Main

High concentrations of inhaled oxygen are used in a variety of diseases to achieve adequate tissue oxygenation. Exposure of the developing lung to such high oxygen concentrations often results in permanent alterations in lung structure and function, including inhibition of alveolar formation, loss of surface area, and development of interstitial fibrosis(1–3). Components of the lung ECM, such as collagen and elastin, are active participants in the pathogenesis of oxygen-induced lung injury(3–8). Proteoglycans, another class of matrix molecules, are important regulators of collagen assembly, receptor-growth factor interactions, and enzyme sequestration in a variety of tissues(9–14). Changes in matrix proteoglycans have been observed with various forms of lung injury(15–17), including hyperoxic injury to the developing lung(18,19), and it is likely these matrix components play important roles during lung injury and repair.

There are at least five ECM-associated proteoglycans in lung. Of these, biglycan and decorin, the small (≈40 kD core protein) chondroitin/dermatan sulfate proteoglycans, share amino acid homology and bind TGF-β(20, 21). We have previously demonstrated an increase in biglycan core protein mRNA and immunostaining in developing rat lung after several weeks of chronic oxygen exposure(18). The closely related proteoglycan, decorin, which is widely distributed in the ECM in association with collagen fibrils(20, 22), is thought to regulate collagen fiber thickness in fibrotic reactions through inhibition of fibril formation(23) and is therefore likely to participate in the response to injury. We hypothesize that exposure of the developing lung to chronic hyperoxia will result in spatial and temporal changes in decorin expression. To test this hypothesis, newborn rats were exposed to hyperoxia for 6 wk and decorin expression was studied.

METHODS

Animals and exposures. Newborn Sprague-Dawley rats were exposed to continuous flow (4 L/min) 80-90% O2 for up to 6 wk as previously described(18). Briefly, oxygen-exposed animals(n = 70) were maintained in Plexiglas chambers where oxygen concentration was monitored twice daily, humidity maintained at greater than 80%, and CO2 removed by soda lime absorption. Nursing mothers were rotated between control (room air-exposed) and hyperoxia-exposed litters every 24 h. Control littermate pups (n = 72) were raised in the same room under normal vivarium conditions. Mean weights for control animals were similar to published norms for healthy rats(24), and oxygen-exposed animals gained weight at rates similar to control littermates(p = 0.38). Mortality was 30% in the oxygen-treated group, with most deaths occurring in wk 4 and 5, and 2% in the control group. At weekly intervals, animals were killed, and lungs from control and oxygen-exposed animals were perfused with ice-cold PBS, then either inflation-fixed at 20 cm H2O pressure with 10% buffered formalin, or flash-frozen in liquid nitrogen for RNA isolation. For cell studies, macrophages and PMNs were obtained by BAL from lungs at 6 wk before freezing for RNA isolation or from lungs used for type II cell isolation (lavaged lungs were not used for inflation fixation). Type II cells were isolated from control and oxygen-exposed lungs at 6 wk.

All study protocols were reviewed and approved by the University of North Carolina at Chapel Hill Institutional Review Committee for Animal Studies.

