INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pleurodesis involves the introduction of sclerosing agents into the pleural space to relieve dyspnea in patients with recurrent pleural effusions. A wide variety of sclerosing agents in various forms of solutions, slurries, and powders, including tetracycline, bleomycin, Corynebacterium parvum, and nitrogen mustard have been introduced into the pleural space with varying response rates (1). Most recently, talc has been viewed as the sclerosing agent of choice to cause pleurodesis in patients with symptomatic malignant pleural effusions (2, 3). In general, pleural sclerosing agents cause an acute inflammatory response in the pleural space, with a predominance of pleural-fluid leukocytes noted early in the process of pleural symphysis (4, 5). Although leukocytes are noted early in the process of pleural symphysis, the mechanism of leukocyte recruitment remains unknown.

Mesothelial cells are metabolically active cells that continuously line the pleura in a monolayer. They are the first cell type to encounter foreign substances instilled into the pleural space, and are therefore likely candidates for initiating and propagating an acute inflammatory reaction. Recent investigations suggest that mesothelial cells initiate pleural inflammation in response to various cytokines or particulates, at least in part by production and release of C-X-C and/or C-C chemokines and expression of adhesion molecules (6). Interleukin-8 (IL-8) (a C-X-C chemokine), is an 8.5-kD peptide, preferentially attracts and activates neutrophils. Monocyte chemotactic protein-1 (MCP-1) is an 8.7-kD protein that belongs to the C-C supergene family of chemokines (9), and has specific chemoattractant and activating activity for monocytes in acute inflammatory conditions. Monocytes are the predominant cellular sources of IL-8 and MCP-1 in acute pleural inflammatory processes, but these chemokines can be produced by nonimmune cells in response to endogenous and exogenous stimuli (10, 11). Recently, mesothelial cells were observed to produce C-C and C-X-C chemokines upon stimulation by inflammatory mediators (12); however, their role in talc-mediated pleurodesis is still undefined.

The pleural inflammatory response is a tightly organized sequence of events controlled in part by the regulated expression of cell-surface and soluble molecules (8, 12, 13). Adhesion molecules expressed on the surfaces of various cell types are important in cell-cell interactions. Adhesion molecules play a key role in cellular traffic through the pleural mesothelium, particularly during the inflammatory response, when leukocytes migrate from the peripheral circulation to the pleural space (14). ICAM-1 is involved in cell-specific extravasation into inflamed tissue. ICAM-1 expression on pleural mesothelial cells (PMC), and its binding with natural ligand (CD11a/CD18) receptors on leukocytes, may influence the local inflammatory process (8, 15).

In the present investigation, the concept of mesothelial cells as an important component to the early host response to talc was applied to the understanding of talc-induced pleurodesis. We found that talc induces biologically active chemokine synthesis and expression of ICAM-1 on PMC. These observations suggest that PMC, by producing chemokines and expressing ICAM-1, participate in the regulation of leukocyte migration into the pleural space, and may therefore actively participate in talc-mediated pleurodesis.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Isolation and Culture of Human PMC

Pleural fluid was obtained via thoracentesis from patients with transudative pleural effusions resulting from congestive heart failure (CHF), according to a protocol approved by the Indiana University Institutional Review Board. The majority of patients had intractable CHF and symptomatic pleural effusions. None of the subjects had evidence of an infectious etiology for their pleural effusions. The pleural fluid was removed into a heparinized container and centrifuged at 1,000 × g for 10 min, and the supernatant was discarded. The cell pellet was briefly exposed to cold hypotonic solution to lyse red blood cells (RBCs). The cells were resuspended in Ham's 199 culture medium (Gibco Laboratories, Grand Island, NY) containing 15% fetal bovine serum (FBS) (Harlan Bioproducts, Indianapolis, IN), penicillin (100 U/ml), and streptomycin (100 µg/ml). The cells were plated in 75 cm² culture flasks (Corning Costar Corporation, Acton, MA) and incubated overnight at 37° C in 5% CO₂/95% air. On the following day the medium was changed to remove nonadherent cells. The mesothelial cells were characterized by the presence of classic cobblestone morphology (16), absence of Factor VIII antigen, and presence of cytokeratin (17). When the cells were confluent they were trypsinized and seeded into 24-well culture plates as required for different assays. All cells were utilized between the second and fourth passages.

