INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Asthma is associated with cellular infiltration of the bronchial mucosa with T cells and eosinophils (1). This airway inflammation is thought to increase bronchial hyperreactivity and mucus secretion and cause episodic airway obstruction. Several studies have shown an association between the number of activated T cells and eosinophils in the airways and airway reactivity, lung function, and clinical severity in patients with asthma. Antigen-specific Th2 lymphocytes play a critical role in the generation of allergic inflammation through the release of cytokines, such as interleukin (IL)-4, IL-5, IL-9, and IL-13, which promote the activation and survival of eosinophils. Eosinophils are thought to be an important effector cell involved in the induction of bronchial mucosal damage by the release of cytotoxic proteins, reactive oxygen metabolites, and proinflammatory and profibrotic cytokines (2). Although increased numbers of T cells and eosinophils are present in the airways and bronchoalveolar lavage (BAL) fluid of patients with atopic asthma, the mechanisms responsible for their preferential migration and accumulation are not completely understood.

The regulation of leukocyte migration is a complex process involving the participation of selectins, integrins, and chemokines. Chemokines are a large superfamily of secreted proteins that direct the migration of leukocytes through specific seven transmembrane-spanning G protein-coupled receptors (3). It has been suggested that the specific expression of chemokines in tissues will be an important determinant of the composition, intensity, and chronicity of an inflammatory response (3). Although the chemokines have overlapping activities in vitro, it is the timing and pattern of expression in vivo that are important determinants of their roles in the pathology of a disease process.

In this regard, a number of chemokines have been shown to be increased in lungs of patients with asthma compared with normal control subjects. Monocyte chemoattractant protein (MCP)- 1, RANTES, MCP-3, and eotaxin have been shown to be increased in the epithelium and bronchial mucosa in patients with asthma compared with normal control subjects (4). Eotaxin and MCP-1 were also increased in BAL fluids from patients with asthma (6, 8). In addition, RANTES, MIP-1, MCP-1, and eotaxin were released into the BAL fluid following endobronchial antigen challenge of patients with allergic asthma (9).

We have molecularly cloned and functionally characterized a chemokine with a high degree of homology to MCP-3 and eotaxin, called MCP-4 (12). MCP-4 is a potent chemoattractant of eosinophils, monocytes, lymphocytes, and basophils (12). It also induces histamine release from interleukin (IL)-3-primed basophils (12, 14) and activation of the respiratory burst in eosinophils (15). MCP-4 signals cells though several chemokine receptors, including CCR2, CCR3, and CCR5 (12, 13, 16). We and others have previously demonstrated that MCP-4 mRNA can be induced in endothelial cells and epithelial cells by tumor necrosis factor- (TNF-) and IL-1, but not IL-4 or interferon- (IFN-) (12, 13). TNF- and IL-1 levels are increased in the lungs of patients with asthma and have been implicated in the pathophysiology of asthma (20, 21). TNF- has been shown to induce the characteristic asthma phenotype, including eosinophilic pulmonary inflammation and airway hyperreactivity (20). However, TNF- is not directly an eosinophil chemoattractant, suggesting that in vivo chemotactic effects of TNF- are indirect, perhaps through the induction of chemokines. In fact, TNF--induced leukocyte recruitment in an air-pouch model has been shown to be dependent on chemokines (22). In addition, IL-4-induced eosinophil accumulation in the skin has been shown to be dependent on endogenous TNF-, suggesting an important role for TNF-- induced chemokines in the regulation of Th2 responses (23).

MCP-4 has been shown to be expressed in a limited number of human disease tissue. We have found that MCP-4 mRNA is expressed in the epithelium and submucosa of nasal biopsies obtained from patients with both allergic and nonallergic sinusitis and allergic rhinitis (12, 24, 25). MCP-4 was also detected in the luminal endothelium of an atherosclerotic vessel (16). Because MCP-4 is broadly active on immune cells implicated in the pathogenesis of asthma, we have sought to determine if MCP-4 is expressed in the airways of patients with asthma. In this report we show that MCP-4 messenger RNA (mRNA) and protein are expressed at higher levels in the epithelium and bronchial mucosa of patients with asthma compared with normal control subjects and that levels of MCP-4 protein in the BAL correlate with eosinophil influx. In addition, we also show that MCP-4 expression can be induced in monocytes and lymphocytes by cytokines similar to those that induce its expression in epithelial cells. Thus, the data are consistent with the hypothesis that MCP-4 plays a role in the recruitment and activation of eosinophils in the airway of patients with asthma and thus the pathogenesis of this disease.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Materials

Recombinant human cytokines, including IFN-, TNF-, MCP-4, MCP-1, MCP-2, MCP-3, eotaxin, and RANTES, were obtained from Peprotech Inc. (Rocky Hill, NJ). IL-1 was obtained from R&D (Minneapolis, MN).

