Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Hirai and coworkers (1) originally reported that a novel, 150-kD stromal protein, epimorphin, is essential for epithelial morphogenesis in both skin and lung. In organ cultures of embryonic tissues, a monoclonal antibody (Ab) against epimorphin inhibited various components of epithelial morphogenesis, including tubular formation, ductal branching formation, and luminal formation (1, 2). When NIH 3T3 cells were transfected with epimorphin complementary DNA (cDNA), they expressed the exogenous epimorphin on their cell surfaces; whereas lung epithelial cells cocultured with the transfectants exhibited normal tubular morphogenesis, epithelial cells cultured with untransfected NIH 3T3 cells did not (1). Moreover, in three-dimensional culture systems, epimorphin was found to be expressed in fetal lung and skin rudiment cells, where it was localized to the mesenchymal-epithelial interface (1), and in endothelial cells, where it stimulated development of a branched morphology (3).

Epimorphin is a member of a gene family that includes SED 5, Pep 12, Sso 1p, Sso 2p, and the syntaxin family. In particular, syntaxin 2 shares 96% amino-acid identity with mouse epimorphin, and it has been suggested that syntaxin 2 is, in fact, rat epimorphin (4). The epimorphin gene is known to be highly conserved among mice, rats, and humans (1, 5): the 289 amino-acid sequence of rat epimorphin/syntaxin 2 exhibits 86% homology to human epimorphin (6). The unit structure of epimorphin predicted from the cDNA sequence contains three major domains, with the epitope for the monoclonal anti-epimorphin Ab located in the central domain (7).

It was recently reported that whereas epimorphin is not mitogenic, it is a primary morphogen in mammary epithelial cells, acting in concert with epidermal growth factor, fibroblast growth factor, keratinocyte growth factor, and hepatocyte growth factor (2). Indeed, morphogenesis of epimorphin-negative epithelial cells (Scp2, a mouse mammary epithelial cell line) was induced by application of epimorphin alone but not by any growth factors alone (2). Conversely, morphogenesis of epithelial cells expressing epimorphin (D6 and I6, mouse mammary epithelial cell lines) was completely blocked by anti-epimorphin Ab, even in the presence of growth factors (2). Using primary cultured rat hepatocytes, a form of epithelial cell, it was also shown that epimorphin induced the formation of hepatocyte spheroids with a bile canaliculi-like structure, which maintained albumin production even in the absence of growth factors (8). Furthermore, expression of epimorphin messenger RNA (mRNA) was increased during hepatocyte repair after partial hepatectomy or damaged by CCI₄ administration (9, 10).

The fetal development and morphogenesis in the hair follicle, tooth, salivary gland, mammary gland, kidney, liver, pancreas, and lung depend on epithelium-mesenchyme interactions (11, 12); such interactions are also believed to be important for the tissue regeneration necessary for wound healing. Although repair after injury in the adult lung involves some of the same factors that regulate lung development (13, 14), the distribution and function of epimorphin in the postnatal lung, including the situation of injury, are entirely unknown.

Pulmonary fibrosis is thought to be a result of the process of wound healing or regeneration after lung injury. The associated re-epithelialization is similar to the process of fetal epithelial development (14), suggesting that epithelium-mesenchyme interactions may play a key role in the development of pulmonary fibrosis. The process takes place mainly within the intra-alveolar space (15). The epithelial basement membrane is destroyed, enabling migration of interstitial cells into intra-alveolar spaces, where they produce and deposit extracellular matrices (ECM). The surface of the intra-alveolar fibrosis is then overlaid by regenerating epithelial cells, and the resultant fibrotic mass is incorporated into the alveolar wall (15, 16, 19, 20, 22).

