Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Urban air particulate matter  10 µm in aerodynamic size (PM₁₀) has received increasing attention because of its role in adverse respiratory health effects, yet little is known regarding the biologic effects of PM₁₀ on pulmonary cell types (1). PM₁₀ are complex particles derived from both anthropogenic sources (e.g., fossil fuel combustion) and natural sources (e.g., pollen, microbial contaminants) (4). Recent epidemiologic studies have correlated episodes of elevated PM₁₀ levels with increased mortality and morbidity (1, 9, 10), and with increased hospital emergency-room visits for asthma and acute respiratory symptoms in adults and children (2, 3, 11).

The acute inflammatory effects of PM₁₀ on the lung are likely mediated in part by resident populations of pulmonary macrophages. Upon inhalation, environmental particles are phagocytized by alveolar macrophages, resulting in the release of cytokines or the production of reactive oxygen intermediates (14, 15). Macrophage-derived mediators may act in a paracrine fashion to trigger inflammatory cascades in other pulmonary cell types such as epithelial cells and mesenchymal cells residing in close proximity to the airways. We postulated that PM₁₀ particles could influence the growth response of mesenchymal cells residing beneath the airway epithelium via a macrophage-dependent mechanism. This is a potentially important issue, because hypertrophy and hyperplasia of mesenchymal cells in close proximity to airways, as well as subepithelial fibrosis, are morphologic features of asthma and chronic airway inflammation (16, 17). No studies to date have determined whether PM₁₀ particles could alter the expression of growth factor receptors on mesenchymal cells and thereby influence the mitogenic response of these cells to polypeptide growth factors.

Platelet-derived growth factor (PDGF) is a major mitogen and chemoattractant for cells of mesenchymal origin that has been implicated in the pathobiology of a variety of pulmonary fibroproliferative diseases, including pulmonary fibrosis (18), asthma (19), and lung cancer (20). PDGF-A and -B polypeptide chains dimerize via sulfhydral linkages to form functional PDGF-AA, -AB, and -BB isoforms that bind cell-surface receptors termed PDGF -receptor (PDGF-R) and PDGF -receptor (PDGF-R), which also dimerize to form three possible receptor combinations (, , and ) (21). The expression of both PDGF-R and PDGF-R is necessary for the maximal mitogenic and chemotactic responses to PDGF, yet the expression of PDGF-R is normally very low or not detectable on cultured rat lung myofibroblasts (RLMF) (22, 23). PDGF-R is inducible by cytokines such as interleukin-1 (IL-1) (22) and basic fibroblast growth factor (24), or by bacterial lipopolysaccharide (LPS) (25). We recently reported that an emission-source residual oil fly-ash (ROFA) particle upregulates the PDGF-R through a macrophage-dependent pathway involving IL-1 (26). ROFA particles are LPS-free and it has been suggested that soluble transition metals associated with these particles mediate the inflammatory effects of ROFA in vivo (27).

In this study, we investigated the possible effects of PM₁₀ particles on the PDGF receptor system in vitro. PM₁₀ samples were obtained from three different regions of Mexico City, a city with high particulate air pollution (7), and we investigated the mechanisms whereby these particles affect PDGF-R expression on lung myofibroblasts. Mexico City PM₁₀ particles were found to upregulate PDGF-R expression through at least two possible mechanisms: (1) PM₁₀-derived LPS and metals (specifically vanadium pentoxide [V₂O₅]) stimulated alveolar macrophages to release IL-1-like activity which then induced PDGF-R on myofibroblasts in a paracrine manner, and (2) PM₁₀ directly upregulated PDGF-R on myofibroblasts because of the presence of particle-associated LPS.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