Particle Preparation

Talc (3 MgO · 4 SiO₂ · H₂O) (Humco Laboratory, Texarkana, TX) particles were suspended in endotoxin-free H₂O at a concentration of 4,000 µg/ml. Particle size was 2.1 ± 0.89 µm as determined with a Sony CCD-IRIS/RGB video camera (Sony Corp., Tokyo, Japan) attached to an Olympus IMT-2 microscope (Olympus Corp., Tokyo, Japan) interfaced with a DataStar 486-66 computer containing Image-Pro Plus software (Media Cybernetics, Silver Spring, MD). Glass (SiO₂) microspheres (Sigma Chemical Co., St. Louis, MO) with a certified mean diameter of 2.1 ± 0.5 µm were processed in a similar manner as the talc particles. The particles were washed and then autoclave-sterilized. The concentrated stock samples of talc and glass microspheres had undetectable levels of endotoxin as determined with the Limulus amebocyte lysate assay (Sigma).

Talc Stimulation and PMC Viability

Confluent PMC were exposed to a range of (2, 4, 8, 16, 32, and 64 µg/ cm²) talc concentrations in serum-free medium in 24-well culture plates. Cell viability was tested at the end of 72 h by trypan blue dye exclusion. On the basis of cell viability, the confluent PMC were subsequently exposed to graded concentrations of 2 to 64 µg/cm² of talc in serum-free Ham's 199 medium for 24 h in order to establish a dose- response curve. The PMC culture medium was centrifuged briefly to remove excess talc, and the chemokine levels were estimated. On the basis of the dose-response curve, an optimal concentration of 4 µg/ cm² of talc was selected for further study. In order to establish the time-dependent response, 4 µg/cm² of talc was applied to confluent mesothelial cells in 24-well plates for time intervals of 1 to 72 h, and the mesothelial cell culture medium was collected at the end of each time period and tested for chemokine release. To evaluate the specificity of the effect of talc, 4 µg/cm² of glass microspheres of a similar size were included as parallel controls.

Enzyme-linked Immunosorbent Assays for IL-8 and MCP-1

The IL-8 and MCP-1 levels of PMC culture supernatant were quantified with sandwich-type enzyme-linked immunoassays (Quantikine; R&D Systems, Minneapolis, MN). For the IL-8 and MCP-1 ELISAs, flat-bottomed 96-well microtiter plates were coated with an excess murine monoclonal antibody to either IL-8 or MCP-1. The PMC culture medium was then added in serial dilutions. After incubating, any unbound protein was removed by washing with phosphate-buffered saline (PBS) and an enzyme-linked polyclonal antibody specific to either IL-8 or MCP-1 was added. The antibody "sandwiches" the antigen that was immobilized during the first incubation period. After again washing with PBS, substrate solution was added to the wells. The presence of IL-8 or MCP-1 was quantitated by comparing the optical density (OD) of the samples with the standard curve.

Isolation of Human PBM and Neutrophils

Mononuclear cells and neutrophils were isolated as reported earlier (12). Venous blood was collected from healthy volunteers into heparinized tubes. The heparinized blood was overlayered on a Ficoll- Hypaque density gradient (Histopaque; Sigma) and centrifuged at 1,600 × g for 20 min at 4° C. The buffy coat between the liquids, containing mononuclear cells, was washed three times in Hanks' balanced salt solution (HBSS). The cells were tested for viability. The resulting cell population contained > 95% viable mononuclear cells. Neutrophils were isolated from normal healthy volunteers via phlebotomy. The heparinized blood was allowed to sediment after the addition of 70% dextran. The neutrophils were purified by differential centrifugation with Histopaque (Sigma). Contaminating erythrocytes were removed by brief exposure to cold hypotonic solution. The neutrophils were then washed and resuspended at a concentration of 2.0 × 10⁶ cells/ml in HBSS.