BAL and type II cell isolation. BAL cells and type II cells and were isolated from control and oxygen-exposed lung at 6 wk according to the methods of Finkelstein and Shapiro(25) with modifications. Briefly, after thoracotomy, the lungs were perfused in situ with ice-cold solution I (140 mM NaCl, 5 mM KCl, 2.5 mM Na2HPO4, 5 mM glucose, 10 mM HEPES, pH 7.4), removed, and lavaged with 5-mL aliquots of solution I to a total volume of 50 mL. The lavage fluid was then centrifuged at 500 × g to pellet BAL cells. Solution II (solution I + 1.9 mM MgCl2, 1.4 mM CaCl2, pH 7.4) containing Pen/Strep (Sigma, St. Louis, MO), elastase (Worthington, Freehold, NJ; 4.3 U/mL), and DNase I (Sigma; 1 mg/60 mL) was instilled into the lungs, which were then placed at 37 °C for 35 min. Enzyme activity was halted by instilling ice-cold Joklik's modified essential medium with DNase I(1 mg/100 mL), 10% calf serum, and trypsin inhibitor (250 mg/100 mL). The lungs were finely minced on ice, filtered, and separated from red blood cells by centrifugation through Percoll (Sigma; 1.08 g/mL). Remaining macrophages and fibroblasts were removed by IgG panning(26). These methods yielded >95% type II cells as determined by fluorescein-labeled Maclura pomifera lectin binding to type II cells and vimentin immunostaining for mesenchymal cells. For immunocytochemistry BAL cells and type II cells were centrifuged directly onto slides using the Cytospin apparatus (Shandon Lipshaw, Pittsburgh, PA), fixed in 95% ethanol, and air-dried. Cells for RNA isolation were quick frozen in liquid nitrogen and stored at -80 °C.

RNA isolation. Total RNA was isolated, as previously described(18), from individual control or oxygen-exposed rat lungs by homogenization in 4 M guanidinium thiocyanate and sedimentation through a 5.7 M CsCl cushion(27). Total RNA was isolated from type II cells using the Trizol reagent (Life Technologies, Inc., Gaithersburg, MD) and Trizol protocol. RNA concentrations were determined by spectrophotometric absorption at 260 nm, and RNA was stored frozen at -80 °C until used for Northern hybridization or RT-PCR analyses.

Northern hybridization analyses. For Northern hybridization analyses, 15 μg/lane total RNA was denatured in glyoxal and DMSO, size-fractionated by electrophoresis through a 1.0% agarose gel, transferred to nylon membranes, immobilized by UV cross-linking, and baking at 80 °C in a vacuum oven(28). Hybridizations were performed in hybridization buffer [6 × SSC (1 M NaCl, 0.1 M sodium citrate, pH 7.0), 10 × Denhardt's solution [0.1% BSA, 0.1% polyvinylpyrrolidone-360, 0.1% Ficoll (Sigma)], 2.0 mM EDTA, 0.1% SDS, 50% formamide, 0.01% sonicated salmon sperm DNA] containing 106 cpm probe/mL buffer at 42 °C for 16-20 h. After hybridization, membranes were washed in successive changes of 2 × SSC, 0.1% SDS and 0.1 × SSC, 0.1% SDS, as previously described(18). Hybridized membranes were subjected to autoradiography at -80 °C using Kodak X-AR x-ray film and Quanta III(DuPont, Wilmington, DE) intensifying screens. Changes in steady-state mRNA abundance was determined by measuring OD of hybridizing bands using the Image-Pro Plus Image Processing System (Media Cybernetics, Silver Spring, MD). Values for band densities were normalized to the 28 S ribosomal bands. Results for oxygen exposed lungs are expressed as percent of age-matched room air control lung expression.

Probe generation and labeling. A rat-specific cDNA for decorin(290 bp in length, corresponding to bases 519-809 of human decorin) was generated by RT of 5 μg of adult rat lung total RNA followed by amplification in the PCR using the following oligodeoxynucleotides: DEC5′ (5′- CTACATCCGCATTGCTGATACCAATATC-3®, identical to bases 519-546 of human decorin)(29) and DECAS[5′-GAAGGTAGAGACAACCTGGATGTACTTATG-3′, complementary to bases 782-809 of human decorin)(29). Amplified products were cloned into the plasmid vector PCR II (Invitrogen Corp., San Diego, CA) for propagation. Identity of the decorin cDNA was confirmed by sequence analyses(30). The cDNA for rat TGF-β1 was obtained from American Type Culture Collection (Rockville, MD) no. 63107(31). The cDNAs were radiolabeled to high specific activities (108 cpm/μg input DNA) by random hexamer labeling using the Prime-a-Gene system (Promega, Madison, WI) and [32P]dCTP (Amersham Corp., Arlington Heights, IL)(32).

cRNA probes. Riboprobes for in situ hybridization studies were made using linearized plasmid containing the full-length rat decorin clone(33) (a gift from Dr. Susan Abramson, University of Miami School of Medicine). Sense and antisense probes labeled with [33P]UTP were generated to give a specific activity of 1.67× 109 dpm/μg according to previously published methods(34). Probes were hydrolyzed to an average length of 200 bp in 40 mM NaHCO3, 60 mM Na2CO3, pH 10.2, at 60°C.