Recombinant endotoxin neutralizing protein (rENP) was kindly provided by Dr. Paul Ketchum (Associates of Cape Cod, Inc., Falmouth, MA). The ENP is normally sold coated on silica beads (catalog no. MQC10), but we used free ENP. The ENP is an 11.8-kD protein purified from the amebocytes of the horseshoe crab, Limulus polyphemus. It is a basic, amphipathic protein, with an isoelectric point of approximately 10. ENP neutralizes the bioactivity of LPS as measured by the Limulus amebocyte lysate assay when used in a 1:1 ratio (weight) of ENP/LPS. Escherichia coli LPS was purchased from Sigma (Serotype 026:B6, catalog no. L-8274; St. Louis, MO). IL-1 and the IL-1 receptor antagonist protein (IRAP) were purchased from R&D Systems (Minneapolis, MN). PDGF isoforms and antibodies to PDGF-R and -R were purchased from Upstate Biotechnology (Lake Placid, NY). [¹²⁵I]PDGF-AA was purchased from Biomedical Technologies, Inc. (Stoughton, MA). Antibodies for myofibroblast characterization were alpha-smooth muscle actin antibody (Sigma), Vimentin antibody (Biogenex, San Ramon, CA), Factor VIII and OX-1 antibodies (DakoPatts, Carpinteria, CA).

PM₁₀ Sampling

PM₁₀ concentration levels were routinely monitored for 4 consecutive days per week over the course of 1 yr in 1993 at three different sampling sites in Mexico City: north (industrial zone), center (business zone), and south (residential area). PM₁₀ were sampled using a Sierra Andersen (Monterey, CA)/GMW Model 1200 VFC HVPM10 sampler. The sampler flow rate was kept constant at 1.13 m³/ min¹ using G313 flow control modules. For sample collection, glass fiber filters were used. Filters were kept cold and dry until used. After each 4-d collection period, PM₁₀ were gently brushed off the filters in a laminar flux hood after dry sonication for 60 min. Samples from each zone were pooled and care was taken to recover, store, and weigh the samples in glassware that had been baked at 200°C for 4 h. All samples were sterilized dry after collection. Just before each experiment, the PM₁₀ particles were suspended in appropriate medium and briefly sonicated to ensure an even suspension.

Endotoxin

Endotoxin was measured by a Limulus amebocyte lysate assay kit according to manufacturer's specifications (BioWhittaker, Inc., Walkersville, MD), after sonication of particles in endotoxin buffer (0.05 M potassium phosphate and 0.01% triethylamine, pH 7.5) for 60 min at 20°C (Bath Sonicator 5200; Branson Ultrasonics, Danbury, CT). Measurements in filter blanks gave 0.23 ± 0.10 ng/ml endotoxin levels. Endotoxin determinations in PM₁₀ samples analyzed within 24 h of collection gave results similar to those obtained in stored samples.

Elemental Analysis of PM₁₀ Samples

Elemental analysis of water extracts and 1 M HCl hydrolysates of PM₁₀ samples was determined by ICP-AES (Plasma 40; Perkin-Elmer, Norwalk, CT). The instrument was calibrated using a five-point curve prepared from solutions of pure element standards (Aldrich Chemical Co., Milwaukee, WI). Performance check solutions were analyzed before and after the unknowns. Measurement accuracies were within 4% for the transition metals, 10% for lead, and 7% for sulfate.

Isolation and Characterization of RLMF

Early passage RLMF were isolated as described previously from male Sprague-Dawley rats (22). RLMF isolates at passage 1 or 2 were plated onto 3-aminopropyltriethoxysilane-coated glass chamber slides and grown to confluence, then fixed briefly in ice-cold acetone. Fixed cells were then subjected to overnight incubation with a murine monoclonal antibody to the antigen of interest, followed by a biotinylated horse antimouse antibody (Vector Labs, Burlingame, CA), avidin-biotin immunoperoxidase (Vector), and 3,3'-diaminobenzidine chromogen (Vector). An irrelevant monoclonal antibody (anti-5-bromo-2'-deoxyuridine; Becton Dickinson, San Jose, CA) at equivalent IgG concentration served as control for nonspecific immunoreactivity. Concentrations of primary antibodies were set by titration on appropriate positive control cells: rat aortic smooth muscle cells (American Type Culture Collection [ATCC, Rockville, MD] CRL 1476) were used for actin, rat dermal fibroblasts (ATCC CRL 1213) for vimentin, bovine pulmonary artery endothelial cells (ATCC CCL 209) for factor VIII, and freshly isolated rat alveolar macrophages for OX-1. RLMF stained positively for vimentin and alpha-smooth muscle actin and negatively for factor VIII and rat leukocyte common antigen (OX-1). In addition, examination of glutaraldehyde-fixed pellets of RLMF by transmission electron microscopy showed ultrastructural features consistent with a myofibroblast phenotype (abundant intermediate filaments and rough endoplasmic reticulum, and lack of Weibel-Palade bodies characteristic of endothelial cells). Myofibroblasts were grown to confluence in 10% fetal bovine serum/Dulbecco's modified Eagle's medium (FBS/DMEM) before being seeded for the assays described below.

