INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Recent reviews (1) have concluded that exposure to high levels of mineral dust is associated with the development of chronic airflow obstruction. The anatomic basis of this physiologic effect remains unclear, but, by analogy with cigarette smoke, there are at least two candidate lesions: dust-induced fibrosis of the small airways, and dust-induced emphysema (4).

Mineral dust-induced fibrosis of the small airways appears pathologically as thick-walled fibrotic airways that are often markedly distorted and that usually contain readily visible dust-laden macrophages and free dust particles (see Reference 4 for illustrations). Lesions of this type have been found after exposure to a wide variety of dusts including asbestos, coal, silica, silicates, iron oxide, and aluminum oxide, and similar changes can be reproduced in animals (4).

The mechanism behind the development of dust-induced airway fibrosis is uncertain. Several possibilities exist. Dusts might produce airway wall fibrosis by incorporation of granulation tissue from inflammatory processes occurring in the lumen. Although this genesis of interstitial fibrosis is well documented in the alveolar region in conditions such as idiopathic interstitial fibrosis or organizing diffuse alveolar damage (5), there is no evidence that dusts provoke a similar response in the airways. In vivo, dusts do evoke an inflammatory reaction in the airway walls (6) and another possibility is that the fibrotic reaction is a reaction to mediators secreted by these exogenous inflammatory cells. However, dusts themselves have been shown to elicit production of fibrogenic cytokines such as platelet-derived growth factor (PDGF) in vitro (7) as well as other fibroblast growth factors from airspace and interstitial macrophages (8, 9). Thus a third possibility is that airway wall fibrosis is a direct effect of dust particles penetrating into the tissues and causing production of cytokine mediators and matrix components by the airway epithelium and interstitial cells. In this study we test the last hypothesis using rat tracheal explants as a model system of the airway wall.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Dusts

The asbestos used was the International Union Against Cancer (UICC) amosite standard reference sample. The geometric mean fiber size (GSD), as determined by counting by electron microscopy in our laboratory, was 3.8 µm (2.7) in length and 0.26 µm (1.9) width. The iron oxide was a hematite sample obtained from Reichard Coulson, Inc. (Bethlehem, PA). This is a compact particle with a geometric mean diameter of 0.45 µm (2.3). The titanium dioxide was anatase, geometric mean diameter 0.12 µm (1.4), obtained from Aldrich Chemicals (Milwaukee, WI).

Explant Preparation and Culture

Tracheal explants were prepared from 250-g male Sprague-Dawley rats using a modification of the method of Mossman (10) as previously described (11). Each explant was approximately 2 × 2 mm. Because most of the explant by weight is cartilage and the amount of tissue that actually contributes RNA is extremely small, 3 explants were used to prepare RNA for each data point; this procedure, by initial testing, provided a reliable signal. Freshly prepared explants from several different animals were used for each experiment and mixed to ensure that all explants for a given data point did not come from the same animal. Each test group shown in Figures 1 and 2 consisted of three data points.

View larger version (36K): [in this window] [in a new window]   Figure 1.   Representative ethidium bromide-stained gels of PCR products showing expression of PDGF-A chain in asbestos-exposed explants, and TGF- in iron-exposed and titanium dioxide-exposed explants at 1, 3, and 7 d of culture compared with control nondust-exposed explants. An increase in PDGF-A and TGF- is seen in the dust-exposed 7-d cultures.

View larger version (31K): [in this window] [in a new window]   View larger version (28K): [in this window] [in a new window]   View larger version (24K): [in this window] [in a new window]   View larger version (24K): [in this window] [in a new window]   View larger version (21K): [in this window] [in a new window]   Figure 2.   Graphical results of densitometry of ethidium bromide- stained gels showing expression of various genes of interest in dust-exposed and culture medium-exposed control explants after 7 d in organ culture. *p < 0.01 or less. (A) PDGF-A chain; (B) PDGF-B chain; (C ) TGF-; (D) TGF-; (E ) procollagen type I. Values are mean ± SD.