Bleomycin-induced lung fibrosis in mice is a well established histologic and biochemical model of human pulmonary fibrosis (22). In the present study, we used that model to assess the temporal and spatial changes in the localization of epimorphin protein and in the expression of its mRNA during development of fibrotic lesions. We discuss the process of pulmonary fibrosis in the context of our findings, taking into consideration epithelium-mesenchyme interactions in wound healing.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animal Models

Pulmonary fibrosis was induced in 8-wk-old male ICR mice (35 to 40 g) by a single intratracheal instillation of bleomycin hydrochloride (25 mg/kg body weight; Nippon Kayaku Co., Tokyo, Japan), which was administered intraperitoneally under pentobarbital sodium anesthesia using a 25-gauge needle inserted between the cartilaginous rings of trachea. Control animals were injected with an identical volume of saline. On selected days after injection (Days 0, 3, 7, 14, 21, 28, 42, and 56), eight bleomycin-treated and five control animals were anesthetized by an intraperitoneal injection of 40 mg/ kg of pentobarbital sodium and exsanguinated by cutting the abdominal aorta. The trachea was then cannulated, and the lungs were removed from the thorax. The right lungs were immediately frozen at 80°C and later used for RNA extraction. The left lungs, which were used for microscopy, were fixed for 8 h at 4°C in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) and inflated at a pressure of 20 cm H₂O. They were then sequentially washed in 10, 20, and 30% sucrose in 0.01 M phosphate-buffered saline (PBS; pH 7.4, 4°C) for 4 h, later washed with 30% sucrose in 0.01 M PBS overnight, after which they were snap-frozen in O.C.T. embedding medium and stored at 80°C.

Light Microscopic Immunohistochemistry

Frozen tissues were cut into 4-µm sections and incubated overnight at 4°C with 10 µg/ml of rat antimouse monoclonal epimorphin Ab (MC-1, a gift from Dr. Hirai), and then for 1.5 h at 37°C with 1 µg/ml biotinylated goat antirat immunoglobulin (Ig) G (ZYMED, San Francisco, CA) and streptavidin-biotin-horseradish peroxidase complex (Dakopatts, Glostrup, Denmark). The bound Ab was visualized by incubation for 10 min in a Coplin jar with 100 ml Tris HCl buffer (pH 7.6) containing 20 mg of diaminobendizine (DAB) and 17 µl of H₂O₂, and counterstained with Mayer's hematoxylin. Nonspecific labeling of primary antibody was evaluated using normal rat serum. Serial sections of each tissue sample were stained for keratin (rabbit antibovine Ab raised against the 58-, 56-, and 52-kD subunits of muzzle epidermal keratin; Dacopatts, Santa Barbara, CA), vimentin (goat polyclonal antihuman vimentin Ab; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), or type IV collagen (goat polyclonal antihuman type IV collagen Ab; Southern Biotechnology Associates, Inc., Birmingham, AL). Serial sections were also stained with hematoxylin and eosin and elastica Masson-Goldner (EMG).

Confocal Microscopy

To more precisely localize epimorphin, epithelial and stromal cells were double-labeled using immunofluorescent probes, and the distribution of epimorphin was compared with that of keratin or vimentin in the same samples. Briefly, after sections were exposed to primary Abs, they were exposed to fluorescein isothiocyanate (FITC) or Texas red-conjugated second Abs (FITC-conjugated goat antirat IgG, American Qualex, San Clemente, CA; Texas red-conjugated goat antirabbit IgG, Molecular Probes, Inc., Eugene, OR; FITC-conjugated rabbit antirat IgG, ZYMED; Texas red-conjugated rabbit antigoat IgG, EY Laboratories, Inc., San Mateo, CA), and the nuclei were counterstained with 4,6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Inc., Burlingame, CA). Specimens were examined under a confocal laser scanning microscope (CLSM, TCS-SP; Leica Lasertechnik, Heidelberg, Germany) based on an upright microscope (DMRB; Leica Lasertechnik) equipped with a krypton/argon laser (23). The excitation wavelengths for FITC and Texas red were 498 and 568 nm, respectively. Green FITC emission was selected and recorded using a 500 to 550 nm bandpass filter, whereas red Texas red emission was selected and recorded using a 581 to 631 nm bandpass filter. In addition, DAPI was excited at 350 nm using a ultraviolet laser, and its blue emission was recorded using a 401 to 551 nm bandpass filter. Each field was also observed using Nomarski optics. Images were collected using a Leitz 63× PL APO objective (numerical aperature [NA] = 1.4). The means of 32 scans were obtained at a resolution of 512 × 512 pixels and analyzed using TCS-NT (Leica Lasertechnik). Figures were printed on a Fuji 3000 Pictrography printer (Fuji Medical Systems, Tokyo, Japan).