Isolation of Alveolar Macrophages

Macrophages from male Sprague-Dawley rats were obtained by bronchoalveolar lavage as previously described (26). Macrophages were cultured in 75-cm² flasks that had been coated with poly(2-hydroxyethyl methacrylate) (Sigma), which prevents macrophage attachment to the culture surface but does not interfere with particle phagocytosis. Cells were incubated for 24 h at 37°C in 5% CO₂ in the absence or presence of 10 µg/cm² of Mexico City PM₁₀ or Mt. St. Helen's volcanic ash (MSHA) particles. The culture medium was centrifuged at 1,500 rpm to pellet macrophages and the supernatant filtered (0.45 µm) and stored at 80°C. A preliminary dose-response experiment, wherein macrophages were exposed to 0, 1, 5, 10, 50, or 100 µg/cm² of each particle, demonstrated that particle densities below 50 µg/cm² did not appreciably affect cell viability as determined by trypan blue exclusion. In some experiments, macrophages were cultured in the presence of metals (V₂O₅, VSO₄, NiO, NiSO₄, Fe[SO₄]₃, CuSO₄, ZnSO₄) or E. coli LPS (Sigma), because these components were constituents of Mexico City PM₁₀ samples (see Tables 12).

[¹²⁵I]PDGF-AA Receptor Assays

Binding of [¹²⁵I]PDGF-AA (Biomedical Technologies) was assayed on confluent, quiescent cell cultures. RLMF in 24-well plates were grown to confluence in 10% FBS/DMEM and then rendered quiescent in serum-free defined medium (SFDM) for 24 h. Cells were then incubated with particles, IL-1, or LPS for 24 h as described in the figure legends. Our previous studies have shown that maximal expression of PDGF-R occurs 24 h following stimulation with IL-1 or LPS (22, 25). The following day, cultures were chilled to 4°C, rinsed in cold binding buffer (Ham's F-12 with N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid, CaCl₂, and 0.25% BSA), and exposed to 2 ng/ml of [¹²⁵I]PDGF-AA in the absence or presence of nonradioactive 500 ng/ml PDGF-AA to measure total and nonspecific binding, respectively. Binding was allowed to occur for 3 to 4 h at 4°C on an oscillating platform. Cells were then rinsed 3 times in ice-cold binding buffer and solubilized in 1% Triton X, 0.1% BSA, and 0.1 N NaOH; and cell-associated radioactivity was counted in a -counter. For each experiment, total and nonspecific binding were performed in triplicate and all data shown are specific binding after correction for cell number.