For dust exposure the explants were submerged, epithelial side up, in a 5 mg/ml (500 µg/cm²) suspension of dust in Dulbecco's modified Eagle medium (DMEM) without serum for 1 h. Controls were exposed only to culture medium. At the end of this time the explants were transferred to Petri dishes containing DMEM in agarose supplemented with 1% glutamine, 1% penicillin-streptomycin-fungizone, 1 µg/ml insulin, 0.1 µg/ml hydrocortisone, 1.5× amino acids, and 10% chicken serum. Explants were maintained in air plus 5% CO₂ organ culture with basal feeding in an incubator at 37° C for 1, 3, or 7 d. At the end of each time period explants were harvested for RNA extraction and for light microscopy.

Counting of Fibers or Particles in the Epithelium and Subepithelial Space

Explants were fixed in formalin, embedded in paraffin, and 5-µm-thick sections counterstained with nuclear fast red to allow easy visualization of particles. The number of asbestos fibers in the epithelium (defined as tissue above the basement membrane) and subepithelial (interstitial) space (defined as tissue between the basement membrane and the underlying cartilage) was determined by counting fibers with a light microscope; results were expressed as fibers/mm epithelial basement membrane length. For iron oxide and titanium dioxide, individual "particles" visible at ×450 magnification were counted (although these are actually aggregates of nonresolvable finer particles) and again expressed as particles/mm epithelial basement membrane length.

Expression of Growth Factors, Fibrogenic Mediators, and Matrix Components by Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR)

At 1, 3, and 7 d of organ culture, explants were harvested and RNA extracted. Because of the small amount of RNA available, even combining 3 explants to produce one data point, analysis of levels of expression of genes encoding various cytokine mediators and matrix proteins that have been suggested to play a part in dust-induced fibrosis was determined by RT-PCR. The asbestos and iron data were generated first, and because no effects were seen at 1 or 3 d, the titanium dioxide experiments were only carried out at 3 and 7 d.

Total RNA was extracted from cultured tracheal explants by the method of Chomczynski & Sacchi. First-strand complementary DNA (cDNA) was synthesized using superscript ribonuclease (Rnase) H reverse transcriptase (GIBCO-BRL, Grand Island, NY) according to the manufacturer's instructions. Briefly, 5 µg RNA were added to a reaction mixture of 1× first-strand buffer, 200 ng oligo(dT)12-18 primer, 0.5 mM each deoxyadenosine triphosphate (dATP), deoxythymidine triphosphate (dTTP), deoxyguanosine triphosphate (dGTP), and deoxycytidine triphosphate (dCTP), 0.1 M dithiothreitol (DTT), plus water to 49 µl. Two hundred units superscript RT was added and the reaction incubated at 42° C for 1 h.

PCR reactions contained 1-µM primers, 1.5 mM Mg⁺⁺, 200 µM deoxynucleotide triphosphates, reaction buffer, 2.5 U of Taq DNA polymerase (Perkin-Elmer Cetus Instruments, Norwalk, CT) and 1 or 5 µl of cDNA in a final volume of 20 µl. The PCR temperature profile consisted of 25 or 28 cycles of denaturation at 94° C for 45 s, primer annealing at 60° C for 45 s, and extension at 72° C for 1.25 min, followed by an additional 5 min final extension at 72° C. The PCR products were size fractionated on 1.5% agarose gel and quantitated from this ethidium bromide-stained gel using a gel Documentation System (Bio-Rad Laboratories, Hercules, CA).

Primer design was based on sequences from the Genbank database (Table 1). We optimized the reaction conditions (magnesium concentration, thermocycle temperature, etc.) to produce the greatest amount of a single PCR product. The amount of cDNA used and number of cycles of amplification were adjusted to stay within the linear region of amplification in order to allow quantification (Figure 3). Specificity of the various products was confirmed by restriction digests. Expression of malate dehydrogenase (12) was used as control (housekeeper) gene.

View larger version (8K): [in this window] [in a new window]   Figure 3.   Linearity of results using RT-PCR and densitometry on ethidium bromide-stained gels. Known amounts of RNA were used to prepare first-strand cDNA and the products amplified through 28 PCR cycles. Densitometry was performed using the Bio-Rad Gel Documentation System as described in METHODS. The amount of PCR product correlates in a linear fashion with input RNA template.

