INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Platelet-derived growth factor (PDGF) is synthesized by a number of cell types, including megakaryocytes (as the source of platelets) (1), lung macrophages (2), fibroblasts (3), endothelial cells (4), and epithelial cells (5). Active PDGF is a 32-kD glycoprotein dimer of PDGF-A and -B chains linked by disulfide bonds, and includes the AA and BB homodimers as well as the AB heterodimer. These ligands, in dimeric form, bind to receptor dimers composed of alpha and beta subunits (6). The PDGF-BB isoform can bind to the PDGF-, -, or - receptors, whereas the PDGF-AA isoform only binds with high affinity to the receptor. Binding of the PDGF dimers to the receptor subunits results in transphosphorylation and subsequent activation of signal transduction pathways, which in turn leads to increased cell division (4). PDGF isoforms are potent mitogens (4) and chemoattractants (7) for mesenchymal cells in vitro. They are responsible for greater than 50% of the serum-induced mitogenic activity for smooth muscle cells. Anti-PDGF antibodies have been shown to block mesenchymal cell proliferation in animal models of vascular (8) and renal (9) fibroproliferative disease. In addition, many other fibroblasts growth factors exert their mitogenic activity, at least in part, through a PDGF-dependent pathway (10).

Inhalation of asbestos fibers is well recognized to result in lung fibrosis. Rats or mice exposed to a single aerosolized- asbestos insult for 1 to 5 h develop a fibroproliferative lesion at the predominant site of fiber deposition, i.e., the alveolar duct bifurcation (ADB) (13, 14). The time-related anatomic details of this lesion have been studied by several techniques, including morphometry, immunohistochemistry, and in situ hybridization. Interestingly, there is a significant increase in interstitial cell number and volume as well as an increase in the connective tissue matrix that the interstitial mesenchymal cells secrete and maintain (14). The peak proliferative index in mouse lungs, as assessed by thymidine incorporation, occurs 33 h after the onset of exposure (15). Alveolar macrophages are rapidly attracted to the sites of fiber deposition in animal models (13), and these cells have been shown to secrete all three PDGF isoforms when exposed to asbestos fibers in vitro (16). Some inhaled asbestos fibers are cleared via the mucocilliary escalator; yet many cross the epithelial monolayer and reside in the interstitium where they come into contact with resident fibroblasts and interstitial macrophages (17).

We have shown previously that lung fibroblasts proliferate under serum-free conditions when exposed to asbestos fibers, and that the asbestos-induced mitosis is specifically blocked with an anti-PDGF antibody (18). Northern analysis revealed that rat lung fibroblasts synthesize PDGF-A chain, but not PDGF-B chain, in culture (18). Furthermore, Western analysis confirmed that asbestos exposure results in the secretion of immunoreactive PDGF by lung fibroblasts in vitro (18). Radioreceptor and Northern analysis revealed that the PDGF-R, but not PDGF-R, is upregulated after asbestos exposure in vitro (19). Moreover, the increased PDGF-R expression resulted in a greater proliferative response of rat lung fibroblasts (RLFs) to recombinant PDGF-AA (19).

Using a well-characterized rat model of asbestos-induced lung fibrosis, we have recently used ribonuclease protection assay, in situ hybridization, and immunohistochemistry to demonstrate that both PDGF-A and -B chains are upregulated at the sites of the fibroproliferative lesions immediately after asbestos exposure and for at least 2 wk thereafter (20). Little is known regarding expression of PDGF receptors in the lung. In the present report, we describe the time course and localization of PDGF receptor expression in vivo during the development of asbestos-induced lung fibrosis. With the experiments presented here, we have shown that the PDGF-R, but not the PDGF-R, is upregulated during the initial development of fibrogenesis.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Asbestos and Iron Exposure

Eight-wk-old male Sprague-Dawley rats were exposed to 10 mg/m³ of chrysotile asbestos in inhalation chambers for a period of 5 h as described previously in detail in our earlier publications (15, 17). This asbestos exposure induces fibroblast proliferation and a morphometrically characterized lesion at the alveolar duct bifurcations (14, 15). A separate group of sham-exposed (ambient air) animals and a group of animals exposed to 50 mg/m³ of carbonyl iron (a nonfibrogenic particle) were used as controls. Animals were killed for Northern and Western analysis, as well as for histologic examination and immunohistochemical staining immediately after exposure, which we refer to as the 5-h time point. Other similarly exposed animals were killed 1, 2, or 8 d after exposure.