In Situ Hybridization

The RNA probes used for in situ hybridization consisted of the 1.9-kb BamH1/Xba1 fragment of epimorphin cDNA containing the open reading frame, which was obtained by transcription of the entire epimorphin cDNA template (a gift from Dr. Hirai). Sense and antisense RNA probes were labeled with digoxigenin (DIG) by incorporating DIG-11-deoxyuridine triphosphate using T3 and T7 promoters (Promega, Madison, WI) and a DIG RNA labeling kit (Boehringer-Mannheim, Mannheim, Germany), after which they were hydrolyzed to a length of ~ 300 bp by alkaline hydrolysis. For hybridization, 5-µm sections of the lung tissue harvested on Day 28 after bleomycin treatment were incubated overnight at 45°C with samples of the labeled RNA probe (1 µg/ ml). The sections were then washed under stringent conditions to reduce the background and incubated overnight at 4°C with alkaline phosphatase-labeled anti-DIG Ab (Boehringer-Mannheim). The resultant hybridization signal was visualized with 5-nitroblue tetrazolium chloride, as recommended by Boehringer-Mannheim.

Immunoelectron Microscopy

Blocks of the lung tissue harvested on Day 28 after bleomycin treatment were fixed for 8 h in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4), cut into 6-µm frozen sections, and incubated first with anti-epimorphin Ab overnight at 4°C, and then with horseradish peroxidase conjugated goat-rat IgG F(ab') 2 (Biosource International, Camarillo, CA). The sections were then fixed an additional 5 min in 1% glutaraldehyde, reacted with DAB and H₂O₂, postfixed in 2% osmium tetroxide for 60 min, dehydrated in a graded ethanol series, and embedded in Epon 812 resin. The ultrathin sections were observed using a transmission electron microscope (model H 7100; Hitachi Ltd., Tokyo, Japan) at 75 kV with either no counterstaining or slight counterstaining with lead citrate.

Northern Blot Analysis

Total cellular RNA was isolated from lungs on Days 0, 3, 7, 14, 28, 42, and 56 after bleomycin administration using the acid guanidium-thiocyanate-phenol-chloroform method. The RNA samples (10 µg/lane) were fractionated by electrophoresis on 1% agarose-formaldehyde gels under denaturing conditions and transferred to Nytran by capillary action. The blots were then probed with the BamH1/Xba1 fragment of epimorphin cDNA labeled with [³²P]deoxycytidine triphosphate (3,000 Ci/mmol) using a Nick Translation System kit (GIBCO/BRL, Grand Island, NY). After hybridization, the blots were washed and autoradiographed. Washed blots were analyzed using a Fujix BAS2000 Bio-imaging Analyzer System (BAS; Fuji Photo Film Co., Ltd., Tokyo, Japan) and visualized on Kodak X-OMAT AR film (Eastman Kodak, Rochester, NY). Signal intensity was quantified in digital images using BAS analysis (FUJIX BAStation, Fuji Photo Film Co., Ltd.). To control for differences in gel loading (24, 25), for each sample, RNA hybridized with the epimorphin probe was normalized to the expression of -actin mRNA in the same sample, and values obtained were then expressed as a percent of control (Day 0).