Assessment of Chemokine Bioactivity by Chemotaxis

To evaluate whether the IL-8 and MCP-1 released from cells were biologically active, neutrophil and monocyte chemotactic activity was determined in the supernatant of talc-stimulated PMC, using modified Boyden chambers as previously described (12). The mesothelial cell supernatant or the appropriate control was added to the lower compartment, and a 3.0-µm polycarbonate filter (Nuclepore Corporation, Pleasanton, CA) was placed and secured in the assemblage to separate the upper and lower compartments. After a 30-min incubation period, 500 µl of the neutrophil/monocyte solution was added to the upper compartment and incubated for 90 min at 37° C in an atmosphere of 5% CO₂. The filters were removed, fixed, stained with Diff-Quik (Harleco, Gibbstown, NJ) and mounted on glass slides. The chemotactic response was quantitated as the "chemotactic index" and defined as the number of migrated neutrophils/monocytes on the inverse surface of the filter in 10 high power fields (hpf). Ten percent zymosan-activated serum (ZAS) and N-formyl-methionyl-leucyl phenylalanine (FMLP) (10⁷ M) served as positive controls for neutrophils and monocytes, respectively, and HBSS served as the negative control. Purified mouse antihuman IL-8 or mouse-antihuman MCP-1 antibodies (R&D Systems) was added in excess to some of the lower chambers to determine whether this would decrease the chemotactic index. Mouse IgG isotype (Sigma) was added to some of the lower compartments as a nonspecific binding agent.

Isolation of PMC RNA and Reverse Transcriptase-Polymerase Chain Reaction

Total cellular RNA was isolated from PMC using Tri-reagent (phenol and guanidine-thiocyanate; Molecular Research Center Inc., Cincinnati, OH) (18) according to the manufacturer's recommendations. One microgram of total RNA was reverse transcribed into complementary DNA (cDNA). The first strand of cDNA was synthesized in a total volume of 20 µl in the presence of 5 mmol/L MgCl₂; 50 mmol/L KCl; 10 mmol/L Tris-HCl, pH 8.3; 1 mmol/L deoxynucleotide triphosphates (dNTPs); 1 U/ml ribonuclease (RNase) inhibitor; primer (15 µM); and 2.5 U/ml of murine leukemia virus (MuLV) reverse transcriptase (RT) (Perkin Elmer Cetus, Norwalk, CT). The reverse transcription was conducted at 42° C for 15 min, and the reaction was stopped by incubation at 99° C for 5 min.

The cDNA was then amplified, using specific primers for human IL-8 (hIL-8) (GeneBank accession no. Y00787) and hMCP-1 (GeneBank accession no. S69738). Human -actin was amplified as a positive control. Primers for amplification were synthesized by the phosphoroamidite method with a Beckman 200 A synthesizer (Beckman Instruments, Fullerton, CA), and were found to be specific for the chemokine being investigated, and not for any other chemokine family member. The primers used were 5'-GCACTCTTCCAGCCTTCCTTCC-3' (sense) and 5'-TGCTTGCTGATCCACATCTGCT-3' (antisense) for human -actin; 5'-TGAAATATCCAGAACATACTTA-3' (sense) and 5'-GCAAAATTTATTGTCCCATCAT-3' (antisense) for IL-8; and 5'-TGCTGCTATAACTTCACCAATA-3' (sense) and 5'-TGGGGAAAGCTAGGGGAAAATA-3' (antisense) for MCP-1. The polymerase chain reaction (PCR) was performed with 5 µl of reverse-transcribed product in a reaction mixture containing 2 mmol/L MgCl₂; 50 mmol/L KCl; 10 mmol/L Tris-HCl, pH 8.30; specific oligonucleotide primers (15 µM); and 2.5 U Taq DNA polymerase (Perkin Elmer Cetus). The samples were amplified in a thermal cycler (GeneAmp PCR System 9600; Perkin Elmer Cetus) preheated for 90 s at 95° C, followed by 30 cycles. Each cycle was conducted at 95° C for 15 s of denaturation, 58° C for 30 s of primer annealing for MCP-1 and 50° C for 30 s for IL-8, and extension for 30 s at 72° C. The amplified product obtained was subjected to electrophoretic analysis. RT-PCR products were separated in 3% agarose gel with TBE (Tris, 21 g/L, boric acid, 11 g/L¹, and ethylenediamine tetraacetic acid (EDTA) 0.002 M, pH 8.00) running buffer. The gels were stained with ethidium bromide, and the bands were visualized under ultraviolet (UV) light and photographed on polaroid film.