Lung Tissue Procurement

At designated time points after exposure, the rats were given an intraperitoneal injection of tribromoethanol (Aldrich Chemical Co., Milwaukee, WI) to render them insensitive to pain prior to being killed. After ligation of the renal artery the chest cavity was exposed. The left lung was tied off at the mainstem bronchus and removed. The left lung was then immediately wrapped in foil and placed in liquid nitrogen for at least 30 min before storage at 70° C for later Northern analysis. In other animals, the left lung was not frozen but immediately prepared for Western analysis. The right lung was perfused with 10% neutral formalin through a tracheal canulus at a pressure of 15 cm H₂O for 15 min. Then, the trachea was clamped and the right lung removed from the chest cavity and placed in fresh fixative overnight at 4° C before processing in paraffin.

Northern Analysis

RNA isolation. Left rat lungs were removed from the 70° C freezer and immediately dissociated in 2 ml of 4.2 M GTC using the Tissue Tearor (Biospec Products, Bartlesville, OK). The samples were transferred to Corex tubes and centrifuged at 10,000 × g to remove unwanted precipitate. The supernatant was gently layered over 1 ml of 5.7 M CsCl and the samples were centrifuged in a Beckman tabletop centrifuge (Beckman Instruments, Fullerton, CA) for 5 h at 21° C. The supernatant was carefully aspirated and discarded. The remaining pellet was suspended in 200 µl of water. Thereafter, 8 µl of 5 M NaCl along with 500 µl of ice-cold ethanol were added to precipitate the RNA. The RNA was washed once with cold 80% ethanol. The pellet was resuspended in DEPC-treated water and stored at 70° C until use.

Twenty micrograms of total rat lung RNA in denaturation buffer were added to each well of a 1.2% agarose gel and separated overnight via electrophoresis. RNA was transferred from the gel to a nylon membrane (Immobilon-N) by capillary action overnight. Prehybridization solution and hybridization diluent consisted of 6× SSPE, 5× Denhardt's, 1% SDS, and 20 µg/ml salmon sperm DNA. Prehybridization took place for 3 to 5 h at 62° C, and hybridization took place overnight at 62° C. cDNA templates were random-primed with ³²P radiolabeled CTP and the Ready-To-Go DNA labeling kit (Pharmacia Diagnostics, Fairfield, NJ). Labeled probes were separated from unincorporated nucleotides using the TE Midi SELECT-D, G-50 centrifuge columns (5 Prime-3; Prime, Inc., Boulder, CO). Hybridized membranes were washed with 2× SSC with 0.5% SDS, then 0.2× SSC with 0.5% SDS at 37° C and 62° C. The hybridized autoradiographic signal was detected using Biomax Film (Kodak, Rochester, NY) and quantified with a densitometric scanner (BioRad Laboratories, Richmond, CA).

The rat PDGF-R plasmid was a generous gift from Randall R. Reed (Johns Hopkins Medical School, Baltimore, MD). The full-length PDGF-R probe was subcloned to include the first 1,718 bases, which encode for the extracellular portion of the receptor, to optimize probe specificity. This was accomplished with the use of the Not 1 restriction endonuclease and ligation as described by Sambrook and colleagues (21). The identity of the PDGF-R subclone was confirmed with both restriction digest analysis and sequencing. The PDGF-R cDNA was generated from a plasmid (generously provided by Micheal Pech, F. Hoffman-La Roche, Basel, Switzerland) (22). The PDGF-R cDNA included the 400 bases between the plasmid EcoRI and EcoRV sites; 36B4 (a human ribosomal phosphoprotein) was used as a loading control (23), in addition to ethidium staining of 28s and 18s ribosomal bands.

Western Analysis

Rats lungs were diced immediately after resection and dissociated in 1 ml of homogenization buffer (25 mM Tris at pH 7.6, 2 M EDTA, 0.5 M EGTA, 5 mM DTT) complete with antiprotease cocktail (PMSF, Aprotinin, Pepstatin, E-64) using the Tissue Tearor (Biospec. Products, Bartleville, OK). Then the samples were sheared through a 26-guage syringe five times and a 27.5-gauge syringe twice. Samples were next centrifuged at 2,000 × g at 4° C and the supernatant was transferred into an ultracentrifuge tube and centrifuges at 100,000 × g for 1 h at 4° C. The remaining supernatant was aspirated and discarded, and the remaining pellet was resuspended in 300 µl of homogenization buffer with antiprotease cocktail and 1% Triton-X. Samples were stored at 70° C until use.