In situ hybridization histochemistry. Paraffin-embedded lung tissue sections 6 μm thick were cut and mounted on Probe on Plus slides(Baxter Scientific Products, McGaw Park, IL). Slides were baked overnight at 55 °C, then stored at room temperature until processed for in situ hybridization analysis. Slides were deparaffinized with xylene before hybridization. Type II cells and macrophages were centrifuged onto Probe on Plus slides, fixed in 70% ethanol, and stored at room temperature until processed for in situ hybridization. Before hybridization, all slides were rehydrated through a graded ethanol series. In situ hybridization was done according to previously published methods(35). Briefly, slides were treated with 1 μg/mL proteinase K for 30 min at 37 °C, equilibrated in 100 mM triethanolamine-HCl, and treated with fresh 0.25% acetic anhydride. Slides were then washed in 2 × SSC, dehydrated in a graded ethanol series, and air-dried. Prehybridization was at 53 °C for 2 h in hybridization buffer without dextran sulfate or probe. Slides were then rinsed in 2 × SSC and dehydrated. Hybridization was performed for 16 h at 53 °C using hybridization solution containing 39 ng/mL 33P-labeled decorin probe(6.5 × 107 dpm/mL). After hybridization, the slides were washed, treated with RNases A and T1, rinsed in RNase buffer, then passed through a series of washes of 2 × SSC, 0.1 × SSC at 60 °C, and 0.1× SSC at room temperature. The slides are dehydrated, dipped in a 1:1 dilution of NTB-2 photographic emulsion (Eastman Kodak, Rochester, NY), exposed at 4 °C for 2 wk, developed, and counterstained with hematoxylin and eosin before photomicrography. Slides of adjacent sections were hybridized with sense (control) probe to assess background binding.

Immunostaining. Lungs from control and oxygen-exposed animals were inflation fixed in 10% buffered formalin, dehydrated, paraffin embedded, and sectioned (6 μm). Cells from bronchoalveolar lavage or from type II cell isolation were centrifuged directly onto slides and fixed in 95% ethanol. Decorin immunostaining was performed as previously described(18), using the rabbit polyclonal antibody LF 113, which was generated against a synthetic peptide corresponding to amino acids 38-48 of mouse decorin (a gift from Dr. Larry Fisher of the National Institutes of Health, Dental Research). This antiserum cross-reacts with rat decorin, but no cross-reactivity is observed with the closely related proteoglycan, biglycan(L. Fisher, personal communication). A set of sections was pretreated with chondroitinase ABC lyase (Sigma Chemical Co., St. Louis, MO) 0.1 U/mL for 30 min before immunostaining to determine whether changes in immunoreactivity might be related to changes in glycosaminoglycan side chain masking of peptide epitopes. Antibody preabsorbed with the synthetic immunogen was used as a negative control for immunostaining. Antibody-antigen complexes in tissue sections were visualized by the avidin-biotin-alkaline phosphatase method using the Vectastain ABC-AP kit (Vector Laboratories, Burlingame, CA) and the substrate Vector Red. Slides were counterstained with Harris hematoxylin, dehydrated in ascending ethanol concentrations, and coverslipped.