Study Population

Patients and control subjects were recruited from the asthma clinic (Notre-Dame Hospital, Montreal) and their clinical characteristics were previously described (6). Ten patients with positive skin tests to at least one aeroallergen and the diagnosis of asthma (as defined by the American Thoracic Society) (26) were studied. As control subjects, nine nonasthmatic nonallergic individuals with negative skin tests and normal spirometry were also studied (recruited from the Notre-Dame Hospital, Montreal). Written informed consent was obtained, and the study was approved by the appropriate Institutional Review Board. None of the subjects was a current smoker and all had less than a 5 pack/yr history of smoking. Patients had not received inhaled or systemic corticosteroids in the last 3 mo and were not receiving medications other than inhaled -agonists. Patients or control subjects had not suffered symptoms of an upper respiratory tract infection within the past month. Nebulized salbutamol was given to all subjects (patients with asthma and control subjects) before bronchoscopy.

Bronchial Biopsies and Lavage Fluid Processing

Bronchoscopy was performed for bronchoalveolar lavages and endobronchial biopsies were obtained as previously described (6). Bronchoalveolar lavage was performed by injection of 50 ml of saline into the right middle lobe. Approximately 25 ml of BAL fluid was recovered and centrifuged at 600 g for 10 min at 4° C. The cell-free supernatant was concentrated ten-fold using the centricon system (Amicon, Beverly, MA) and stored at 80° C until further analysis.

In Situ Hybridization

Bronchial biopsies were immediately fixed as previously described (6) and kept at 80° C. BAL cells were suspended at a concentration of 10⁶ cells/ml in phosphate-buffered saline (PBS) and centrifuged in a Cytospin III (Shandon, UK) at 400 rpm onto superfrost slides (Fisher, Pittsburgh, PA), fixed, dried, and stored at 80° C. In situ hybridization was performed as previously described (6). Briefly, the complementary DNA (cDNA) for human MCP-4 (12) was subcloned into the pBluescript vector (Stratagene, La Jolla, CA). The sense and antisense strands were generated by T3 and SP6 RNA polymerases, respectively. Labeling of RNA probes with fluorescein isothiocyanate-11-uridine triphosphate (FITC-11-UTP) was performed according to manufacturer's recommendations (Boehringer Mannheim, Mannheim, Germany). Preparations were acetylated in 0.1 M triethanolamine for 5 min and then in 0.1 M triethanolamine containing 0.25% acetic anhydride for 10 min. To avoid nonspecific hybridization, prehybridization was carried out with a solution containing 100 mM dichlorodiphenyltrichloroethane (DTT) and irrelevant riboprobes for 2 h at 42° C. RNase A (20 µg/ml) was used during posthybridization washing. Hybridization was carried out with an antisense MCP-4 riboprobe for 16 h at 50° C. As a negative control, preparations were also hybridized with FITC-11-UTP-labeled sense probes in the same conditions. Specimens were then stained for 1 min with Hoechst 33258 dye (bisbenzimide; Sigma Chemical Co., St. Louis, MO) for visualization under a Zeiss Axiophot fluorescence microscope.

Generation of MCP-4 Antibodies

To prepare a polyclonal serum, New Zealand White rabbits (Pocono Rabbit Farm and Laboratory, Canadensis, PA) were immunized with recombinant human MCP-4 protein. Monoclonal antibodies (mAb) to MCP-4 were prepared according to standardized methods established in our laboratory (27). Briefly, 6-wk-old BALB/c mice were immunized intraperitoneally with 50 µg of human recombinant MCP-4 emulsified in 100 µl of complete Freund's adjuvant (Sigma). Three weeks later these mice were immunized again intraperitoneally with 50 µg of human recombinant MCP-4 emulsified in 100 µl of incomplete Freund's. Three days before fusion, animals were injected intraperitoneally with 50 µg of human MCP-4 protein in incomplete Freund's adjuvant. Spleen cells from immunized mice were fused to the 63A8.653 hybridoma cell line using polyethylene glycol. Hybridoma supernatants were screened for MCP-4 reactivity using a direct enzyme-linked immunosorbent assay (ELISA) as previously described (6) and then tested for specificity using direct ELISA and Immunoblot analysis. Three positive clones were specific for human MCP-4 and did not react to human MCP-1, MCP-2, MCP-3, RANTES, and eotaxin by direct ELISA or Immunoblot and were single cell cloned by two rounds of limiting dilution. These monoclonal antibodies, L1D2, L3B5, and L3B3, were further purified by affinity chromatography over a Protein-G Sepharose column (Pharmacia, Piscataway, NJ) from serum-free culture medium.

Immunostaining Studies

Fresh frozen sections (5 µm) and cytospin slides were saturated with universal blocking solution for 10 min (Dako, Carpenteria, CA), slides were incubated with L1D2 mAb overnight at 4° C, washed twice with Tris-buffered saline (TBS), followed by an incubation for 1 h at room temperature (RT) with 5 µg/ml swine anti-mouse alkaline phosphatase-conjugated antibody (Sigma). Positive cells stained red after development with Fast Red (Sigma). Omission of the primary antibody and incubation of the primary antibody with excess recombinant MCP-4 protein as well as staining with an isotype-matched control mAb were used as negative controls. To determine the phenotype of the inflammatory cells in the BAL, cytospins were stained with the appropriate mAb, anti-CD-3, anti-CD68, antineutrophil elastase, and anti-Major Basic Protein (MBP) antibody as previously described (6). After each incubation with antibodies, slides were extensively washed with TBS. Nuclei of cells were stained for 1 min with hematoxylin.