Samples were removed from the 70° C freezer and their protein concentration was determined colorimetrically using the BioRad DC Protein Assay Kit and a spectrophotometer. Then the samples (100 µg/lane) were separated electrophoretically using Novex 8% Tris-glycine precast gels. Protein was transferred to nitrocellulose membrane (Hybond ECL; Amersham Life Science, Arlington Heights, IL) using a semidry transfer cell. Membranes were blocked overnight in 5% powdered milk in TBST before exposure to a 1:500 dilution of the primary antibody (rabbit anti-PDGF-R; Santa Cruz Biotechnology Inc., Santa Cruz, CA) for 1 h at room temperature in 2% milk-TBST. After washing, the membranes were incubated with a 1:1,000 dilution of horseradish peroxidase donkey antirabbit IgG in 2% milk-TBST for 1 h at room temperature. Then, after washing, the membranes were exposed to the ECL developer and exposed to ECL Hyperfilm (Amersham).

Immunohistochemistry

Immunohistochemical staining for the PDGF-R protein was performed using the immunoperoxidase technique previously described (20). Briefly, formalin-fixed deparaffinized lung tissue sections were preincubated in 0.03% hydrogen peroxide in methanol for 20 min to block endogenous peroxidase activity. Nonspecific antibody binding was blocked by a 30-min incubation in 2.5% fish gelatin-5% normal goat serum in 1% bovine serum albumin-PBS at pH 7.4. Then the slides were incubated at room temperature with a rabbit antirat PDGF-R IgG polyclonal antibody (no. 8526, 1 µg/ml; kindly provided by Dr. M. Pech, Hoffmann-Laroche Ltd., Grenzacherstr, Switzerland) in 0.1% gelatin-1% BSA-PBS for 1 h. After washing, the slides were incubated with a 1:4,000 dilution of biotinylated goat antirabbit IgG (Jackson Immunoresearch, West Grove, PA) in 0.1% gelatin-1% BSA-PBS for 1 h. Then the slides were incubated for 1 h with streptavidin-horseradish peroxidase (1:2,000; Jackson Immunoresearch) after washing. After further washing, the peroxidase activity was visualized with a 10-min incubation in 0.05 M Tris-HCl at pH 7.6 containing 0.02% diaminobenzidine (Sigma, St. Louis, MO) and 0.006% hydrogen peroxide. The slides were counterstained with Lerner-3 hematoxylin (Lerner Inc., Pittsburgh, PA). An equivalent dilution of normal rabbit IgG was used in place of the primary antibody as a specificity control.

Statistical Analysis

In order to obtain statistics from densitometric analysis of Northern or Western blots run on different days, the data were expressed as a ratio of asbestos-exposed over sham-exposed ± SEM. Student's t tests for unpaired values were employed to test for significance (p < 0.05).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Northern analysis revealed a striking and reproducible increase in PDGF-R mRNA isolated from asbestos-exposed lungs 5 h after the onset of exposure, compared with sham controls. A composite of consecutive PDGF-R Northern blots from four asbestos-exposed and four sham-exposed rats is shown in Figure 1. The size of the PDGF-R message was as expected (6.4 Kb) (4). Ratio of PDGF-R to loading control RNA were derived using a scanning densitometer. There was a statistically significant increase in PDGF-R mRNA expression in asbestos-exposed versus sham-exposed rat lung at the 5-h time point (2.55 ± 0.68, p < 0.01, n = 4). Data for the 1-, 2-, and 8-d time points were derived from three asbestos-exposed and three sham-exposed rats at each time point and are reported as a ratio of asbestos-exposed over sham-exposed. There were no significant differences in PDGF-R mRNA expression between the control and asbestos-exposed animals at the 1-d (0.68 ± 0.22), 2-d (0.67 ± 0.25), and 8-d (1.01 ± 0.29) time points (Northern blots to 1-, 2-, and 8-d time points not shown).