Western Blotting

RLMF were grown to confluence in 75-cm² flasks and rendered quiescent for 24 h in SFDM. Cultures were exposed to 0.25× macrophage-conditioned medium (MCM) or 1 ng/ml IL-1 for 24 h. Cells were washed with phosphate-buffered saline (PBS) and 250 µl of lysis buffer (50 mM Tris-HCl; 1% Triton X-100; 150 mM NaCl; 1 mM ethyleneglycol-bis-(-aminoethyl ether)-N-N'-tetraacetic acid; 1 mM phenylmethylsulfonyl fluoride, 0.25% Na-deoxycholate; 1 µg/ml each of aprotinin, leupeptin, pepstatin; 1 mM Na₃VO₄, 1 mM NaF) was added to cover the surface of the attached cells for 20 min. Extracts were stored at 70°C. A total of 20 µl of each sample mixed with sample buffer was boiled for 5 min before electrophoresis in a 2- 15% Tris-glycine sodium dodecyl sulfate polyacrylamide gel (Integrated Separation Systems, Hyde Park, MA) for 2 h at 130 V and 30 mA. The protein on the gel was transferred to a nitrocellulose membrane (Hybond; Amersham, Arlington Heights, IL). The membrane was blocked with 3% milk/PBS for 1 h before addition of a rabbit antimouse PDGF alpha or beta receptor antibody (Upstate Biotechnology) overnight. After washing 3 times with PBS-Tween, a secondary horseradish peroxidase-conjugated swine antirabbit antibody (Dako, Carpinteria, CA) was added for 90 min. An ECL luminol kit (Amersham) was used for detection of bound secondary antibody.

IL-1 Assay

Macrophage supernatants were analyzed for rat IL-1 with a commercially available enzyme-linked immunosorbent assay (ELISA) kit according to the manufacturer's instructions (Endogen, Woburn, MA).

Statistical Analysis

The Systat statistical package was used for all analyses (Systat, Evanston, IL). Two sample t tests were performed to compare the control group with a treatment group, or to compare a treatment group with the same treatment in the presence of an inhibitor (e.g., IRAP or rENP).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Induction of RLMF PDGF-R by Conditioned Medium from Macrophages Stimulated with Mexico City PM₁₀ Samples

PM₁₀ samples from the southern, central, and northern regions of Mexico City were compared with MSHA for their ability to induce PDGF receptors as measured by Western blotting assays and [¹²⁵I]PDGF-AA binding assays. Mexico City PM₁₀ stimulated cultured alveolar macrophages to release a factor(s) that upregulated the PDGF-R on RLMF as shown by Western blot analysis (Figure 1A). Expression of the PDGF-R was not affected. IL-1 was used as a positive control to induce expression of the PDGF-R (22). MCM from PM₁₀-treated macrophages also upregulated the specific binding of [¹²⁵I]PDGF-AA to RLMF (Figure 1B). MSHA particles did not induce PDGF-R when compared with MCM alone (Figures 1A and 1B). All Mexico City PM₁₀ samples contained a variety of transition metals that were detected by ion-coupled plasma emission spectrometry, including Cu, Fe, Ni, V, Zn, and Pb (Table 1). The PM₁₀ samples also contained endotoxin that was detected by the Limulus amebocyte lysate assay, whereas the MSHA particle sample did not contain any detectable endotoxin (Table 2). MSHA particles have been previously characterized as 1.8 µm in mean diameter (0.7 to 6.0 µm range) and composed of silica and aluminum with minor amounts of Na, K, and Ca (15). Our analysis also detected Fe in MSHA (Table 1).

View larger version (34K): [in this window] [in a new window]   Figure 1.   Upregulation of PDGF-R expression on RLMF following stimulation with supernatants from rat alveolar macrophages stimulated with PM10 particles. Macrophages were stimulated for 24 h with MSHA or PM10 (10 µg/cm2) from the southern, central, and northern regions of Mexico City (MEX-S, MEX-C, MEX-N, respectively). The MCM (0.25×) was added to confluent, quiescent cultures of RLMF for 24 h in serum-free defined medium prior to collecting cell lysates for Western blot analysis of PDGF receptors (A) or performing a [125I]PDGF-AA binding assay (B). IL-1 (2 ng/ml) was tested on parallel cultures of RLMF, a positive control for PDGF-R induction. All Mexico City PM10 samples stimulated macrophages to release upregulatory activity for the PDGF-R on RLMF; P < 0.01 compared with control medium, +MCM treatment, or MCM (MSHA) treatment. Western blotting data are from one experiment typical of four separate experiments. Receptor binding data are the means ± SEM of three experiments.