In Situ Hybridization

cDNA probes for transforming growth factor-₁ (TGF-₁) and PDGF-B were prepared as described previously by RT-PCR and subcloned into pCR II (Invitrogen Corp., Carlsbad, CA). A rat procollagen type I probe was the kind gift of Dr. E. Crouch, Washington University and has been described in Reference 6. Probes were linearized and labeled with ³⁵S and in situ hybridization carried out on paraffin sections of formalin-fixed explants according to our previously published protocols followed by autoradiography. Sense probes were also prepared and a similar protocol followed; all sense probes showed only minimal background grains.

Hydroxyproline Content of Explants

Hydroxyproline was measured by high-performance liquid chromatography (HPLC) as previously described (13). In brief, explants were dried and derivatized with phenylisothiocyanate using a Water Pico-Tag Station. Analysis was performed on a Waters HPLC system (Waters, Mississauga, ON, Canada) with a model 486 detector set at 254 nm, and a Whatman Partisil ODS-2 C18 (Whatman Inc., Clifton, NJ), 10 µm, 4.6 × 250 mm column. Results were expressed per milligram of tissue. A detailed protocol is available in Reference 13. Only Day 7 explants were analyzed.

Statistics

Comparisons were made between test and control groups at each time by analysis of variance using SYSTAT (14).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Particle Uptake

Epithelial uptake was visible to a minor degree by 24 h and steadily increased over time (Table 2); transport to the interstitium was much slower, so that no particles were seen in the subepithelial space at 24 h, small numbers were present by 3 d, and considerably greater numbers by 7 d (Table 2 and Figure 4). The overall pattern was similar for all three dusts, although the number of particles (actually aggregates) reaching the interstitial space was much greater for iron oxide and titanium dioxide compared with asbestos.

View larger version (141K): [in this window] [in a new window]   View larger version (119K): [in this window] [in a new window]   Figure 4.   Tracheal explant exposed to amosite asbestos (A) or iron oxide (B) and maintained in organ culture for 7 d. There is uptake of asbestos fibers and iron particles into the epithelium (E) and transport through the epithelium to the subepithelial interstitium (S). C = underlying cartilage plate, A = asbestos fibers on surface (moved off during tissue cutting); arrows show fibers in tissue. Iron appears as dark staining particles in air space, in epithelium, and in subepithelial interstitium. Explants exposed to titanium dioxide appear very similar to those exposed to iron. Both stained with nuclear fast red, original magnification: ×400.

RT-PCR Analysis of Fibrogenic Mediators and Matrix Components

Representative photographs of ethidium bromide-stained gels of the PCR products are shown in Figure 1 and graphical results from densitometry in Figure 2. No differences between test and control segments were seen at 1 and 3 d of culture for any dust. At 7 d in explants treated with amosite asbestos, there was about a 60% increase in PDGF-A chain, a 30% increase in PDGF-B chain, a 35% increase in procollagen type I, and a 25% increase in TGF-₁ expression compared with control. In explants treated with iron oxide, there was a 30% increase in PDGF-A and a 75% increase in TGF-₁ expression, but no increase in PDGF-B or procollagen. After titanium dioxide treatment there was a 40% increase in PDGF-A chain, a 35% increase in PDGF-B chain, a 100% increase in TGF-, and a 140% increase in TGF- expression, but no increase in procollagen. All the differences were, statistically, significantly different (p < 0.01 or less) from control values. Tumor necrosis factor- (TNF-) expression levels were statistically identical in dust-treated and control explants (data not shown). Fibronectin expression levels were not affected by titanium dioxide or iron oxide treatment, but were, surprisingly, very slightly decreased (about 10%) at 3 and 7 d after asbestos exposure (data not shown). No signal for tropoelastin was detected with any dust or in controls, indicating that there is no or only very low level expression of tropoelastin in these preparations.

In Situ Hybridization

After examining these data, in situ hybridization for TGF-₁, PDGF-B chain, and procollagen messenger RNA (mRNA) was carried on 7-d explants. Within the limits of resolution afforded by autoradiography, TGF-₁ expression appeared to be present in both the tracheal epithelial cells and the subepithelial space (Figure 5A), whereas procollagen expression (Figure 5C) was seen only in the subepithelial space. PDGF expression showed a similar distribution to that of procollagen but a less intense signal. There were no obvious differences in signal distribution among the dust-exposed groups and the control explants.