View larger version (39K): [in this window] [in a new window]   Figure 1.   Northern analysis for PDGF-R in RNA (20 µg) isolated from four asbestos-exposed (A) and four sham (air-exposed) (S) rat lungs at the end of a 5-h exposure to chrysotile asbestos or air. The 36B4 signal serves as a loading control. There is a reproducible upregulation of PDGF-R mRNA in rat lung in response to asbestos exposure.

Increased expression of PDGF-R mRNA was not seen after exposure to a nonfibrogenic particle (carbonyl iron). Northern analysis of RNA from the lungs of iron-exposed rats revealed no changes in PDGF-R expression at 5 h or at 1 or 2 d after exposure (Figure 2a). Similar results were obtained on a second Northern blot using another set of three iron- exposed rats killed at the 5-h and 1- and 2-d time points. In Figure 2b, a lack of change in PDGF-R expression can be seen at the 5-h time point and is representative of data from three pairs of asbestos- and sham-exposed rats. The ratio of PDGF-R mRNA expression in lungs of asbestos-expressed over sham-exposed rats at the 5-h time point was only 1.24 ± 0.26 (p = 0.4, not statistically significant). Thus, the increase in PDGF-R levels appears specific for this PDGF receptor subunit since PDGF-R mRNA levels appeared statistically unchanged in response to asbestos exposure at the time points tested (5 h, and 1, 2, and 8 d after exposure).

View larger version (27K): [in this window] [in a new window]   Figure 2.   (a) Northern analysis for PDGF-R in RNA (20 µg) from iron-exposed rat lungs at various time points after the onset of a 5-h exposure. Lane 1, 5-h asbestos; lane 2, 5-h air control; lane 3, 5-h iron; lane 4, 1-d iron; lane 5, 2-d iron. Iron exposure did not result in upregulation of the PDGF-R mRNA expression at the time points tested. (b) There is no change in the expression of PDGF-R mRNA immediately after a 5-h asbestos exposure in exposed rats (lane 6) compared with sham-exposed control rats (lane 7 ). This PDGF-R Northern blot is representative of three blots using different animals from the same exposure.

Lungs used for the Western analysis were derived from an identical exposure to the one reported above for the Northern analysis. A Western blot of two sets of lung membrane preparations from asbestos-exposed rats killed at 5 h, and at 1 and 2 d after the onset of the asbestos exposure is shown in Figure 3a. Lung membrane preparations were obtained from four asbestos-exposed animals at each of the three time points and from four sham-exposed animals killed at the 5-h time point. There is a clear increase in immunoreactive peptide at 175 kD in the asbestos-exposed rat lung compared with that in the sham-exposed lung at 5 h after onset of exposure.

View larger version (24K): [in this window] [in a new window]   View larger version (23K): [in this window] [in a new window]   View larger version (23K): [in this window] [in a new window]   Figure 3.   (A) Western blots for PDGF-R in two separate membrane preparations from both air and asbestos-exposed rat lung (100 µg/lane). Lanes 1 and 5, air 5 h; lanes 2 and 6, asbestos 5 h; lanes 3 and 7, asbestos 1 d; lanes 4 and 8, asbestos 2 d. These two Western blots are representative of four blots using different animals from the same exposure. (B) Western blot of a 5-h asbestos sample in which the primary antibody was incubated for 1 h at room temperature without (lane 1) or with (lane 2) 20 µg of the recombinant peptide used to generate the primary antibody prior to its addition to the membrane. The immunoreactive band seen in lane 1 is not present after preincubation of the primary antibody with the recombinant peptide, demonstrating that the band is specific for PDGF-R. (C ) A graph depicting the results of densitometric scanning of the Western analysis for PDGF-R on four separate sets of rat lungs from the same exposure. There is a significant increase in PDGF-R in the asbestos-exposed animals when compared with air control animals at 5 h after the onset of exposure, which gradually returns toward baseline by 2 d (p < 0.05).

The PDGF-R peptide is expected to be approximately 180 kD (24). Preincubation of the primary antibody with the recombinant PDGF-R peptide fragment (20 µg) used to generate the primary antibody resulted in the disappearance of the immunoreactive band and confirms the specificity of the Western analysis of the PDGF-R peptide (Figure 3b).