Detection of ICAM-1 Expression with Flow Cytometry

The mesothelial cells were stimulated with talc (4 µg/cm²) for 24 h at 37° C in a 5% CO₂ atmosphere. PMC were also stimulated with glass beads (4 µg/cm²) as an internal control and tumor necrosis factor- (TNF-) (20 ng/ml) as a positive control. The PMC were trypsinized and washed three times in PBS with 5% bovine serum albumin (BSA) and 5 mM sodium azide, and were incubated for 45 min at 4° C, in the presence of either mouse monoclonal antihuman ICAM-1 (1 µg/10⁶ cells) antibody or mouse IgG isotype. Cells were washed three times and labeled with rabbit antimouse IgG₁ FITC conjugate to detect the antibody bound to the antigens, with a similar isotype included as a positive control. After incubation, the cells were washed three times and fixed in 4% paraformaldehyde. The fluorescence associated with the cells was analyzed flow cytometrically using a FACStar (Becton-Dickinson Immunocytometry Systems, Mountain View, CA). Fluorescence data were collected on a log scale, and the relative fluorescence intensity was reported by comparing the cells' light scattering characteristics with those of normal cells analyzed in the same experiment.

Statistical Analysis

Data were analyzed with the Sigmastat statistical software package (Apple Computer, Cupertino, CA) and expressed as mean ± SEM. The difference between the group means were analyzed by analysis of variance (ANOVA), with use of the student-Newman-Keuls test. Data were considered statistically significant at p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Isolated Pleural-fluid Cells are Characterized as PMC

In general, approximately 200 ml of centrifuged transudative pleural fluid yields a 75-cm² confluent flask of mesothelial cells within 7 to 14 d when cultured as described. The mesothelial origin of cells in the second passage was confirmed by positive staining for vimentin, cytokeratin, and hyaluronic acid mucin (12). The cell morphology was documented as having a cobblestone pattern through phase-contrast microscopy, and numerous microvilli were noted on transmission electron microscopy. All cells were utilized between the second and fourth passage in 24-well plates.

Talc Stimulates Mesothelial Cells to Release C-X-C and C-C Chemokines

Confluent PMC cultures were stimulated with varying doses of talc (2 to 64 µg/cm²) for 24 h in tissue-culture plates. Cell viability was documented by trypan blue dye exclusion and visual inspection with phase-contrast microscopy. Viable cells were expressed as the percent viable cells of all cells (Table 1). The PMC viability decreased with increasing talc concentration. The PMC viability found with a talc concentration of 64 µg/cm² was about 75%. Cultured mesothelial cells release small amounts of chemokine constitutively; however, all concentrations of talc used in the study stimulated the mesothelial cells to significantly increase their release of IL-8 and MCP-1 over that of unstimulated mesothelial cells. A talc concentration of 4 µg/cm² was utilized in further studies because this dose resulted in the most significant chemokine release while maintaining a > 95% PMC viability (Figure 1). A time-course analysis revealed that talc-stimulated mesothelial cells released more IL-8 and MCP-1 than did unstimulated mesothelial cells beginning at the 3-h time point, and that this continued for up to 72 h. Glass beads of a similar size distribution as talc were also tested, to determine whether the mesothelial-cell chemokine production was specific to talc. Talc-stimulated mesothelial cells produced IL-8 and MCP-1 (Figures 2 and 3) to a significantly greater degree than did cells treated with glass beads, and both particles caused an increase in IL-8 and MCP-1 chemokine release over that of unstimulated cells. Because chemokine levels plateaued at 24 h, this time period was used for the remaining experiments unless otherwise indicated.