Quantification

For immunostaining and in situ hybridization, positive cells were counted in a blinded fashion in a random coded order using a Zeiss Axiophot microscope at ×200 magnification. Cells exhibiting positive hybridization signals or immunoreactivity were counted in at least two fields (counting a minimum of 400 total cells) for BAL cytospin preparations. In the airway submucosa, positive cells were counted along the entire length of the epithelial basement in a minimum of six sections. The percentage of MCP-4-positive cells was calculated, and results reported as the mean percentage ± SD.

ELISA

A sandwich ELISA was generated to detect MCP-4 in biological fluids. Purified L1D2 (100 ng) in 100 µl of PBS was used to coat 96-well microtiter plates at RT for 2 h. Unbound sites were blocked with 100 µl/well of blocking buffer (3% bovine serum albumin in PBS). Standard dilutions of MCP-4 in blocking buffer or BAL fluids (10-fold concentrated) were added in duplicate to the coated wells and incubated for 2 h at room temperature. After washing with wash buffer (PBS containing 0.1% Triton X-100), 100 µl of the polyclonal rabbit anti-MCP-4 antibody (1:50,000) in blocking buffer was added to each well at RT for 2 h. Following washing a horseradish peroxidase-labeled goat anti-rabbit IgG antibody (Kirkagaard & Perry, Gaithersburg, MD) diluted 1:10,000 in blocking buffer was added to each well and incubated for 1 h at RT. The plates were washed and the bound peroxidase activity developed with 3,3',5,5'-tetramethylbenzidine substrate (Kirkagaard & Perry). The reaction was stopped by the addition of 1 M H₃PO₄ per well and the absorbance at 450 nm was measured. This assay was sensitive to 60 pg/ml.

Epithelial Cell Culture

A549 cell line derived from a lung adenocarcinoma with the alveolar type II cell phenotype (American Type Culture Collection, Rockville, MD) was cultured in Dulbecco modified Eagle medium (DMEM)-F12 with 10% fetal bovine serum (FBS). Cells were grown to confluence and serum starved for 24 h before cytokine stimulation. In experiments involving dexamethasone (Sigma) or cycloheximide (Sigma), the agents were added 30 min before cell stimulation. In time course experiments, the cells were harvested 1, 2, 4, 8, 16, 24, 32, and 48 h after stimulation. In concentration-response studies, the cells were harvested at the time of peak expression, which was 4 h after stimulation with IL-1 or TNF-. Protein synthesis was inhibited by the addition of cycloheximide (10 µg/ml) to the medium that inhibited > 90% protein synthesis in A549 cells (27).

Isolation and Stimulation of Peripheral Blood Mononuclear Cells (PBMCs), Monocytes, and Lymphocytes

PBMCs were isolated from the heparinized venous blood of 16 healthy volunteers by density gradient centrifugation with Histopaque 1077 (Sigma). This procedure yielded ~ 1 × 10⁸ PBMCs from 100 ml of venous blood from each volunteer, with PBMC purity > 98%. None of the subjects had allergic disease or peripheral blood eosinophilia (eosinophil percentages were < 5%), and all gave written informed consent with the prior approval of the appropriate institutional review board. PBMCs were cultured in RPMI-1640 with 10% heat-inactivated FBS on 10-cm plates at 5 × 10⁶ cells/ml. Monocytes were purified by adherence following an overnight culture in RPMI-1640 with 10% type AB human serum (Sigma) and lymphocytes were purified from the nonadherant population by discontinuous density gradient centrifugation with Percoll (Sigma). Lymphocyte purity was determined by flow cytometry (FACScan, Becton Dickinson) with antibodies to human CD14, CD3, CD19, and CD56 (Phamingen, San Diego, CA). Lymphocyte purity was greater than 98%, and the percentages of T cell, B cells, and natural killer (NK) cells were 75-89%, 5-12%, and 8-11%, respectively. PBMCs, monocytes, and lymphocytes were cultured in RPMI-1640 with 10% heat-inactivated FBS on 10-cm plates at 5 × 10⁶ cells/ml and were stimulated with TNF- (10 ng/ml) and IL-1 (10 ng/ml) and harvested for RNA isolation at 1, 2, 4, 8, 24, and 48 h after stimulation.