The Upregulatory Activity Released by PM₁₀-Treated Macrophages for the PDGF-R on Myofibroblasts Is Due in Part to IL-1

We postulated that IL-1 secreted by PM₁₀-stimulated macrophages was driving upregulation of the PDGF-R. Therefore, RLMF were incubated with MCM in the absence or presence of IRAP to block the binding of IL-1 to its receptor. IRAP inhibited the induction of PDGF-R by recombinant IL-1 and MCM from Mexico City north PM₁₀-stimulated macrophages by 100% and ~ 70%, respectively (Figure 2). All three Mexico City PM₁₀ samples stimulated alveolar macrophages to release IL-1 as determined by ELISA, whereas MSHA particles did not cause an increase in IL-1 release when compared with unstimulated macrophages (Table 3). The concentration of IL-1 present in the MCM in Table 3 (i.e., ~ 100 pg/ ml) has been demonstrated by our laboratory to upregulate PDGF-R (22). Together these data demonstrate that PM₁₀ particles stimulated the release of IL-1 by macrophages and that the majority of RLMF PDGF-R upregulatory activity in the MCM was due to IL-1.

View larger version (21K): [in this window] [in a new window]   Figure 2.   Upregulation of [125I]PDGF-AA binding to RLMF by MCM (0.25×) from PM10-stimulated macrophages is inhibited by the IRAP protein. RLMF were treated for 24 h with MCM derived from macrophages incubated in serum-free defined medium alone (control) or macrophages stimulated with Mexico City north PM10 (10 µg/cm2). Parallel cultures of RLMF were treated with MCM in the presence of IRAP (2 µg/ml). IRAP significantly inhibited upregulation of PDGF-R induced by IL-1 or MCM (MEX-N) (P < 0.01) and MCM (P < 0.05). Data are the means ± SEM of three experiments wherein the total and nonspecific binding were each assayed in triplicate.

Endotoxin Associated with Mexico City PM₁₀ Stimulates Alveolar Macrophages to Secrete Upregulatory Activity for the Myofibroblast PDGF-R

We sought to determine the component(s) of PM₁₀ that mediated the release of IL-1-like activity by macrophages. Because endotoxin (LPS) has been reported to induce the production of IL-1 by macrophages (28, 29), we stimulated macrophages with Mexico City PM₁₀ in the absence or presence of rENP and then measured the release of macrophage-derived upregulatory activity for the PDGF-R on RLMF (Figure 3). The rENP blocked the release of macrophage-derived upregulatory activity for the myofibroblast PDGF receptor by ~ 50%, indicating that the PM₁₀ were exerting their stimulatory effect on macrophages, at least in part, through particle-associated LPS.

View larger version (32K): [in this window] [in a new window]   Figure 3.   Effect of recombinant endotoxin-neutralizing protein (rENP) on alveolar macrophage release of upregulatory activity for the PDGF-R on RLMF. Macrophages were stimulated for 24 h in the absence or presence of Mexico City PM10 particles (10 µg/cm2). Parallel cultures of macrophages were stimulated with PM10 in the presence of rENP (1 µg/ml). The MCM was then harvested and added to RLMF for 24 h prior to performing a [125I]PDGF-AA binding assay. rENP significantly inhibited the PM10-stimulated release of macrophage-derived upregulatory activity for the RLMF PDGF-R (**P < 0.01, *P < 0.05). Data are the means ± SEM of three experiments wherein the total and nonspecific binding were each assayed in triplicate.

Mexico City PM₁₀ Stimulate Lung Myofibroblasts to Upregulate PDGF-R through a Macrophage- Independent Pathway Involving Endotoxin

We also addressed the possibility that PM₁₀ could directly stimulate the expression of the PDGF-R on RLMF without involvement of the macrophage. We previously reported that LPS at a concentration as low as 10 ng/ml is a potent inducer of PDGF-R on RLMF (25). PM₁₀ from all three regions of Mexico City induced PDGF-R upregulation on RLMF directly and this induction was ~ 50% blocked by rENP (Figure 4). MSHA particles had no discernible effect on the PDGF receptor system on RLMF in this direct-exposure strategy.