It can be seen in Figure 3c that there is a statistically significant increase in PDGF-R peptide expression in rat lung immediately after asbestos exposure. The ratio of PDGF-R immunoreactivity in asbestos-exposed rat lung over sham- exposed lung at the 5-h time point as determined by densitometry was 2.6 ± 0.3 (p < 0.05). The amount of receptor protein gradually decreased over time, and the ratio of asbestos exposed over sham-exposed lung was 2.2 ± 0.6 (p = 0.08) 1 d after exposure and 1.5 ± 0.4 (p = 0.4) by 2 d after exposure.

Immunohistochemical staining was performed to confirm the Western analysis and establish the distribution of the receptor protein. A would be expected for immunostaining of mesenchymal cells that bear the PDGF-R, the reaction product was confirmed to the lung interstitium. Figure 4 is representative of anti-PDGF-R immunostaining. Only trace immunoreactivity could be detected in sections from sham-exposed rat lung. Definitive staining was observed in the asbestos-exposed rat lung sections at both 5 h and 2 d after exposure. Replacement of the primary antibody with similar concentrations of rabbit IgG did not result in staining (data not shown). The staining is localized to the subepithelial portions of the ADBs and bronchioles. Notably, the upregulation of the PDGF-R immunostaining is seen at the sites of initial lung injury and the developing fibrotic lesions.

View larger version (151K): [in this window] [in a new window]   Figure 4.   Immunohistochemical staining for PDGF-R in sections from air- and asbestos-exposed rat lung at various times after the onset of exposure. Panel A, air-exposed at 5 h; Panel B, asbestos-exposed at 5 h; Panel C, an enlargement of the inset in Panel B; Panel D, asbestos-exposed at 2 d. The bars represent 20 microns. There is more staining apparent at the alveolar duct bifurcations and along the bronchioles in the asbestos-exposed rat lung sections than in the air-exposed sections. Staining is confined to the interstitial compartment with no staining apparent in the bronchiolar or alveolar epithelium (arrows = interstitial cells exhibiting abundant PDGF-R protein).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Asbestosis is a fibroproliferative lung disease in which asbestos exposure causes an increase in lung mesenchymal cell and interstitial matrix components (25). This distorts the lung architecture and eventually disrupts lung function as collagen deposition produces a "stiff" lung and reduced gas diffusion (14, 25). We have been studying the effects of asbestos fibers on lung fibroblast proliferation in animal models (13) as well as in vitro (18, 26). In human and in animal lungs, asbestos fibers are found in contact with and within interstitial lung fibroblasts (14). The mechanisms through which these fibers influence fibroblast proliferation are not well-defined. Hence, we have previously studied asbestos-induced lung fibroblast proliferation in vitro (18, 19), and here we have shown that similar events appear to be operative in the animal model. Studies of asbestos exposure in vitro have demonstrated that rat and human lung fibroblast proliferation are induced under serum-free conditions in cells that have been rendered quiescent (18, 27). Furthermore, the asbestos-induced proliferation can be blocked specifically with an anti-PDGF antibody. Northern analysis of RNA isolated from RLFs demonstrates that they make abundant PDGF-A-chain, but not PDGF-B-chain mRNA (18). In addition, the PDGF-R is upregulated after asbestos exposure, which in turn makes the fibroblasts more responsive to the PDGF-AA isoform that they secrete (19). We believe these data support the hypothesis that asbestos exposure causes fibroblast proliferation through a PDGF-A-chain/PDGF-R-dependent pathway in vitro.

The data presented here, along with our previous report describing the upregulation of the PDGF ligands in this model (20), establish that both components of the PDGF-A-chain pathway, ligand and receptor, are upregulated in vivo after asbestos exposure. PDGF-A-chain and PDGF-R expression are increased at a time just preceding the peak in asbestos- induced lung cell proliferation (15). There was a statistically significant increase in PDGF-R expression immediately after exposure (Figure 1), which returns to baseline by 1 d in the case of mRNA expression, and more gradually (by 2 d) in the Western analysis (Figure 3). Another method for investigating time-related PDGF-R peptide expression is shown by the strong immunohistochemical staining for the PDGF-R at 5 h and at 2 d postexposure (Figure 4). The immunohistochemistry data show that the PDGF-R staining is not limited to the ADB but also includes the subepithelial compartment along the small airways. The early, vigorous expression of the PDGF-R, taken together with our earlier work showing upregulation of PDGF-A-chain expression in this model (20), suggests that the PDGF-R may play a significant role in the initial events of asbestos-induced lung fibrosis. Although the majority of our work regarding the role of PDGF in asbestos- induced lung fibrosis has been with rats, we have also observed upregulation of PDGF-alpha receptor mRNA and peptide in lungs of asbestos-exposed mice (preliminary unpublished data). These findings regarding early changes in PDGF-alpha receptor expression after asbestos exposure in rats and mice, along with the known similarities in PDGF receptors and ligands among a number of species, suggest that these early events are relevant to lung injury in humans.