RNA Analysis

Total RNA was isolated from freshly harvested cells by guanidinium-thiocyanate-phenol chloroform extraction (Stratagene Inc., La Jolla, CA). For Northern analysis, 10-20 µg of total epithelial cell RNA was subjected to gel electrophoresis on a formaldehyde-2% agarose gel and transferred to a nylon membrane (Schleicher & Schuell Inc., Keene, NH). After ultraviolet (UV) cross-linking, the membrane was hybridized at 42° C in a 50% formamide buffer (pH 7.5), containing 10% dextran sulfate, 5% SSC, 1% Denhardt's solution, 1% (wt/vol) sodium dodecyl sulfate (SDS), 100 mg of herring sperm DNA/ml, and 20 mM Tris either with a ³²P-labeled 850 bp EcoRI fragment of the human MCP-4 cDNA (12) or with a glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA probe. The membranes were washed for 2 × 20 min at 42° C in 2% SSC-0.1% SDS and then for 2 × 20 min at 55° C in 0.2% SSC-0.1% SDS. RNA expression was quantitated by densitometry. To control for RNA loading, the hybridization signal obtained for MCP-4 was normalized to that of GAPDH for each sample. All experiments involving the quantitation of RNA were performed at least in triplicate and a representative blot is displayed. However, levels of MCP-4 can be accurately compared within a given blot only as a result of differences in exposure times between blots.

Statistical Analysis

Statistical comparisons of the results were performed by a non-parametric Mann-Whitney test. The strength of the association between infiltrating cells as determined by immunocytochemistry and the BAL concentration of eotaxin and MCP-4 was determined by linear regression analysis. All statistical tests were performed using Instat Software 2.0 (GraphPad Software, San Diego, CA). p Values less than 0.05 were considered significant. For Northern blot analysis, data were tested for normalcy and equal variance and were analyzed by both ANOVA and ANOVA based on ranks. Differences between groups were determined by the method of Dunn or Newman-Keuls as appropriate. Analysis by ANOVA and ANOVA based on ranks gave similar results. The data are presented as mean ± SEM as all data that did not contain zero values met assumptions of normalcy.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

MCP-4 mRNA Expression in the Airways

The expression and distribution of MCP-4 mRNA in the tissues of 10 patients with asthma and in 9 normal control subjects were examined (Figure 1). In normal control subjects, MCP-4 cRNA hybridized weakly in human bronchial epithelial cells and submucosa cells (Figure 1A). In patients with asthma, an increase in the hybridization signal was seen in epithelial cells and inflammatory cells in the submucosa (Figure 1C). Positive in situ hybridization signals were observed only when the antisense probe was employed and not when the sense probe was used for either normal tissue (Figure 1B) or tissue from patients with asthma (data not shown). The quantification of the in situ hybridization revealed a statistically significant increase in the number of MCP-4 mRNA positive cells in the airway submucosa of patients with asthma compared with normal control subjects (28.8 ± 6.3 versus 11.7 ± 4.2, p < 0.01) (Table 1).

View larger version (112K): [in this window] [in a new window]   Figure 1.   Expression of MCP-4 mRNA (A, B, C ) and protein (D, E, F ) in airways of patients with asthma (C, F ) and normal control subjects (A, B, D, E ). MCP-4 mRNA is increased in airways of patients with asthma (C ) compared with normal airways (A). Representative examples of in situ hybridization using an FITC-labeled MCP-4 antisense riboprobe in airway biopsies of normal subjects (A) and patients with asthma (C ) revealed MCP-4 mRNA (bright yellow) in both the epithelium and submucosa in biopsies from patients with asthma with less staining in normal airways. As controls, no signal was seen with the sense riboprobe on normal airways (B) or airways of patients with asthma (data not shown). The epithelium is on the top of each panel. Representative examples of immunohistochemical staining using a mouse anti-MCP-4 mAb in airway biopsies of normal subjects (D) and patients with asthma (F  ). MCP-4 immunoreactivity was visualized with fast red chromagen and revealed an increase in MCP-4 protein both in the epithelium and submucosa of airways of patients with asthma compared with airways of normal control subjects. As controls, no signal was seen with an isotype-matched control mAb on normal airways (E ) or airways of patients with asthma (data not shown). Original magnification: ×200.

MCP-4 Protein Expression in the Airways

A specific mAb to MCP-4 was developed and used to study the expression of MCP-4 protein in lung biopsy tissue obtained from patients with asthma and normal volunteers. Analysis of normal airways by immunohistochemistry demonstrated weak MCP-4 immunoreactivity in airway epithelial cells and submucosal mononuclear cells (Figure 1D). An analysis of airways of patients with asthma revealed that MCP-4 protein expression was increased in epithelial cells and inflammatory cells in the submucosa (Figure 1F). The pattern of staining for MCP-4 protein appeared to be intracellular rather than membrane bound. No immunoreactivity was found in any cell type, when the first antibody was omitted or preadsorbed with excess recombinant MCP-4 or when an isotype control mAb was used in either normal (Figure 1E) or airway tissue from patients with asthma (data not shown). The quantitation of the immunostaining revealed a significant increase in the number of MCP-4-positive cells in the airway submucosa of patients with asthma compared with normal control subjects (33.4 ± 6 versus 13.8 ± 4.2, p < 0.01) (Table 2).