View larger version (18K): [in this window] [in a new window]   Figure 4.   The rENP blocks upregulation of the PDGF-R on RLMF by Mexico City PM10 particles. RLMF were stimulated with PM10 (10 µg/cm2) for 24 h in the absence or presence of rENP (1 µg/ml) prior to performing a [125I]PDGF-AA binding assay. E. coli lipopolysaccharide (LPS, 10 µg/ml) was used as a positive control for PDGF-R upregulation. rENP significantly reduced upregulation of [125I]PDGF-AA binding in response to LPS (**P < 0.01) or Mexico City PM10 (*P < 0.05). Data are the means ± SEM of three experiments wherein the total and nonspecific binding were each assayed in triplicate.

Vanadium Pentoxide, a Component of Mexico City PM₁₀, Induces PDGF-R on RLMF through a Macrophage-Dependent Pathway Involving IL-1

Recent studies have shown that metals (Fe, V, Ni) associated with emission source particles could also serve as potential stimulators of inflammatory mediators (e.g., IL-1) by pulmonary cells (30). Therefore, we examined whether Mexico City PM₁₀-associated metals (Table 1) could participate in a similar mechanism in vitro. A variety of metals shown to be present in the PM₁₀ samples were tested for their potential to stimulate macrophages to release upregulatory activity for the PDGF-R on RLMF (Table 4). Of these, only V₂O₅ stimulated the secretion of macrophage-derived factor(s) that upregulated PDGF-R on RLMF (Table 4, Figure 5). Therefore, we further investigated V₂O₅ as a candidate metal for mediating upregulation of PDGF-R on RLMF through either a macrophage-dependent or -independent pathway. Direct treatment of RLMF with V₂O₅ (0.01 to 1 µg/cm²) caused only a minor stimulatory effect on induction of the PDGF-R as determined by the [¹²⁵I]PDGF-AA binding assay (Figure 5A). However, alveolar macrophages treated with V₂O₅ secreted a factor(s) that maximally upregulated the PDGF-R when the MCM was added to RLMF (Figure 5B). Relatively high concentrations of V₂O₅ (1 µg/cm²) were required to induce PDGF-R. The majority of this PDGF-R upregulatory activity was due to IL-1, because IRAP inhibited the increase in [¹²⁵I]PDGF-AA binding to RLMF (> 75%) mediated by MCM. Also, MCM from macrophages that were stimulated with 1 µg/cm² V₂O₅ contained 160 pg/ ml IL-1 as determined by ELISA, compared with 30 pg/ ml in MCM from unstimulated cells.

View larger version (21K): [in this window] [in a new window]   Figure 5.   Effect of V2O5 on PDGF-R expression by direct stimulation of RLMF or treatment of RLMF with MCM from V2O5-stimulated macrophages. (A) Direct exposure of RLMF to V2O5 induced only a minor increase in [125I]PDGF-AA binding, but V2O5 strongly induced macrophages to release upregulatory activity for the PDGF-R on RLMF (*P < 0.05; **P < 0.01). (B) Upregulation of PDGF-R on RLMF by MCM from V2O5-treated macrophages was significantly inhibited (**P < 0.01) by the IRAP. Data are the means ± SEM of three experiments wherein the total and nonspecific binding were each assayed in triplicate.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We have previously reported that induction of the PDGF-R by IL-1 and LPS renders RLMF hyperresponsive to the mitogenic and chemotactic effects of PDGF (22, 23). Recently, we found that ROFA particles upregulated PDGF-R on RLMF through a macrophage-dependent pathway involving IL-1, but direct treatment of the RLMF with ROFA did not cause induction of PDGF-R (26). In the present study, we found that PM₁₀ particles collected from three regions of Mexico City, which represent a complex mixture of organic and inorganic constituents derived from natural and anthropogenic sources (6, 7), are stimulators of the PDGF-R through at least two different possible mechanisms: (1) endotoxin or vanadium stimulates macrophages to release IL-1, which then functions as a paracrine signal for PDGF-R upregulation on RLMF; and (2) PM₁₀-associated endotoxin directly stimulates RLMF. These macrophage-dependent and -independent mechanisms are illustrated in Figure 6.