RT-PCR. Five micrograms of type II cell total RNA from control or oxygen-exposed rat lung were reverse-transcribed for 1 h at 37 °C using Superscript reverse transcriptase (Life Technologies, Inc.) and the decorin 3′ oligo (DECAS) in a total reaction volume of 20 μL according to Frohman et al.(36). One microliter of the RT product was used in a PCR amplification with the DECAS and DEC5′ oligos and Taq polymerase (Boehringer Mannheim, Indianapolis, IN)(36). PCR conditions were 1 cycle of denaturation at 94°C, 1 min of annealing at 60 °C, and 2 min of extension at 72 °C, followed by 36 cycles of 2 min of denaturation at 94 °C, 1 min of annealing at 60 °C, and 1 min of extension at 72 °C. RNA integrity was determined by gel electrophoresis and ethidium bromide staining before RT, and the rat decorin cDNA was used as a positive control. The PCR products were separated by electrophoresis in a 1.0% agarose gel, stained with ethidium bromide, and photographed.

Statistical analysis. Where indicated, statistical analysis of data were performed using the program StatView II (Abacus Concepts, Inc., Berkeley, CA) and the 2-tailed, unpaired t test. Results are expressed as the mean ± SEM.

RESULTS

Cloning of a rat decorin cDNA. A rat decorin cDNA was generated by RT-PCR for use in Northern hybridization analyses using total RNA extracted from adult rat lung. The 290-bp cDNA is 90% homologous with human decorin(29) at the nucleotide level. The rat decorin sequence can be found under accession no. L75825 in GenBank/MBIR (Santa Fe, NM).

Decorin mRNA expression in normal and injured whole lung. Northern hybridization analyses of RNA extracted from control (n = 5) and oxygen-injured lungs at 1, 2, 3, 4, 5, and 6 wk (n = 5 at each time point) demonstrate a decrease in the relative abundance of decorin mRNA with chronic oxygen exposure (Fig. 1A). Densitometric analyses of the autoradiograms reveal that, relative to control lung mRNA expression, exposure to hyperoxia is associated with a progressive reduction in whole lung decorin mRNA. After 6 wk of hyperoxia, decorin mRNA decreases to 23% of control levels (p = 0.0001) (Fig. 1B).

Figure 1
figure 1

Decorin expression in control and oxygen-injured newborn rat lung. (A) Representative autoradiograph from Northern hybridization analyses of control and oxygen-injured lung RNAs. Each lane represents 15 μg of total RNA hybridized with a [32P]dCTP-labeled cDNA for decorin (lower panel). Age in weeks is given at the bottom of the figure, with (*) indicating RNA from oxygen-exposed animals. Top panel shows ethidium bromide staining of 28 S ribosomal RNA. (B) Graphic summary of relative changes in steady-state abundance of decorin mRNA in animals exposed to hyperoxia for up to 6 wk expressed as a percentage of age-matched control values. Results are presented as means ± SEM of at least six animals. p values are from 2-tailed, unpaired t tests comparing age-matched control and oxygen-exposed animals.

Full size image

TGF- β mRNA expression in normal and injured lung. The expression of TGF-β mRNA was determined in RNA extracted from control and oxygen-injured lung at 1, 2, 3, 4, 5, and 6 wk. Densitometric analyses of the autoradiograms failed to demonstrate any change in TGF-β expression with exposure to hyperoxia (results not shown).

Localization of decorin RNA and core protein in lung. To localize decorin expression in whole lung, in situ hybridization was used to demonstrate decorin mRNA in normal and oxygen-exposed lung. Representative bright and darkfield microscopic images of lung sections from control and oxygen-exposed animals at 2, 4, and 6 wk are shown(Figs. 2 and 3, respectively). In control lung(panels A, C, and E) decorin mRNA is scant at 2 wk (Fig. 3A), increases in cells surrounding airway epithelium at 4 wk (Fig. 3C; arrow) and is abundant in cells of the loose connective tissues surrounding large airways (a) and blood vessels (v) at 6 wk of age (Fig. 3E; arrows). Exposure to chronic hyperoxia (panels B, D, and F) results in no change at 2 wk (Fig. 3B), but at 4 wk (Fig. 3D) there appears to be a generalized decrease in expression with localized areas of increase in perivascular tissues (arrows). After 6 wk of hyperoxia (Fig. 3F), loose connective tissue expression appears unchanged, whereas fibroblast-like cells in focal areas of thickened alveolar septa demonstrate increased expression relative to surrounding lung (arrows). Cells isolated from bronchoalveolar lavage and type II cells from control and oxygen-exposed lung failed to demonstrate decorin mRNA by in situ hybridization techniques (results not shown).