Statistics

All values were expressed as the mean ± standard deviation. Comparisons were made using Mann-Whitney tests and unpaired t tests. Values of P < 0.05 were considered to be statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Immunohistochemical Analyses

Control lungs. Immunohistochemical staining revealed that in the normal adult mouse lungs (Day 0), epimorphin was distributed in the connective tissue components of the walls of the bronchi and the bronchioles, and was also weakly detected in the vascular and alveolar walls (Figure 1a). On Day 3 or Day 7 after instillation of normal saline, mild edema of the interstitial connective tissue and infiltration of a few inflammatory cells were observed. Nonetheless, the pattern of epimorphin immunoreactivity was very similar to that seen before treatment (Day 0). Furthermore, no specific changes were found among lung tissues harvested on Day 28 or Day 56. 

View larger version (96K): [in this window] [in a new window]   Figure 1.   Immunohistochemical detection of epimorphin in the lungs of normal adult mice (a) and 3 d after intratracheal bleomycin administration (b). Epimorphin immunoreactivity is seen in the bronchiolar, vascular, and alveolar walls of normal lungs, but the intensity of the reaction is mildly increased on Day 3 after bleomycin. (c and d) In lungs harvested on Day 7 after bleomycin, epimorphin immunoreactivity is more intense, and areas of epimorphin-positive, early intra-alveolar fibrosis can be seen (arrowheads). All bars = 50 µm.

Bleomycin-treated lungs. In contrast, acute inflammatory changes, including edema and infiltration of neutrophils and mononuclear cells, were observed by Day 3 after instillation of bleomycin, mainly around the alveolar ducts. A few polymorphonuclear leukocytes, mixed with occasional desquamated epithelial cells and alveolar macrophages, were found in the air spaces. As with normal lungs, epimorphin immunoreactivity was found in the bronchiolar, alveolar, and vascular walls, but the intensity of the labeling appeared stronger than in the control lungs (Figure 1b).

By Day 7, inflammatory changes were apparent and more diffuse, and early fibrotic lesions associated with inflammatory cells were detected. The thickened alveolar septa associated with lymphomonocytic infiltration encroaching upon the lumens were notable. Some interstitial cells migrated into the intra-alveolar spaces together with inflammatory cells (Figures 1c and 1d). Epimorphin immunoreactivity in the bronchiolar, alveolar, and vascular walls was stronger than on Day 3, and positive staining was also observed in the areas of early intra-alveolar fibrosis (Figure 1d).

Fourteen days after administration of bleomycin, intra- alveolar fibrosis was clearly detectable, and epimorphin immunoreactivity was detected in the alveolar walls and in the areas of intra-alveolar fibrosis (Figure 2a). At this time, keratin-positive epithelial cells were observed in the normal alveolar walls, but positive labeling for keratin was not yet present in the intra-alveolar fibrosis (Figure 2b). This finding was confirmed by confocal immunofluorescent images showing the absence of epimorphin within keratin-positive regenerating epithelial cells in areas of intra-alveolar fibrosis (Figure 2c).

View larger version (97K): [in this window] [in a new window]   Figure 2.   Immunohistochemical analysis of the lungs harvested on Day 14 after bleomycin treatment. (a) Epimorphin immunoreactivity is detected in alveolar walls and intra-alveolar fibrotic areas. (b) Keratin immunoreactivity is detected in alveolar epithelial cells of normal alveolar walls but not in intra-alveolar fibrotic lesions that lack epithelial cells. (c) Fluorescent confocal image showing the codistribution of epimorphin (green) and keratin (red) immunoreactivity. Note the keratin-positive epithelial cells and the epimorphin-positive intra-alveolar fibrosis. Nuclei are shown as DAPI-positive (blue). Asterisks denote intra-alveolar fibrosis. All bars = 20 µm.

