time period (NAD).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Both tyrosine kinase activation and oxygen radical formation have been implicated as critical components in the signal transduction pathway leading to the transcriptional activation of NF-B-dependent cytokine production (29- 33). Adherence provides a robust signaling event in MOs leading to transcriptional activation of several cytokines, including IL-1 and IL-8 (8, 9). Much less is known about adherence and particle-induced signaling in AMs. This study investigated and compared the involvement of tyrosine phosphorylation and oxidant radical generation in NF-B translocation and IL-8 gene induction in stimulated MOs and AMs.

Adherence to tissue culture plastic or fibronectin induced tyrosine phosphorylation in both MOs and AMs, with a similar array of proteins being phosphorylated in both cell types. The tyrosine phosphorylation response to the pollution particle ROFA was much stronger in AMs compared with MOs. A number of proteins in the size range < 45 kD were heavily phosphorylated after ROFA exposure. As these particles contain large concentrations of vanadium (34), a known inhibitor of tyrosine phosphatase activity, suppression of endogenous phosphatases would presumably facilitate appearance of the supranormal levels of tyrosine phosphorylation observed here. Oxygen radical formation was stimulated in both cell types upon adherence or ROFA exposure and, as reported in granulocytes, was linked to tyrosine phosphorylation (35). Furthermore, the respiratory burst triggered in mouse macrophages by phagocytosis of zymosans required tyrosine phosphorylation (29). In this study, adherence- induced ROI production in both MOs and AMs was inhibited by tyrosine kinase inhibitors genistein and herbimycin A (data not shown).

Numerous studies have shown that ROIs are involved in NF-B translocation, which subsequently promotes expression of NF-B-dependent genes (12, 14). Treatment of cells with antioxidants, such as N-acetylcystein, -lipoic acid, and vitamin E derivatives, suppressed NF-B activation (29, 34). Schreck and coworkers (16) reported that stimulation of cells with exogenous oxidants, such as hydrogen peroxide, resulted in NF-B activation. Kaul and Forman (19) showed that activating the respiratory burst in rat AMs by phorbol myristate acetate induced NF-B translocation. Therefore, it was surprising to find that neither adherence nor exposure to ROFA resulted in NF-B activation or expression of the chemokines IL-8 and GRO in AMs, despite strong tyrosine kinase activation and oxidant stress induced in these cells. Other studies have also suggested that oxidant stress may not be sufficient or necessary for NF-B activation (36, 37). Transient overexpression of catalase, a scavenger of superoxide ions, did not block NF-B activation in COS cells after cytokine stimulation (37). Brennan and O'Neill (36) showed that, whereas hydrogen peroxide activated NF-B in Jurkat T cells, it failed to activate NF-B in another T cell line (EL4.NOB-1) or the KB epidermal cells. Moreover, antioxidants inhibited TNF-, and IL-1 induced NF-B activation of Jurkat cells but had no such inhibitory effect on the EL4.NOB-1 or KB cells.

Multiple pathways leading to NF-B activation appear to be used by different cell types. Evidence is conclusive that cytoskeletal regulatory guanidine triphosphatases, Rho and Rac, are early mediators of the signaling cascade that results in activation of members of the MAPK family (38, 39). This directly implicated adhesion and phagocytosis-associated changes in the actin cytoskeleton with downstream regulation of gene expression. Activation of JNK/ MAPK is known to result in phosphorylation of c-Jun and subsequent activation of the AP-1 transcription factor (40). Recent evidence also demonstrates that JNK/MAPK activation may lead to activation of NF-B (41, 42), although this is not true of all cells and signal events (27). While p38/MAPK is often implicated in activation of NF-B (25, 43), a recent report indicates that ERK/MAPK activated via a c-src-MAPK-pp90rsk pathway can also be a sufficient signal for NF-B activation (28). Hwang and colleagues (26) have demonstrated in macrophages that this is sufficient to activate NF-B in response to LPS. In MOs, the three kinase pathways JNK, p38, and ERK/MAPK were all activated upon adherence of MOs. This would be in keeping with the generalized activation of cytokine gene transcription that occurs within minutes of adhesion (9). However, whereas both JNK and p38/MAPK were constitutively active in AMs, and moderately induced by adherence and ROFA exposure, ERK/MAPK was minimally activated by adherence and not at all by ROFA particle exposure. In contrast, LPS exposure, which rapidly induced NF-B translocation and IL-8 expression, activated ERK/MAPK in AMs. This suggests that an uncoupling of cell surface receptor activation from the downstream events leading to ERK/MAPK activation may be an important regulator of AM environmental responses. Physiologically, it makes sense that not all types of adhesive interactions will trigger cytokine production in AMs because these cells are continuously involved in particle clearance. Cytokine and chemokine production may be selectively induced when the AMs encounter pathogenic microorganisms.

The relationship between tyrosine kinase-dependent MAPK activation and transcription of IL-8 was further investigated in adhering MOs. Pretreatment of the cells with genistein or herbimycin A had no effect on NF-B translocation and transcriptional activation of IL-8, whereas steady-state IL-8 mRNA expression was drastically decreased by genistein treatment. p38, JNK, and ERK/MAPK phosphorylation was strongly inhibited. This would suggest that transcriptional activation of IL-8 may not be totally regulated via the MAPK signaling pathway. Indeed, it has been reported that activation of NF-B and p38/MAPK may be mediated by separate pathways (44). On the other hand, the genistein-dependent decrease in IL-8 expression reported in this study appears to be associated with inhibition of the MAPKs, as we have previously reported that specific inhibition of p38/MAPK or ERK/MAPK resulted in a marked decrease in mRNA stability (14).

These data strengthen the argument that the oxidant stress model of NF-B activation may be restricted to specific cell types. The data with MOs imply that NF-B translocation and IL-8 gene induction is independent of oxidant stress, and the results with AMs suggest that induction of oxidant stress is an insufficient signal for NF-B-dependent gene activation. Furthermore, failure of either adhesion or ingestion of ROFA particles to activate the ERK/MAPK pathway may signify an uncoupling of chemokine responses to repetitive lung stimuli from those of lung pathogens.