Figure 2
figure 2

Morphology of lung sections used for in situ hybridization histochemistry. Shown are photomicrographs of lung sections from control (panels A, C, and E) and oxygen-exposed (B, D, and F) animals at 2 wk (panels A and B), 4 wk (panels C and D), and 6 wk (panels E and F) corresponding to darkfield photomicrographs shown in Figure 3. Oxygen exposure is associated with a loss of surface area and abnormal alveolarization.

Full size image
Figure 3
figure 3

In situ hybridizations for decorin RNA in control and oxygen-exposed rat lungs. Darkfield images of lung sections from control (panels A, C, and E) and oxygen-exposed (B, D, and F) animals at 2 wk (panels A and B), 4 wk (panels C and D), and 6 wk (panels E and F). Lung tissue sections were hybridized with 33P-labeled decorin antisense riboprobe, and the emulsion was exposed for 2 wk. Large airways (a) and blood vessels (v) are labeled for clarity. At 2 wk, decorin mRNA is scant in control (A) and oxygen-exposed(B) lung. At 4 wk, there is an increase in decorin RNA in cells surrounding airway epithelium in control lung (C; arrow) and a generalized decrease in expression with localized areas of increase in perivascular tissues (D; arrows) with oxygen exposure. In control lung at 6 wk, most decorin RNA is localized to cells in the loose connective tissues surrounding large airways and blood vessels (E; arrows), whereas exposure to hyperoxia results in focal increases in decorin RNA in thickened areas of alveolar septa (F; arrows). Bar = 100 μm.

Full size image

Figure 4 shows representative photomicrographs of lung sections from control and oxygen-exposed animals at 2, 4, and 6 wk, immunostained with antibody LF 113, which recognizes the core protein of decorin. In control lung (panels A, C, and E), loose connective tissue (arrows) surrounding bronchi (b) and blood vessels (a, artery; v, vein) demonstrates strong immunoreactivity for decorin (red staining), and airway epithelium also demonstrates immunoreactivity at 4 and 6 wk (open arrows, panel C; inset and panel E). At 6 wk in control lung, focal staining of alveolar walls is evident (arrow, panel E; inset). With oxygen exposure there is a decrease in connective tissue immunostaining at 2 wk(arrows, panel B) and no discernible change in connective tissue staining at 4 and 6 wk (panels D and F). Oxygen-exposed lung does not demonstrate alveolar wall immunoreactivity at 2 wk (panel B; inset), but at 4 and 6 wk, focal increases in decorin immunostaining in alveolar walls is evident (short arrows, panels D and F; insets). Pretreatment of sections with chondroitinase ABC lyase before immunostaining had no effect on immunoreactivity in sections from control or oxygen-treated animals (results not shown).

Figure 4
figure 4

Localization of decorin protein in control and oxygen-injured lung. Photomicrographs of decorin immunostaining (red staining) at 2 wk (panels A and B), 4 wk (panels C and D), and 6 wk (panels E and F) in control(panels A, C, and E) and oxygen-injured lung(panels B, D, and F). Large vessels (a, artery; v, vein) and bronchi (b) are labeled for clarity.