MCP-4 mRNA and Protein Expression in Cells Recovered from BAL

As previous studies have demonstrated that BAL cells from patients with asthma are capable of producing several chemokines (6, 8), it was of interest to determine whether BAL cells would express MCP-4 mRNA and protein. BAL cells from subjects with asthma had a significantly increased number of cells expressing positive signals for MCP-4 mRNA (28.5 ± 10.4 versus 10.6 ± 3.5) (Table 1). Morphologically identified lymphocytes, macrophages, and eosinophils stained positively in subjects with asthma, whereas in normal control subjects, mRNA MCP-4 was expressed only sporadically by macrophages and occasionally by epithelial cells (data not shown). Results obtained with in situ hybridization were also confirmed by immunostaining as the number of cells expressing MCP-4 immunoreactivity was significantly increased in BAL from subjects with asthma when compared with control subjects (28.9 ± 8 versus 10.8 ± 3.6, p < 0.01) (Table 2).

MCP-4 Protein Expression in BAL

BAL fluid was collected from 10 patients with asthma and 9 normal control subjects. MCP-4 protein levels were measured by a sandwich ELISA using 10-fold concentrated BAL fluid. We detected significantly higher levels of MCP-4 protein in BAL fluid obtained from patients with asthma compared with normal control subjects (188.2 pg/ml ± 92.9 versus 54.4 pg/ml ± 28.1, p < 0.001) (Figure 2A). Because the BAL fluid of patients with asthma contain high levels of eotaxin and eosinophils compared with control subjects (6), it was of interest to determine whether there is a correlation between eotaxin, eosinophils, and MCP-4 protein. In the BAL fluid of patients with asthma there was a significant correlation between eotaxin and MCP-4 (r = 0.78, p < 0.05) (Figure 2B), MCP-4 and MBP⁺ cells (r = 0.85, p < 0.05) (Figure 2C), and eotaxin and MBP⁺ cells (r = 0.77, p < 0.05) (Figure 2D). No correlation was observed between macrophages, lymphocytes, neutrophils, and MCP-4 protein (data not shown).

View larger version (18K): [in this window] [in a new window]   Figure 2.   Detection of MCP-4 protein in BAL fluid by ELISA and relationship between levels of MCP-4, eotaxin, and eosinophils in patients with asthma. (A) BAL from patients with asthma (n = 9) and normal subjects (n = 9) was concentrated (10 times) and used in an MCP-4-specific sandwich ELISA as described in METHODS. MCP-4 levels in BAL obtained from patients with asthma were significantly higher than in normal control subjects (p < 0.001). Regression analysis of the relationships between (B) BAL fluid concentration of MCP-4 and eotaxin (previously reported in [6]), (C) BAL fluid concentration of MCP-4 and numbers of MBP+ eosinophils, and (D) BAL fluid concentration of eotaxin and numbers of MBP+ eosinophils. The correlation coefficients (r) were significant at indicated p values.

Cytokine Regulation of MCP-4 mRNA Expression in Epithelial Cells

Airway epithelium participates in the generation and propagation of the inflammatory response observed in the lung of patients with asthma by producing a number of proinflammatory mediators, such as cytokines. We have previously shown that the respiratory epithelial cell lines, BEAS-2B and A549, expressed MCP-4 mRNA in response to stimulation with IL-1 or TNF-, but not in response to IFN-, transforming growth factor (TGF)-, or TGF- (12). To further examine the regulation of MCP-4 in epithelial cells, we performed kinetic and dose-response studies. When A549 cells were treated with IL-1 (10 ng/ml), MCP-4 mRNA accumulation was detected at 1 h, peaked at 2-4 h, and declined over the subsequent 48 h (Figure 3A). A similar kinetic response was seen for MCP-4 accumulation in response to TNF- (10 ng/ml) (Figure 3B).

View larger version (48K): [in this window] [in a new window]   Figure 3.   Time course of IL-1 and TNF--induced MCP-4 mRNA accumulation in A549 epithelial cells. Northern analysis of 20 µg of total RNA harvested from A549 cells at various times after treatment with 0.1 ng/ml IL-1 (A) or 10 ng/ml TNF- (B). To control for RNA loading a quantitative comparison at each time point between the MCP-4-specific signal and the GAPDH-specific signal is presented below each blot as mean (n = 3 independent experiments) ± SEM.

To assess the dose-response of cytokine-induced MCP-4 mRNA expression, geometrically increasing doses of IL-1 and TNF- were added to cultures of A549 cells and the cells harvested 4 h later. Increasing doses of cytokine was associated with increasing MCP-4 mRNA accumulation (Figure 4). The effect was maximal for IL-1 and TNF- at 10 ng/ml.