Strong epimorphin immunoreactivity was seen in alveolar walls and in the areas of intra-alveolar fibrosis 21 d after administration of bleomycin (Figure 3a). Moreover, keratin labeling was seen not only in alveolar epithelial cells but also in the regenerating epithelial cells that partially overlaid the intra-alveolar fibrotic lesions (Figure 3b). In addition, immunolabeling of vimentin and type IV collagen, respectively, showed that the fibrotic lesions contained spindle mesenchymal cells (Figure 3c) and that the fibrosis had formed within the intra-alveolar spaces (Figure 3d). Double-labeling confocal fluorescent images clearly confirmed that at this stage, epimorphin-positive, intra-alveolar fibrotic lesions were being re-epithelialized by regenerating keratin-positive epithelial cells (Figure 3e). The same view of confocal images using Nomarski optics clearly showed the intra-alveolar fibrotic lesions with three-dimensional effect and matching the colors of confocal images (Figure 3f). Double-labeling confocal fluorescent images also clearly confirmed that epimorphin was localized in vimentin-positive stromal cells and in surrounding ECM (Figures 3g, 3h, and 3i).

View larger version (64K): [in this window] [in a new window]   Figure 3.   Immunohistochemical analysis of the lungs harvested on Day 21 after bleomycin treatment. (a) Epimorphin immunoreactivity present in an area of intra-alveolar fibrosis. (b) Keratin labeling in alveolar epithelial cells of alveolar walls and in regenerating epithelial cells overlaying a fibrotic lesion (arrowheads). (c) Vimentin labeling in stromal spindle cells (arrowheads) present within a fibrotic lesion. (d) Type IV collagen labeling in the basement membrane confirms that fibrotic lesions are formed within the intra-alveolar space. (e) Double-labeled immunofluorescent image showing the presence of epimorphin (green) within an intra-alveolar fibrosis, and the presence of keratin (red) within epithelial cells partially overlaying (arrowheads) the lesion. ( f ) The same view of panel e using Nomarski optics clearly shows the intra-alevolar fibrotic lesions with three-dimensional effect and matching the colors of panel e. (g, h, and i) Double-labeled immunofluorescent images showing the presence of epimorphin (green) within vimentin (red)-positive mesenchymal cells (yellow) as well as in the surrounding extracellular matrix. Nuclei are shown as DAPI-positive (blue). Asterisks denote intra-alveolar fibrosis. All bars = 18 µm.

By Day 28, some alveoli exhibited dense fibrotic remodeling, but inflammatory cells were still observed. Strong epimorphin immunoreactivity was observed in the fibrotic areas (Figure 4a), whereas immunolabeling of keratin revealed pronounced re-epithelialization of the fibrotic lesions (Figure 4b). This was confirmed by images of double-labeled samples, which clearly showed that the epimorphin-positive fibrotic lesions were now almost completely overlaid with regenerating epithelial cells (Figure 4c). The same view of confocal images using Nomarski optics clearly showed the intra-alveolar fibrotic lesions by three-dimensional shapes and matching the colors of confocal images (Figure 4d).

View larger version (110K): [in this window] [in a new window]   Figure 4.   Immunohistochemical analysis of the lungs harvested on Day 28 after bleomycin treatment. (a) Intense epimorphin immunoreactivity is observed in areas of fibrosis. (b) Keratin labeling is seen in regenerating alveolar epithelial cells surrounding the fibrotic lesions. (c) Double-labeled immunofluorescent image clearly showing the presence of epimorphin within a fibrotic lesion (green) as well as in surrounding keratin (red)-positive regenerating epithelial cells (arrowheads). (d) The same view of panel c using Nomarski optics clearly shows the intra-alveolar fibrotic lesions with three-dimensional shapes and matching the colors of panel c. Nuclei are shown as DAPI-positive (blue). Asterisks denote intra-alveolar fibrosis. All bars = 18 µm.

Final observations were made on Day 56 after bleomycin treatment. At that time, some of the alveoli exhibiting fibrotic remodeling also showed occasional pleural indentations and focal, emphysema-like dilatation (Figure 5a). There was less epimorphin immunoreactivity in the fibrotic areas than in previous stages (Figure 5b). Other paired serial sections showed the presence of vimentin-positive cells within areas of dense fibrosis (Figure 5d) where decreased expression of epimorphin was also demonstrated (Figure 5c).