View larger version (31K): [in this window] [in a new window]   Figure 1.   Dose-dependent production of IL-8-immunoreactive antigenic protein in PMC as measured with a specific ELISA. PMC were stimulated with varying concentrations of talc for 24 h, and the IL-8 levels in the culture media were estimated. The data shown represent mean ± SEM of six individual studies. The values are significant *(p < 0.001) when compared with control values.

View larger version (22K): [in this window] [in a new window]   Figure 2.   Time course of talc-induced antigenic IL-8 production in PMC. PMC were incubated either in serum-free medium or with 4 µg/cm2 of talc or glass beads. Culture supernatants were collected at each time point and analyzed for immunoreactive IL-8 with a specific ELISA. The data shown at each time point represent the mean ± SEM of six individual experiments. The values are significant (*p < 0.001) when compared with those for serum-free medium and when (!p < 0.001) compared with those for glass beads.

View larger version (27K): [in this window] [in a new window]   Figure 3.   Time course of talc-induced MCP-1 chemokine production in PMC. PMC were incubated either in serum-free medium or with 4 µg/cm2 of talc or glass beads. Culture supernatants were collected at each time point and analyzed for immunoreactive MCP-1 with a specific ELISA. The data shown at each time point represent the mean ± SEM of six individual experiments. The values are significant (*p < 0.001) when compared with those for serum-free medium and (!p < 0.001) compared with those for glass beads.

Neutrophil/Monocyte Chemotaxis

The supernatants from talc-stimulated and unstimulated mesothelial cells were evaluated for chemotactic activity for human neutrophils and monocytes. The culture media from the talc-stimulated mesothelial cells had chemotactic activity comparable to that of the potent chemotactic agent ZAS (p = 0.534). The chemotactic index of the culture media from the talc-stimulated mesothelial cells was significantly greater (p < 0.001) than that from unstimulated mesothelial-cell supernatant (Tables 2 and 3). The chemotactic activity of the talc-stimulated supernatant was significantly reduced, but did not entirely return to the level of unstimulated cells upon the addition of excess monoclonal IL-8 antibody (Table 2) or excess MCP-1 antibody (Table 3). The chemotactic activity of PMC culture supernatant was blocked by 44.2% with IL-8-specific antibody, and by 55.7% with MCP-1-specific antibody, demonstrating that PMC-derived chemokines are bioactive. An excess of IgG isotype had no statistically significant effect on the chemotactic index of the supernatant from the talc-stimulated mesothelial cells.

Talc Stimulates C-X-C and C-C Chemokine mRNA Expression in PMC

PMC were stimulated with talc for 24 h, and 1 µg of total RNA was reverse transcribed and amplified with the PCR. The product lengths selected were 235 bp and 259 bp for IL-8 and MCP-1, respectively. Figures 4 and 5 show that talc induced C-X-C and C-C chemokine expression in PMC, as observed by ethidium bromide staining of RT-PCR product. The transcriptional response of IL-8 and MCP-1 expression was enhanced in talc-stimulated PMC. Mesothelial cells, when either stimulated with glass beads or incubated in the presence of serum-free medium did not show any chemokine expression.