View larger version (43K): [in this window] [in a new window]   Figure 4.   Dose-response of IL-1 and TNF--induced mRNA accumulation in A549 epithelial cells. Northern analysis of 20 µg of total RNA harvested from A549 cells at 4 h after stimulation with increasing concentrations of IL-1 (A) or TNF- (B). To control for RNA loading a quantitative comparison at each time point between the MCP-4-specific signal and the GAPDH-specific signal is presented below each blot as mean (n = 3 independent experiments) ± SEM.

Effects of IFN- on Cytokine-induced MCP-4 mRNA Accumulation

IFN- alone did not induce the expression of MCP-4 mRNA in A549 cells by northern analysis (12) (and data not shown). However, the addition of IFN- to IL-1-stimulated (0.1 ng/ml) A549 cells increased the accumulation of MCP-4 mRNA in a dose- dependent manner (Figure 5A). This synergistic effect was also seen in TNF- (10 ng/ml)-treated A549 cells (Figure 5B).

View larger version (50K): [in this window] [in a new window]   Figure 5.   Dose-response characteristics of IFN- synergy. Northern analysis of 20 µg of total RNA harvested from A549 cells at 4 h after treatment with 0.1 ng/ml IL-1 (A) or 10 ng/ml TNF- (B) in the presence of increasing concentrations of IFN-. Below each blot a quantitative comparison is made at each dose between the MCP-4-specific signal and the GAPDH-specific signal.

Effects of Dexamethasone on Cytokine-induced MCP-4 mRNA Accumulation

We also analyzed the effects of corticosteroids on the expression of MCP-4 in cytokine-stimulated epithelial cells. The addition of dexamethasone to the medium was associated with a dose-dependent decrease in IL-1-induced and TNF--induced MCP-4 mRNA expression (Figure 6). Dexamethasone treatment had no effect on cell morphology or viability (data not shown).

View larger version (49K): [in this window] [in a new window]   Figure 6.   Dose-response characteristics for dexamethasone suppression of IL-1 and TNF--induced MCP-4 accumulation. Northern analysis of 20 µg of total RNA harvested from A549 cells at 4 h after stimulation with 0.1 ng/ml IL-1 (A) or 10 ng/ml TNF- (B) and increasing concentrations of dexamethasone. Below each blot a quantitative comparison is made at each dose between the MCP-4-specific signal and the GAPDH-specific signal.

Effect of Protein Synthesis Inhibition on Dexamethasone-treated Epithelial Cells

To examine the effect of protein synthesis inhibition on MCP-4 mRNA accumulation in cytokine and dexamethasone-treated A549 cells, these cells were treated with cycloheximide. IL-1-induced MCP-4 mRNA accumulation was superinduced when cycloheximide was present in the media and part of the suppressive effect of dexamethasone persisted in the presence of cycloheximide (Figure 7A). TNF--induced MCP-4 mRNA expression was also enhanced in the presence of cycloheximide (Figure 7B). Dexamethasone inhibition of TNF--induced MCP-4 mRNA expression was less apparent in the presence of cycloheximide than that observed for IL-1.

View larger version (32K): [in this window] [in a new window]   Figure 7.   Effects of protein synthesis inhibition on cytokine induction and dexamethasone suppression of MCP-4 RNA accumulation. Northern analysis of 20 µg of total RNA harvested from A549 cells at 4 h after the indicated treatments with 0.1 ng/ml IL-1 (A) or 10 ng/ml TNF- (B). For cells treated with cycloheximide (CHX) (10 µg/ml) and/or dexamethasone (DEX) (1 µM) plus cytokine, the cycloheximide and dexamethasone were added 30 min before the addition of the cytokine. Below each blot a quantitative comparison is made at each dose between the MCP-4-specific signal and the GAPDH-specific signal. US, unstimulated.

MCP-4 mRNA Expression in PBMCs, Monocytes, and Lymphocytes

Since we observed the expression of MCP-4 mRNA and protein in mononuclear cells in the submucosa of airways isolated from patients with asthma, we sought to examine the regulation of MCP-4 expression in these cells ex vivo. MCP-4 mRNA was induced in PBMCs isolated from normal volunteers with the same cytokines that induced its expression in epithelial cells, namely IL-1 and TNF-. Kinetic analysis revealed that the time course of MCP-4 mRNA accumulation was also similar to epithelial cells with a peak accumulation at 2-8 h with levels decreasing thereafter (Figure 8A). The expression of MCP-4 mRNA was further evaluated in purified monocytes and lymphocytes stimulated with IL-1. MCP-4 expression was highly induced in purified peripheral blood monocytes (Figure 8B). Peak MCP-4 expression was seen by 2-8 h in monocytes and levels decreased to baseline over 48 h (Figure 8B). MCP-4 mRNA accumulation was induced to lower levels in purified peripheral blood lymphocytes over the 48 h analyzed (Figure 8C).