View larger version (113K): [in this window] [in a new window]   Figure 5.   Immunohistochemical analysis of the lungs harvested on Day 56 after bleomycin treatment. (a) EMG-stained specimen showing fibrotic remodeling (asterisk) of some alveoli. (b) Serial section of panel a shows epimorphin immunoreactivity is now almost absent from the fibrotic area but observed weakly in the alveolar walls around the lesion (asterisk). (c and d) Paired serial sections showing decreased expression of epimorphin (c) within areas of dense fibrosis where vimentin-positive cells are present (d). All bars = 50 µm.

In Situ Hybridization

In situ hybridization using an antisense probe for epimorphin showed that in the lung tissue samples harvested on Day 28 after bleomycin treatment, epimorphin mRNA was being expressed in the mesenchymal cells present in areas of active fibrosis (Figure 6a). What's more, the pattern of distribution of epimorphin mRNA matched that of the immunohistochemical labeling for epimorphin protein (Figure 6c). No positive reaction was detected when hybridization was carried out using the sense probe (Figure 6b).

View larger version (107K): [in this window] [in a new window]   Figure 6.   In situ hybridization of epimorphin mRNA from a lung harvested on Day 28 after bleomycin treatment. (a) In situ hybridization using an antisense probe reveals the presence of epimorphin mRNA within the mesenchymal cells located in areas of early fibrosis (arrowheads). (b) In situ hybridization using a sense probe for epimorphin mRNA shows no reaction. (c) The pattern of distribution of epimorphin mRNA matches the pattern of immunoreactivity for epimorphin protein. Panels a, b, and c depict serially sectioned specimens. All bars = 40 µm.

Immunoelectron Microscopy

Immunoelectron microscopic analysis of the lung tissue harvested on Day 28 after bleomycin treatment revealed epimorphin to be localized within the endoplasmic reticulum of mesenchymal cells. It also revealed epimorphin to be bound to the basement membrane and collagen fibrils present within the fibrotic lesions (Figure 7).

View larger version (169K): [in this window] [in a new window]   Figure 7.   Immunoelectron micrograph showing the distribution of epimorphin in a lung harvested on Day 28 after bleomycin treatment. Epimorphin immunoreactivity is seen in the endoplasmic reticulum (arrowheads) of the mesenchymal cells, on the basement membrane (arrow), and on collagen fibrils (C) within the fibrotic area. Bar = 1 µm.

Northern Blot Analysis

Expression of a 3.2-kb epimorphin mRNA was detected in normal adult lungs before bleomycin treatment. However, the level of expression increased markedly between Days 3 and 28 after bleomycin treatment; expression peaked around Day 28 and then gradually declined. To confirm that the blots reflected samples of equal size, they were reprobed for expression of -actin mRNA, and it was found that relative to -actin, epimorphin expression was increased nearly 1.5-fold in the lung tissue harvested on Day 28 after bleomycin treatment (Figure 8b).

View larger version (21K): [in this window] [in a new window]   Figure 8.   Representative Northern blots of epimorphin mRNA and -actin mRNA. (a) Epimorphin mRNA expression detected in normal lungs and at the indicated times after bleomycin treatment (upper blot). The intensity of the signal increases until Day 28 (d28) and then decreases gradually. The blot was reprobed for expression of -actin mRNA (lower blot). (b) Time-dependent changes of epimorphin mRNA expression obtained by BAS analysis of digitized images of the blots. Each bar represents the levels of epimorphin mRNA normalized to levels of -actin mRNA in eight mice and is shown as a percentage of the control value (Day 0).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This is the first study to examine in detail the temporal and spatial changes in the distribution and magnitude of epimorphin expression in normal adult and fibrotic lungs. We found that in normal adult lungs, epimorphin was present within bronchiolar, alveolar, and vascular walls, and that bleomycin-induced fibrosis resulted in increased expression of epimorphin mRNA up to Day 28 after treatment. Epimorphin was synthesized by mesenchymal cells and localized to the mesenchymal cells and ECM of early intra-alveolar fibrotic lesions. Later re-epithelialization of the fibrotic lesions suggests that epimorphin may have a role in the epithelial repair following bleomycin-induced injury.