View larger version (77K): [in this window] [in a new window]   Figure 4.   Talc-induced expression of IL-8 mRNA in PMC. PCR-cDNA gel showing 235-bp IL-8-specific amplification product stained with ethidium bromide. -Actin (302 bp) served as a positive control in the RT-PCR. Lane 1: Hae III molecular-weight marker; Lane 2: unstimulated PMC; Lane 3: glass-bead-stimulated PMC; Lane 4: talc-stimulated PMC. These results are single representative results of three individual experiments.

View larger version (74K): [in this window] [in a new window]   Figure 5.   Talc-induced expression of MCP-1 mRNA in PMC. PCR-cDNA gel showing 259-bp MCP-1-specific amplification product stained with ethidium bromide. -Actin (302 bp) served as a positive control in the RT-PCR. Lane 1: Hae III molecular-weight marker; Lane 2: unstimulated PMC; Lane 3: glass-bead-stimulated PMC; Lane 4: talc-stimulated PMC. These are single representative results of three individual experiments.

Talc Induces ICAM-1 Expression on PMC

Figures 6A through 6D, show ICAM-1 expression on PMC as evaluated by fluorescence-activated cell sorting (FACS) analysis. A total of 10,000 cells were gated in a FACstar Sorter and positive events were collected on a log scale. Talc-stimulated PMC were 68.82% ICAM-1 positive, and 6.46% showed activity with a nonspecific isotype antibody (mouse IgG antibody) (p < 0.001). PMC stimulated with glass beads used as an internal control were 21.94% ICAM-1 positive, and those stimulated with TNF- as a positive control were 87.63% ICAM-1 positive.

View larger version (35K): [in this window] [in a new window]   Figure 6.   Expression of ICAM-1 on PMC. Flow-cytofluorometric analysis of stimulated PMC was done with a control nonbinding monoclonal antiisotype antibody (A), or a monoclonal anti-ICAM-1 antibody (B, C, and D). The horizontal axis measures the surface density of the antigen on a 4-log scale; the vertical axis measures the cell number.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Inducing chemical pleurodesis is a common practice in clinical pulmonary medicine, yet the mechanism of the inflammatory response induced by the sclerosing agent remains poorly defined. A diverse array of pharmaceutical and physical agents have been used in the pleural space to induce pleural symphysis. These materials include various antineoplastic agents, antibiotics, radioactive elements, and naturally occurring particles such as talcum powder. The lack of a unifying characteristic among sclerosing agents led to investigation for a common physical property shared by these agents. The acid/base status of the sclerosing agent was explored and disproved as a common link that leads to pleurodesis (4). Subsequently, research was directed at identifying a common host response to explain this phenomenon. Several in vivo studies evaluated pleural-fluid results before and after induced inflammation. Agrenius and coworkers documented a decrease in fibrinolytic activity during induced inflammation, a decrease in pleural-fluid D-dimer levels, and an increase in plasminogen-activator inhibitor following intrapleural treatment with quinacrine (19). Rossi and associates disproved the concept that pleurodesis occurred as the result of local enhancement of cell-mediated immunity (5). However, they did document a change in pleural-fluid leukocyte differential counts, with an increase in the proportion of neutrophils following intrapleural injection of Corynebacterium parvum. Other in vivo studies confirmed the presence of pleural-fluid neutrophilia, and associated this with an increase in several proinflammatory cytokines, including TNF, IL-6 and IL-8, although the source of these cytokines was not fully elucidated (20).