Full size image

To determine the cellular localization of decorin core protein in lung, and to begin to investigate which cells other than fibroblasts contain decorin, type II cells and BAL cells were immunostained for decorin.Figure 5 demonstrates decorin immunostaining of isolated type II cells (panels A and B) and BAL cells(panels C and D) from 6-wk control (panels A and C) and 6-wk oxygen-exposed (panels B and D) animals. Type II cells and BAL cells from control lung do not stain for decorin. With oxygen exposure, PMNs (panels B and D; arrows) as well as occasional macrophages (panel D; arrowhead) stain for decorin (red staining). All slides were processed together, and care was taken during photography and photo processing to ensure that exposure times were constant, so that direct comparisons between photomicrographs could be made.

Figure 5
figure 5

Immunostaining of isolated type II cells and bronchoalveolar lavage cells from control and oxygen-injured lung. Type II cells from both control (A) and oxygen-exposed lung (B) fail to stain for decorin, whereas immunoreactivity for decorin is evident in PMNs contained in the type II cell preparation from injured lungs (B; arrow). Large mononuclear cells isolated by bronchoalveolar lavage from control lung (C) do not stain for decorin; however, with oxygen-exposure, PMNs (D; arrow) and rare mononuclear cells(D; arrowhead) demonstrate positive immunostaining for decorin. All cells were obtained from rat lungs after 6 wk of exposure to either room air or hyperoxia.

Full size image

Analysis of decorin mRNA expression by type II pneumocytes. To determine whether alveolar cells containing decorin mRNA and staining for decorin core protein were type II cells, and because previous investigators have speculated that type II cells synthesize decorin(37), isolated type II cells were examined for expression of decorin mRNA. Because of the small quantities of type II cell RNA obtained, RT-PCR was used to analyze decorin mRNA expression. Figure 6 shows ethidium bromide staining of the amplification products obtained from RT-PCR using RNA extracted from type II cells isolated from normal (lanes 1-3) and oxygen-exposed lungs (lanes 4 and 5). There were no detectable decorin amplification products obtained using type II cell RNA from control or oxygen-exposed lung. Lane 6 contains the amplification product obtained using the rat decorin cDNA as a positive control. This indicates either the absence of decorin mRNA in type II cells or that decorin transcripts are present in such low abunadance that they are undetectable by an extremely sensitive assay.

Figure 6
figure 6

RT-PCR analysis of decorin expression in type II cells. Photograph of ethidium bromide-stained RT-PCR-amplified products from a representative experiment. Lanes 1-3 contain RT-PCR products amplified from type II cell RNA isolated from control animals, lanes 4 and 5 contain RT-PCR products amplified from type II cells isolated from oxygen-exposed animals, and lane 6 contains the RT-PCR product amplified from rat decorin cDNA control. Lane 7 contains an RNA ladder size marker. Sizes of molecular size standards are indicated by arrows on the right.

Full size image

DISCUSSION

Proteoglycans are noncollagenous components of the ECM that change in response to different forms of injury(15, 17–19, 38–40) and are likely to influence the healing process through their effects on matrix assembly and cell adhesion, and through growth factor interactions(23, 41). In this study, exposure of newborn rats to hyperoxia is associated with a decrease in whole lung decorin mRNA and with focal increases in decorin mRNA and protein in areas of distal lung. In normal tissues, decorin is widely distributed in the ECM of normal tissues in association with collagen, and a distinct pattern of developmental expression has been described(42, 43). Respiratory distress syndrome in the premature non-human primate is associated with decreases in collagen-associated dermatan sulfate proteoglycan, although the specific proteoglycans affected have not been identified(17). Injury related changes in decorin expression have been demonstrated in experimentally induced liver fibrosis(44, 45), tissue damage in brain(46), and in bleomycin-induced pulmonary fibrosis(15). The regulation of developmental as well as injury-related changes in decorin expression is not well understood, although in cell culture studies, dexamethasone(47) and basic fibroblast growth factor(48) increase decorin expression, and TGF-β decreases its expression(42). We examined the expression of TGF-β in whole lung from control and oxygen-exposed animals and did not find changes in the abundance of TGF-β mRNA with exposure to hyperoxia, although localized changes in TGF-β expression or activity which might affect decorin expression would not necessarily be reflected in whole lung RNA studies.