View larger version (142K): [in this window] [in a new window]   View larger version (137K): [in this window] [in a new window]   View larger version (134K): [in this window] [in a new window]   View larger version (149K): [in this window] [in a new window]   Figure 5.   In situ hybridization images of explants from 7 d showing expression of TGF- in both epithelium and interstitium (A, antisense probe; B, sense probe), and procollagen type I (C, antisense probe; D, sense probe). E = epithelium, S = interstitium, C = cartilage, A = overlying asbestos on epithelial surface (lifted off during the cutting process). The distribution of signal was the same in dust-treated and control explants, indicating that differences in expression levels arise within the compartments where these substances are normally expressed. All original magnifications: ×400.

Hydroxyproline Content

Hydroxyproline content is shown in Figure 6. Asbestos-exposed explants showed a 15% increase in hydroxyproline content (p = 0.05 by one-tailed test). Iron- and titanium-exposed explants did not differ from control.

View larger version (10K): [in this window] [in a new window]   Figure 6.   Hydroxyproline content of the explants at Day 7. Values are mean ± SD. *Significantly greater than control, p = 0.05 by one-tailed test.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study we have used rat tracheal explants to examine mechanisms of mineral dust-induced fibrosis of the airway wall. Explants offer a number of advantages for this purpose. Because they contain both mesenchymal and epithelial cells in their normal anatomic arrangement, they provide a realistic model in which interactions between mediators produced in the epithelium that affect underlying mesenchymal cells (for example TGF-) or vice versa can be examined. While they contain interstitial macrophages, they are free of airspace inflammatory cells, and, by definition, cannot mount an inflammatory response from the circulation, thus allowing one to examine intrinsic interactions between mineral dust particles and the airway wall.

Tracheal explants also have some distinct limitations. Dust uptake is much slower in the trachea than in the distal parenchyma, and uptake is considerably lower than is found with monolayer culture systems. Respiratory bronchioles, the primary site of dust deposition in vivo, and the region of the airways that shows the most pronounced dust-induced fibrosis, also appear to take up greater amounts of dust. The reasons for these differences are unclear, although slow uptake of dust in the explants does very closely mimic the time course of the same process in vivo in the trachea (15). The relatively low level of dust uptake is probably responsible for the relatively small increases in mediators and matrix components that we observe, but it should be remembered that, compared with a monolayer culture where all cells will produce a product, bulk analysis of tracheal explants by definition dilutes any given cytokine or matrix product with nonproducing cells. Thus the results we observe would probably be much greater in magnitude if we were only dealing with isolated cells of a particular type. However these experiments should not be viewed as indicative of absolute increases in mediator expression but rather as a model of events in the airway wall.

The present experiments strongly support the idea that the quite marked fibrosis seen in the small airways of workers with high level dust exposure (4) results, at least in part, from the penetration of particles into the airway epithelium and airway wall, with subsequent generation of proliferative and fibrogenic mediators. TGF-, for example, is a powerful mediator of matrix, particularly collagen and fibronectin, production (16) and PDGF stimulates fibroblast proliferation and migration (17); TGF- also stimulates both epithelial cell and fibroblast proliferation (18). However, it is equally clear from our experiments that different dusts evoke different patterns of cytokine and matrix production. The exact parameters (particle number, size, shape, and composition) that influence these differences in either a qualitative or quantitative fashion remain to be determined. But it is interesting that, in vivo, iron oxide and titanium dioxide are only weakly fibrogenic whereas amosite asbestos is quite strongly fibrogenic; thus our model system does mimic the in vivo reactivity of these dusts in terms of gene expression and hydroxyproline content.