The purpose of the present investigation was to determine whether mesothelial cells responded to talc, a potent sclerosing agent, by producing C-X-C and C-C chemokines and expressing adhesion molecules in vitro. We selected these chemokines for several reasons. IL-8 has been implicated as a major neutrophil chemotactic factor that mediates neutrophil trafficking in many lung-parenchymal and pleural diseases. MCP-1 is chemotactic for monocytes (7, 21). The attraction and accumulation of neutrophils and monocytes is an essential component of the inflammatory process. The role of the mesothelial cell as a regulatory cell, capable of secreting a wide variety of cytokines that regulate leukocyte traffic between the circulation and the pleural space, has emerged over the past decade. Recently, several investigators have shown that mesothelial cells are capable of producing IL-8 and MCP-1 (8, 12). Goodman and associates revealed that mesothelial cells produce IL-8 in response to TNF-, IL-1, and interferon- (IFN-) (7). Recent studies revealed that PMC release IL-8 in a polarized fashion (22). PMC stimulated with asbestos were found to release factors chemotactic for neutrophils (13). Boylan and coworkers concluded that asbestos stimulates mesothelial cells to synthesize IL-8, which plays an important role in asbestos-induced pleurisy in the rabbit (6). Griffith and associates documented the role of IL-1 in asbestos-induced release of IL-8 by mesothelial cells (23). Earlier studies demonstrated that PMC express C-C and C-X-C chemokines in response to bacterial endotoxins, lipopolysaccharide (LPS), IL-1, and TNF- (12, 24).

We found that mesothelial-cell release of IL-8 and MCP-1 increased in a time- and dose-dependent manner when these cells were exposed to talc. It is possible that similar elements in the chemical composition, geometry, or crystalline structure of particles may account for the accentuated chemokine production. When tested, glass beads of a similar size to talc particles in our control experiments did not stimulate PMC chemokine production as much as did talc (Figures 2 and 3). This testing demonstrated that the large quantity of IL-8 and MCP-1 produced is a specific response to talc particles and is not generalized to all particulates. Talc induces superoxide-anion generation (25), but the mechanisms of talc-mediated release of chemokines remains to be understood. Our RT-PCR studies also confirmed (Figures 4 and 5), that IL-8 and MCP-1 mRNA expression was specific to talc stimulation. Previous investigators have shown a similar difference in IL-8 release by mesothelial cells exposed to asbestos as opposed to other control particles (6, 23).

Neutrophil chemotactic assays confirmed that the IL-8 released by mesothelial cells upon exposure to talc was biologically active. To determine the relative contribution of IL-8 to neutrophil chemotaxis, the culture media in our study were preincubated with control isotype or antihuman IL-8 antibody. Although the chemotactic index did not return to the level of culture media from unstimulated mesothelial cells media with the addition of a neutralizing monoclonal antibody to IL-8, it was decreased by 44.2%, suggesting that most of the observed neutrophil chemotaxis was induced by IL-8 (Table 2). This is consistent with previously published in vivo data obtained with pleural fluid from patients with empyema, in which neutrophil migration was decreased by 32% to 65% (26). Because a monoclonal antibody specific to human IL-8 was used in our experiments, cross-reactivity with other potential neutrophil chemoattractants was unlikely to have occurred.

MCP-1 was both constitutively expressed in mesothelial cells and, by these cells when they were stimulated with talc in a time-dependent manner (Figure 3). Mesothelial-cell-derived MCP-1 may play a role in mononuclear-cell recruitment into the pleural space. Malignant pleural effusions are characteristic of mononuclear-cell infiltration, and contain high levels of MCP-1 (27). In malignant pleural disease, the malignant cells themselves may contribute to the total MCP-1 found in the pleural fluid (27, 28). However, mesothelial-cell expression of MCP-1 in response to talc suggests that mesothelial cells play a role by producing MCP-1 in talc-induced pleurodesis. Talc-stimulated PMC produced significantly higher levels of MCP-1 than did PMC stimulated with glass beads (Figure 3). Culture medium from talc-stimulated PMC was found to be chemotactic for monocytes (Table 3). Neutralization of MCP-1 bioactivity with specific antibody resulted in a 55.7% reduction of overall monocyte chemotactic bioactivity. These findings suggest that the MCP-1 released by PMC is biologically active and is a major source of mononuclear-cell chemotactic activity. The data demonstrate that talc-mediated release of PMC chemokines is an important factor in inducing and amplifying the inflammatory process initiated by talc.