The increase in PDGF-A-chain (20) and PDGF-R expression are present at the site of the developing fibroproliferative lesions, at the ADBs. The finding of increased PDGF-R immunohistochemical staining immediately after the 5-h exposure complements the Western analysis. However, as shown in Figure 4, there was increased immunohistochemical staining at the 2-d time point, and this would not have been predicted based on our Western data in Figure 3. These studies show the value of using more than one analysis for in vivo receptor expression in such experiments. There are several possibilities to account for the observed differences in PDGF-R expression detected at 2 d after exposure between the Western analysis and immunohistochemistry. These possibilities include: (1) the variance in the Western results at the later time points, as demonstrated by the data at the 2-d time point (Figure 3c); or (2) that Western analysis and immunohistochemistry do not have similar sensitivity. However, the most likely explanation is that (3) the immunohistochemical staining is representative of events occurring precisely at the sites of lung injury, in contrast to the Western data, which are diluted by an analysis of whole-lung tissue.

PDGF-R mRNA expression appeared not to be upregulated in this model at the 5-h and the 1-, 2-, and 8-d time points, although PDGF-R mRNA was present in rat lung. However, our earlier studies demonstrated that PDGF-B-chain immunohistochemical staining is also upregulated at the ADB in this murine model of fibrosis (20). Thus, PDGF-B-chain-containing isoforms also may be involved in the pathogenesis of asbestos-induced lung fibrosis, but if so, the B-chain ligand probably is not regulated at the level of receptor mRNA expression.

There is less information in the literature regarding PDGF-A-chain and PDGF-R expression in lungs compared with what is known about PDGF-B-chain. This may be due to the fact that PDGF-AB and -BB are more potent mitogens for mesenchymal cells in vitro. However, the potency of PDGF-AA is directly dependent on the number of PDGF  receptors (19, 28). PDGF-AA is not a strong stimulus for proliferation in primary RLFs because of the low expression of the PDGF-R, but PDGF-AA mitosis does increase substantially after asbestos-induced upregulation of the PDGF-R expression in vitro (19). Swiss 3T3 cells have an equal number of PDGF-R and PDGF-R and have a mitogenic response to PDGF-AA, which is 70% of that induced by PDGF-BB (29).

The mechanism of asbestos-induced upregulation of PDGF-R, both in vitro and in vivo is unknown. Some cytokines are known to regulate the expression of PDGF-R in vitro. For example, IL-1, a cytokine released early in inflammation, is capable of upregulating PDGF-R expression and the subsequent fibroblast proliferative response to PDGF-AA (10). Basic fibroblast growth factor (30), thrombin (12), and inflammogenic lipopolysaccharide (31) have also been shown to increase PDGF-R expression in mesenchymal cells. Other cytokines such as transforming growth factor-beta may decrease PDGF-R expression (11). Interestingly, other investigators (32) have described mechanisms through which crocidolite asbestos upregulates the IL-8 promotor in an epithelial cell line. Apparently, phosphorylation events and asbestos-induced redox changes result in activation of nuclear proteins that recognize the NF kappaB/NF-IL-6 binding sites in the IL-8 promotor. Whether or not such events occur in vivo upon exposure to asbestos is not known at this time.

Unfortunately, it is not currently possible to study PDGF-AA or PDGF-R knockout animals to define the role of PDGF-AA-dependent mitosis in the formation of fibroproliferative lesions because either deletion is lethal (33, 34). Interestingly, the mice exhibit an underdevelopment of mesenchymal components in the lung. Development of such animal models will be essential for future studies to determine which factors are significant in the pathogenesis of fibrotic lung disease.