View larger version (13K): [in this window] [in a new window]   Figure 6.   Hypothetical scheme showing induction of the PDGF receptor system on RLMF by PM10 particles through macrophage-dependent and -independent pathways. Two components of Mexico City PM10 that mediate induction of the PDGF receptor system on RLMF are (1) V2O5, and (2) endotoxin (LPS). Both V2O5 and LPS are strong inducers of IL-1 release by alveolar macrophages; the LPS component is blocked by rENP. The contribution of vanadium to PM10-induced upregulation of PDGF-R remains questionable because of the relatively high concentrations of this metal required to elicit PDGF receptor induction. IL-1 secreted by activated macrophages serves as a paracrine inducer of the PDGF-R on lung myofibroblasts and is inhibited by the IRAP protein. LPS present on PM10 particles may also directly stimulate upregulation of PDGF-R on myofibroblasts without involvement of the macrophage.

Endotoxin has been recognized previously as a mediator of a number of occupational diseases involving particle/fiber inhalation such as cotton dust and grain dust (31- 33). In these pulmonary diseases it is generally accepted that endotoxin mediates its effects by stimulating the release of inflammatory cytokines (e.g., IL-1 and tumor necrosis factor [TNF]-) from macrophages and other pulmonary cell types. A recent study by Becker and coworkers showed that urban ambient air particles (UAP) from four urban centers (Washington, DC; St. Louis, MO; Ottawa, Canada; and Dusseldorf, Germany) stimulated rat and human alveolar macrophages to release TNF and IL-6 (15). These investigators found that cytokine secretion in response to UAP was inhibited by polymyxin B, suggesting endotoxin as a constituent responsible for the particle-mediated inflammatory response. This is consistent with our observation that ambient particles from Mexico City stimulate PDGF-R at least in part through endotoxin. In our experiments we blocked the effect of LPS with an ENP, which has been shown by other investigators to inhibit E. coli-induced sepsis in rats (34). Direct PM₁₀ treatment of RLMF caused a relatively weak induction of PDGF-R, perhaps because of the relatively low levels of LPS present in the particle suspension (i.e., 50 µg/ml PM₁₀ contained 0.5 to 1 ng/ml LPS) as compared with the concentration of pure LPS (10 µg/ml) used to stimulate upregulation of PDGF-R maximally (Figure 4). However, the low levels of LPS in the PM₁₀ were effective in maximally stimulating macrophages to release upregulatory activity (i.e., IL-1) for the PDGF-R on RLMF (Figure 3), and this could be due to a lower dose-response for LPS-induced release of IL-1 by macrophages compared with direct induction of PDGF-R on myofibroblasts.

It is possible that factors other than endotoxin, such as certain transition metals, can mediate the inflammatory effects associated with Mexico City PM₁₀ particles. We recently reported that ROFA particles stimulate the upregulation of the PDGF-R on lung myofibroblasts through a macrophage-dependent mechanism involving IL-1 (26), yet ROFA particles neither contain detectable levels of endotoxin nor directly stimulate myofibroblasts to induce the PDGF-R. ROFA particles contain metals such as V, Fe, Ni, Mn, and Pb in addition to a variety of other inorganic and organic constituents (27). The transition metals V and Ni have been shown to stimulate IL-1 messenger RNA expression in the lung (30), and in this study we observed that V₂O₅ stimulated macrophages to secrete IL-1 and IL-1-like activity that upregulates the PDGF-R on RLMF (Figure 5). Other metals, including Fe(SO₄)₃, NiO, NiSO₄, CuSO₄, and ZnSO₄ did not stimulate macrophages to release PDGF-R upregulatory activity (Table 3). Surprisingly, VSO₄ also did not stimulate PDGF-R via macrophage-derived mediators (Table 3); the reason for this difference between VSO₄ and V₂O₅ remains unclear. It is noteworthy that MSHA, ambient particles from a volcanic source, contain no detectable endotoxin and are composed mainly of silica and aluminum with trace amounts of Na, K, Ca, and Fe (15; Table 1). We found that MSHA particles did not upregulate PDGF-R on myofibroblasts.