View larger version (37K): [in this window] [in a new window]   Figure 8.   Time course of IL-1-induced MCP-4 mRNA accumulation in PBMCs (A), monocytes (B), and lymphocytes (C). Northern analysis of 10 µg of total RNA harvested from PBMCs at various times after treatment with 0.1 ng/ml IL-1. To control for RNA loading a quantitative comparison at each time point between the MCP-4-specific signal and the GAPDH-specific signal is presented below each blot as mean (n = 3 independent donors) ± SEM.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The recruitment and activation of leukocytes in the airways are thought to be a critical step in the pathophysiology of asthma. The molecular mechanisms responsible for the selective recruitment of Th2 lymphocytes and eosinophils into the airways of patients with asthma are incompletely understood and of considerable interest. As chemokines are specific and potent leukocyte chemoattractants, they make good candidates to be the molecular regulators responsible for linking antigen-specific immune activation and leukocyte influx into the airways. The first step in assessing the role of a given chemokine in a disease process is to determine its pattern of expression in diseased tissue. In this regard, we have determined that MCP-4, a chemokine with activities on a number of cells implicated in the pathophysiology of asthma, is upregulated in the bronchial epithelium, submucosa, and BAL of individuals with asthma compared with normal control subjects.

While chemokines do have overlapping activities on cells in vitro, studies using neutralizing chemokine antibodies and chemokine-deficient mice have clearly revealed that in vitro redundancy does not predict overlapping roles in vivo (3). It is likely that timing and pattern of expression in vivo will be critical determinants of a chemokine's function in vivo. In addition, it is also clear that rodent studies need to be correlated to human studies for several reasons, including the inadequacies of rodent models of a human disease and the inherent differences between species. For example, to date there has been no orthologue of MCP-4 isolated in rodents. Therefore, if one is interested in understanding the role of a chemokine in a human disease, its expression will need to be evaluated in that disease.

In this regard, we have found that MCP-4 mRNA and protein are expressed in both the epithelium and infiltrating submucosal mononuclear cells in bronchial biopsies of patients with asthma. Recently, we have extended these observations by demonstrating that MCP-4 expression is also increased in the small airways of subjects with asthma compared with control subjects without asthma (28). To determine the role of MCP-4 in asthma-associated immune response, we analyzed its relationship with eotaxin and inflammatory cells, with emphasis on BAL fluid as potential source of chemokines and proinflammatory cytokines. Although no relationship was found between MCP-4 and neutrophils, lymphocytes, or macrophages, a close relationship was observed between MCP-4 and eosinophils. This selective relationship between MCP-4 and eosinophils was further established with positive correlation between MCP-4 and eotaxin, which is also elevated in patients with asthma (6) and correlates with the degree of airway hyperresponsiveness (7). CCR3 mRNA and protein expression in bronchial biopsies has also been reported to be elevated in patients with asthma (7). The role of CCR3 agonists in regulating the movement of eosinophils into the allergic lung has begun to be elucidated using mice deficient in eotaxin (29, 30) and using neutralizing eotaxin antiserum (31, 32). Although the results of these studies are not in complete agreement, it is clear that eotaxin only partially contributes to the recruitment of eosinophils into the allergic lung. It is thus likely that other eosinophil chemoattractants participate in this process. This is consistent with our previous observation that neutralizing antibodies to MCP-4 or eotaxin individually only partially neutralized (10-30%) the eosinophil chemotactic activity contained in the BAL fluid recovered from individuals with asthma (6). During the preparation of our re-vised manuscript, and in agreement with our studies, it was reported that multiple CCR3 ligands, including eotaxin, eotaxin-2, MCP-3, MCP-4, and RANTES, are all expressed in bronchial biopsies of patients with atopic and nonatopic asthma (33). Thus, multiple CCR3 ligands are likely to contribute to the eosinophil chemotactic activity found in the airway of patients with asthma. However, it is also likely that chemoattractants that are not active on CCR3 also contribute to the eosinophil chemotactic activity found in airways of patients with asthma as a neutralizing mAb to CCR3 inhibited only 50% of the eosinophil chemotactic activity found in allergic nasal polyp tissue fluid (34). Thus, although our studies suggest that MCP-4 is involved in human asthma and likely plays a role in the recruitment of eosinophils into the allergic lung, it appears that it does so in collaboration with other chemoattractants.

In the chemokine system, the apparent in vitro redundancy has not predicted the unique biological functions of individual chemokines in animal models. In general terms, this may relate to differential expression of chemokines in situ and/or differential signaling of the same chemokine receptor by individual chemokine ligands. There is recent evidence that both of these hypotheses may be operational for CCR3 ligands. For example, following allergen-induced late phase cutaneous responses in atopic subjects, eotaxin was associated with early 6 h eosinophils, whereas eotaxin-2 and MCP-4 were associated with later 24 h tissue eosinophilia (35). It has also recently been shown that of the CCR3 ligands, MCP-4 may uniquely activate the eosinophil. Whereas eotaxin and eotaxin-2 induced a sustained change in eosinophil shape, MCP-4 induced a transient shape change (36). Thus, it appears that there may be differential activation of CCR3 by its different ligands leading to unique biological responses.