The observed distribution of epimorphin protein and expression of its mRNA in normal mouse lungs were consistent with previous reports (3, 26) and suggests that epimorphin is active in healthy adult tissue, although its specific function is not yet known. It is known, however, that under physiologic conditions, alveolar epithelial cell turnover time in mouse lung ranges from 28 to 35 d (27), and daily endothelial cell turnover is estimated to be about 1% (28). Thus, in normal adult lung, epimorphin may serve as an epithelial and endothelial cell morphogen, maintaining the cell turnover necessary for normal structure and function.

Our findings clearly demonstrate expression of epimorphin in the mesenchymal cells situated within active fibrotic lesions and the presence of epimorphin in the ECM in the vicinity of those cells. Those findings were consistent with previous reports demonstrating the expression of epimorphin in mesenchymal cells in the fetal and postnatal rat intestine by in situ hybridization (29, 30) and the high affinity of epimorphin to basement membrane component in Matrigel (2, 7). Furthermore, the expression of epimorphin was found to be elevated before and during the period of active re-epithelialization of fibrotic lesions, after which epimorphin expression declined. It is well known that alveolar and bronchiolar epithelial cell regeneration are crucial components of pulmonary fibrosis and involve a process similar to fetal development (14). The time course of the expression of epimorphin in our model was similar to that seen in fetal rat intestine, where epimorphin mRNA was strongly expressed during lumen formation and villus morphogenesis, then decreased gradually but continued at a lower level even in the postnatal period (29).

A 19 amino-acid motif (NL-peptide) within the central domain of epimorphin mediates its binding to cells (31), and the recombinant epimorphin that includes the NL-peptide promotes adhesion of epithelial cells (PAM [murine keratinocyte] cell line) (7), mesenchymal cells (10T1/2 cell line) (7), and endothelial cells (human umbilical vein endothelial cells cell line) (3) to various substrates. It has been suggested that the function of epimorphin is to induce epithelial cell differentiation. However, in the early stages of lung fibrosis in our study, epimorphin expression was strongly detected in areas of active fibrotic lesions in which re-epithelialization had not yet occurred. This high level of expression continued until Day 28, by which time re-epithelialization was virtually complete. The ECM present in early fibrotic lesions is known to contain fibronectin and other adhesion molecules as ligands for regenerating alveolar epithelial cells for successful repair (13, 14, 32). Therefore, we suggest that a key role for epimorphin present in the early fibrotic lesions is to participate in the adhesion of the regenerating epithelial cells in addition to the differentiation in vivo.

Epimorphin has been shown to augment expression of gelatinase A in cultured mammary epithelial cells (2). Gelatinase A is also known to be upregulated and activated in regenerating alveolar epithelial cells in pulmonary fibrosis (35, 36). We therefore will discuss the possibility that some of the effects of epimorphin during re-epithelialization of fibrotic lesions in our model might be mediated by gelatinase A like mammary epithelial cells. Although epimorphin receptors have not yet been identified, it has been suggested that epimorphin binding to ECM molecules may be involved in establishing epithelial polarity, forming an organized basement membrane, and forming cell-ECM junctional complexes (2). Our results are consistent with the ideas that epimorphin has a high affinity for ECM molecules and could alter cellular function as adhesion and differentiation of regenerating epithelial cells by altering signaling through probable ECM receptors.

In conclusion, epimorphin may have important roles as a morphogen not only in embryonic lungs but also in adult lungs during the process of wound healing, and epimorphin may contribute to the remodeling of pulmonary fibrosis via epithelial-mesenchymal interactions.