Localization of decorin mRNA and core protein in normal and oxygen-injured lung demonstrates message and protein in loose connective tissue surrounding large airways and blood vessels as well as in airway epithelial cells and interstitial cells of the peripheral lung. In contrast to Northern hybridization studies, in situ and immunohistochemical techniques demonstrate increases in decorin message and protein in focal areas of oxygen-injured lung after 6 wk of hyperoxia. It is likely that the decrease in whole lung decorin RNA detected by Northern blot hybridization is due either to a small decrease in decorin RNA abundance in a large population of cells or to a decrease in the number of cells expressing decorin RNA. Small changes in decorin expression by a large number of cells would not likely be detected by in situ hybridization or immunohistochemical techniques.

Because oxygen-induced injury to the developing lung is a heterogeneous process characterized by areas of cell proliferation and fibrosis alternating with areas of impaired cell division and coalescence of airspaces(3, 49), it is not surprising that there are regional differences in the expression of molecules that likely participate in the pathogenesis of such injury. Submucosal fibromuscular tissues such as those surrounding large vessels and airways often demonstrate fragmentation of elastic tissue, damage to smooth muscle cells, and collagen deposition(50). Because decorin binds to triple helical collagen and appears to inhibit collagen fibril formation(23, 51, 52), decreased decorin concentrations in these areas might contribute to increased mature collagen formation and fibrosis. Alternatively, at the alveolar level, hyperoxia causes impaired septal formation and leads to a distorted alveolar architecture consisting of large airspaces and irregular tissue masses of undifferentiated cells, PMNs, and collagen bundles(1, 3, 50). Increased decorin expression in these areas may serve to limit collagen fibrillogenesis resulting in a less organized connective tissue. Decorin also binds to fibronectin(53) and thrombospondin(54), inhibiting cell adhesion(55, 56), which might be an additional mechanism by which decorin contributes to the abnormal alveolarization seen in hyperoxic lung injury.

A number of growth factors are involved in the response of the lung to injury. TGF-β plays a key role in ECM gene expression, acting as a powerful modulator of matrix synthesis and degradation, important aspects of connective tissue remodeling in fibrosis(57, 58). Decorin modulates TGF-β activity by reversibly binding to the growth factor and neutralizing its activity(41, 59). The removal of such an inhibitory influence over TGF-β may promote ECM deposition, fibrosis, and tissue remodeling around large vessels and airways, whereas an increase in such regulation in distal lung may lead to disruptions in alveolar formation, such as those described with chronic oxygen-induced injury.

The expression of decorin by fibroblasts is well described(48, 60–62), and it is likely that the peripheral lung cells containing decorin RNA and core protein are fibroblasts. Analyses of BAL cells isolated from oxygen-injured lung indicate that PMNs and rare alveolar macrophages contain decorin core protein but not decorin mRNA, indicating that decorin in these cells is likely taken up by phagocytosis. Because decorin core protein immunoreactivity in distal lung was seen at the corners of alveolar septa and in areas with increased numbers of cuboidal cells, and because of speculation that the small chondroitin/dermatan sulfate proteoglycan synthesized by type II cells might be decorin(37), we investigated the expression of decorin by type II cells. Our studies demonstrate that type II cells isolated from normal and oxygen-exposed lungs do not contain detectable RNA or protein for decorin.

The temporal and spatial changes in decorin expression demonstrated in these studies support a role for this proteoglycan in the cellular response to oxygen-induced lung injury. It is likely that regional differences in decorin expression are related both to the influence of factors produced and acting locally and to regional differences in cell populations. We speculate that changes in decorin expression may contribute to the architectural changes seen in chronic oxygen-induced injury to the developing lung through interactions with extracellular matrix components and growth factors.