Intercellular adhesion molecules play a critical role in cell trafficking from the vascular compartment to the pleural space, particularly during the inflammatory process (29). Mesothelial cells have been shown to express both ICAM-1 and vascular cell adhesion molecule-1 (VCAM-1) (8, 13). Since the development of an inflammatory exudate is one of the major and immediate consequences of talc, we also investigated the effect of this sclerosing agent on the cell-surface expression of ICAM-1. ICAM-1 expression was upregulated in mesothelial cells upon talc stimulation, whereas glass beads did not show a significant effect (Figure 6). ICAM-1 is involved in the transmigration of neutrophils and mononuclear cells from the vascular compartment to the site of inflammation (30). In earlier studies we demonstrated that neutrophil transmigration across mesothelial monolayers was dependent on mesothelial ICAM-1 expression (14). Others studies showed that ICAM-1 expression is upregulated on exposure to TNF- or IL-1, and causes increased binding of monocytes to mesothelial cells (8) Chemokines are known to modulate the expression of adhesion molecule (31), and because talc induces IL-8 and MCP-1 release in PMC, it is tempting to speculate that talc-induced chemokines may regulate ICAM-1 expression on PMC and may thus facilitate neutrophil and mononuclear infiltration into the pleural space.

Increased knowledge in the area of talc-induced pleural fibrosis has several important clinical implications. Through a better understanding of the inflammatory process in the pleural space, we may be able to alter this process to achieve desirable clinical outcomes, such as effective pleurodesis in cases of intractable pleural effusions, or preventing fibrothorax in cases of empyema. Talc-induced mesothelial-cell production of IL-8 may be only one of the initial events in the process of talc-induced pleural inflammation. It is possible that after neutrophils arrive in the pleural space, they amplify and perpetuate the inflammatory process. It is well known that neutrophils are a source of cytokines and other inflammatory mediators when appropriately stimulated (32). It is also conceivable that in cases of overwhelming malignant involvement of the pleura, the pleural inflammatory process is diminished as a result of the paucity of normal mesothelium, contributing to failed chemical or mechanical pleurodesis. Eliciting a common mechanism of action may lead to simplification of the current processes used for inducing pleurodesis, by focusing on the host response.

Further research is required to determine whether the neutrophil itself is essential for the fibrotic process to occur, or whether IL-8 and MCP-1 play an independent role. Previous studies have reported that neutropenia does not suppress pleural fibrosis in the tetracycline model of pleural sclerosis (33). This suggests that the presence of neutrophils may be an unrelated response to eventual pleural sclerosis. It is apparent that further knowledge of potentiators and inhibitors of the pleural inflammatory response is necessary to understand and potentially modulate pleural inflammation, repair and fibrosis. Recent evidence suggests that IL-8 has a broader range of activity beyond neutrophil chemotaxis and activation, including stimulation of angiogenesis, as well as the ability to induce expression of ICAM-1 on endothelial cells and CD11/CD18 on leukocytes. Recent evidence suggests that these adherence molecules may function as growth factors for fibroblasts (34). MCP-1 was detected in inflammatory pleural effusions (27), and PMC were shown to produce MCP-1 in response to inflammatory stimuli (12). Recruitment of monocytes/macrophages to inflammatory sites in the pleura requires the expression of adhesion protein by both monocytes and pleural mesothelial cells, and the generation of chemotactic gradients. MCP-1 can regulate the cell-surface expression of adhesion molecules (35), and the capacity of MCP-1 to stimulate both monocyte chemotaxis and enhanced expression of integrins could therefore be important in inflammatory conditions. These observations suggest that the chemokines IL-8 and MCP-1 partly regulate leukocyte chemotaxis, together with the expression of ICAM-1, and may play a larger role in the process of pleurodesis.

These findings suggest a new perspective for understanding the mechanism of talc-induced pleurodesis, wherein the acute host inflammatory response to talc is mediated by the local production of proinflammatory cytokines by talc-stimulated mesothelial cells. The findings further emphasize the role of mesothelial cells in pleural inflammation. The findings in our study also add to the limited knowledge of the mechanisms of chemical pleurodesis.