In our experiments, we used high concentrations of V₂O₅ relative to the amounts that were measured in the Mexico City PM₁₀ samples (Figure 5, Table 1). While it is clear that V₂O₅ is a strong inducer of PDGF-R on RLMF via a macrophage-dependent pathway involving IL-1 (Figure 5), and since vanadium is present in Mexico City PM₁₀, it is possible that vanadium in the +5 oxidation state could be responsible in part for mediating induction of the PDGF receptor system by Mexico City PM₁₀. However, given the low levels of vanadium detected in Mexico City PM₁₀ (Table 1) relative to the amount of V₂O₅ required to induce PDGF-R expression in vitro (Figure 5), our data do not strongly support V₂O₅ as a PM₁₀-derived mediator of PDGF-R upregulation. In other words, cells treated with 10 µg/cm² of PM₁₀ actually received ~ 3 ng/cm² vanadium, as compared with cells that received pure V₂O₅ at 1 µg/cm² (i.e., an approximate 300-fold difference). It is possible that vanadium could act synergistically with other factors in PM₁₀ (e.g., endotoxin), and thus the low levels of vanadium measured in the Mexico City PM₁₀ could be significant in mediating induction of the PDGF receptor system. Further studies should focus on possible synergistic effects of PM₁₀ components such as metals and endotoxin in mediating inflammatory responses.

Upregulation of PDGF-R by PM₁₀ particles could have a role in remodeling of the pulmonary interstitium or the airways of the lung, although it is currently unknown whether or not PM₁₀ exposure in vivo affects PDGF receptor expression. We recently reported that intratracheal instillation of emission-source ROFA particles or V₂O₅ caused upregulation of the PDGF-R in vivo and receptor induction preceded mesenchymal cell hyperplasia and collagen deposition (26, 35). Whereas ROFA and V₂O₅ cause acute lung injury and subsequent fibrosis, there is no known correlation between PM₁₀ and pulmonary fibrosis in humans or experimental animals. PM₁₀ particles have been associated with an increased incidence of asthma and acute respiratory symptoms in children and adults (2, 3, 11). Although induction of PDGF-R likely does not play a role in the acute effects of inhaled PM₁₀, the increased proliferation of myofibroblasts adjacent to airways could contribute to chronic fibroproliferative changes that contribute to airway narrowing. For example, morphologic features of airways from patients with asthma include hypertrophy and hyperplasia of airway smooth muscle cells and subepithelial fibrosis (16, 17). Interestingly, we recently observed that induction of the PDGF-R on human airway smooth muscle cells serves to increase the mitogenic responsiveness of these cells to PDGF, and this could be a possible mechanism contributing to mesenchymal cell hyperplasia (24). Given the results of the present study, it appears that components of PM₁₀ such as endotoxin and metals could enhance PDGF-stimulated proliferation of airway smooth muscle cells and myofibroblasts adjacent to airways via induction of the PDGF receptor system. Because endotoxin and metals are soluble components of PM₁₀, they could translocate to the interstitium and directly interact with mesenchymal cells. Although we did not assay the production of PDGF isoforms by pulmonary cells in the present study, it has been reported that alveolar macrophages and lung myofibroblasts secrete PDGF following stimulation with inorganic particles (36).

In summary, we report that PM₁₀ from three regions of Mexico City induce expression of the PDGF -receptor subtype on rat pulmonary myofibroblasts. PDGF receptor induction on myofibroblasts is mediated through macrophage-dependent pathway(s) involving IL-1 wherein endotoxin and metal components of PM₁₀ stimulate IL-1 release. The endotoxin present on PM₁₀ particles also elicited upregulation of the PDGF receptor by direct interaction with the myofibroblasts. These data suggest that PM₁₀ exposure leads to remodeling of airways or the pulmonary interstitium by enhancing myofibroblast replication and chemotaxis.