Production of mediators such as PDGF-A and TGF- after exposure to silica or asbestos has been observed in tissue culture systems and in the alveolar interstitium in in vivo animal models, but it is usually assumed that dust-evoked airspace inflammatory cells are the initiators of the fibrogenic response (19). Our experiments suggest that in fact airway wall fibrosis may occur in the absence of airspace inflammatory cells and in the absence of circulating inflammatory cells. Interstitial macrophages are present in the explants and undoubtedly are important in mediator production (8, 9). It is also possible that secretion of these same cytokines from dust-stimulated exogenous inflammatory cells may act to increase the dust effects, but some experiments suggest that many cytokines produced in the airspaces cannot actually cross the epithelial barrier (8, 9). However, what our data imply is that dusts have the intrinsic property of inducing cell proliferation, cytokine secretion, and fibrosis in the airway wall.

Our results also support the growing body of evidence that epithelial cells are important sources of inflammatory and fibrogenic mediators (22). Most of these observations have been made in tissue cultures of alveolar epithelial cells (usually type II cells or A549 cells), but Coker and coworkers (25) recently demonstrated expression of TGF-₁ by in situ hybridization in the small airway epithelium of normal mice. The data reported here show that TGF- is also a product of the normal large airway epithelium. However, our finding that there is expression of PDGF-A and B chain in the subepithelial but not in the epithelial cells, with or without dust exposure, is contrary to that of Liu and coworkers (26) who reported PDGF expression in the epithelium of the bronchoalveolar duct junctions and small airways in asbestos-exposed animals but not (with rare exceptions) in control animals; similarly they found PDGF signal in the interstitium in asbestos-exposed but not control animals. These differences may reflect differences in fiber dose, and the use of in vitro versus in vivo preparations, as well as the time course of events after dust exposure.

It is particularly interesting that all of the dusts studied evoked upregulation of TGF- and PDGF-A expression. On the face of it these findings begin to provide a mechanistic basis for the general observation that any dust is fibrogenic if administered in sufficient quantity (27). However, because only asbestos caused increased procollagen expression and increased levels of hydroxyproline in these experiments, it is clear that the pathway leading to fibrosis in our system is more complicated than simple increases in TGF- or PDGF expression; whether quantitative differences in the levels of cytokines secreted, interactions of the cytokines, or expression of a substance we have not measured (for example, receptors such as PDGF-R) is the crucial element leading to fibrosis needs to be established.

The fact that increased expression of genes for these various mediators and matrix does not occur until the point that substantial numbers of particles have entered the tissues (Table 2) suggests that it is most likely particle penetration rather than just surface contact that is crucial to the induction of these components; whether particles must penetrate to the interstitium or only into the epithelium for these effects to occur remains to be determined. Lasky and coworkers (7) showed that chrysotile asbestos could directly upregulate PDGF expression by contact with cultured fibroblasts. Adamson and coworkers (8, 9) demonstrated that both asbestos and silica increased the production of fibroblast growth factors in cultured interstitial macrophages as well as in alveolar macrophages, but that penetration of fibers or particles to the interstitium was required for the initiation of fibrogenesis in vivo; they suggested that the direct proximity of dust-stimulated interstitial macrophages to fibroblasts was the crucial event in the induction of fibrosis. Their observations are somewhat different from ours in that particles apparently reached the interstitium after inducing epithelial necrosis and ulceration, whereas epithelial necrosis and ulceration are not found in our explants. Nonetheless, their findings would support the idea that the amount of dust reaching the interstitium is important in determining the fibrotic response. However, the process is probably more complicated, because in the experiments reported here, iron oxide and titanium dioxide particles reach the interstitium in considerably greater numbers compared with amosite asbestos fibers, but only amosite causes increased collagen production. Further, the upregulation of TGF- (and probably other mediators) in the airway epithelium after dust exposure may be geographically quite widespread, and thus it is possible that some of the effects of dusts on fibrogenesis are mediated by initial interactions of dust with epithelial cells and subsequent interactions of epithelium-derived mediators with underlying interstitial cells rather than only by direct dust interactions with interstitial cells. These hypotheses need further investigation.

Extrapolation of our findings suggests that all mineral dusts may evoke the production of mediators which potentially cause fibroblast proliferation and increased matrix production, and thus all dusts potentially can produce airway wall fibrosis, even in the absence of inflammatory cells. Although the airway wall is structurally different from the alveolar wall, the fundamental cellular elements are similar, and the same conclusions may apply to the alveolar parenchyma as well.