It is becoming increasingly clear that the airway epithelium participates in the inflammatory process by producing a number of inflammatory mediators including chemokines such as IL-8 (37), RANTES (38), MCP-1 (4), MCP-3 (5), eotaxin (27), and MCP-4 (12, 13). Like eotaxin, TNF- and IL-1 appear to be the main inducers of MCP-4 in epithelial cells in vitro. Although both IL-1 and TNF- have been associated with asthmatic inflammation (21, 39), it remains to be determined, however, what the physiological inducer(s) of chemokine expression in vivo are. In the mouse, passive transfer of Ag-specific Th2 cells (but not Th1 cells) to an unsensitized mouse will transfer sensitivity to aerosol Ag challenge resulting in pulmonary eosinophilia, airway hyperreactivity, and eotaxin expression (40, 41). The mechanism by which Th2 lymphocytes lead to the induction of chemokines from airway epithelium is an important unanswered question and recent studies have suggested that IL-13 may play an important role (42).

Although IFN- alone did not induce the expression of MCP-4 in epithelial cells, IFN- did synergize with TNF- and IL-1 in inducing the accumulation of MCP-4 mRNA. This effect of IFN- on MCP-4 induction was very similar to the effect of IFN- on eotaxin expression in epithelial cells (27, 43). These findings suggest that IFN- produced from Th1 cells can participate in the recruitment of cells typically associated with a Th2 response. This may be relevant to the airway inflammation and airflow obstruction that follow nonallergic conditions, such as viral-induced exacerbations of asthma, that are known to induce airway cytokine and IFN- production.

MCP-4 induction is a primary response to TNF- and IL-1 stimulation in epithelial cells. In fact, inhibition of protein synthesis alone weakly induced the accumulation of MCP-4 mRNA and augmented TNF- and IL-1 induction of MCP-4 mRNA accumulation in A549 epithelial cells. This has been seen for numerous primary response cytokines and suggests that a protein with a short half-life inhibits the transcription of MCP-4 or degrades the MCP-4 mRNA. The plausibility of the later possibility is supported by the fact that MCP-4 mRNA contains AU-rich motifs in its 3' untranslated region that are believed to be responsible for decreasing mRNA stability of cytokine genes.

Glucocorticoids are antiinflammatory steroids that are widely used in the therapy of asthma, where they decrease airway inflammation and restore normal airway tone (44). The mechanism of the glucocorticoid effect is multifactorial involving inhibition of both T-cell function and eosinophil recruitment. We have demonstrated that the glucocorticoid dexamethasone inhibited both TNF- and IL-1 induction of MCP-4 in A549 epithelial cells. This is consistent with the observation that the glucocorticoid budesonide inhibited TNF- induced MCP-4 accumulation in BEAS-2B cell line (13). In addition, our studies, using the protein synthesis inhibitor cycloheximide, revealed that part of dexamethasone's suppressive effect requires new protein synthesis and part of its effect is independent of new protein synthesis. This suggests that glucocorticoids can act directly to suppress MCP-4 mRNA accumulation, which can be at the level of transcription and/or mRNA stability. Recently, it has been shown that budesonide inhibited both RANTES and eotaxin promoter-driven reporter gene activity and selectively accelerated the decay of eotaxin and MCP-4 mRNA (45). Glucocorticoids downmodulate the expression of several chemokines in epithelial cells implicated in asthmatic inflammation, including eotaxin (27), MCP-1 (46), RANTES (38), and IL-8 (47). These findings suggest that epithelial cells are an important source of proinflammatory chemokine production in the airways and a potentially important target of the antiinflammatory effects of inhaled glucocorticoids, which are widely used for the treatment of asthma. In fact, systemic treatment of patients with glucocorticoids reduced nasal polyp mRNA levels for eotaxin, eotaxin-2, and MCP-4 to levels found in normal turbinate mucosa (34). In addition, we have found that topical corticosteroids inhibit allergen-induced expression of MCP-4 in nasal tissue in patients with allergic rhinitis (25).

MCP-4 mRNA and protein was also upregulated in mononuclear cells in the airway submucosa and BAL cells of patients with asthma compared with normal control subjects. To study the regulation of MCP-4 in mononuclear cells the expression of MCP-4 was evaluated in peripheral blood mononuclear cells and purified monocytes and lymphocytes. The time course of MCP-4 expression in these cells after IL-1 and TNF- stimulation was similar to that seen in A549 cells with levels detectable 2 h after stimulation. Chemokines produced in this time frame could certainly contribute to the airway inflammation seen following antigen challenge.

The correlation of MCP-4 and eotaxin protein in the BAL fluid from patients with asthma and the similar pattern of MCP-4 and eotaxin protein and mRNA staining seen on biopsies taken from patients with asthma suggest that these two chemokines may be coordinately regulated in vivo in patients with asthma. This is consistent with our in vitro findings that document a similar pattern of MCP-4 and eotaxin induction in epithelial cells, mononuclear cells, and endothelial cells. Thus, our results suggest that in humans with asthma, MCP-4 and eotaxin contribute to eosinophil tissue influx, and thus the pathogenesis of this disease.