To determine potential relationships between transforming growth factor (TGF)-α and surfactant homeostasis, the metabolism, function, and composition of surfactant phospholipid and proteins were assessed in transgenic mice in which TGF-α was expressed in respiratory epithelial cells. Secretion of saturated phosphatidylcholine was decreased 40–60% by expression of TGF-α. Although SP-A, SP-B, and SP-C mRNA levels were unchanged by expression of TGF-α, SP-A and SP-B content in bronchoalveolar lavage fluid was decreased. The minimum surface tension of surfactant isolated from the transgenic mice was significantly increased. Incubation of cultured normal mice type II cells with TGF-α in vitro did not change secretion of surfactant phosphatidylcholine and SP-B, indicating that TGF-α does not directly influence surfactant secretion. Expression of a dominant negative (mutant) EGF receptor in the respiratory epithelium blocked the TGF-α-induced changes in lung morphology and surfactant secretion, indicating that EGF receptor signaling in distal epithelial cells was required for TGF-α effects on surfactant homeostasis. Because many epithelial cells were embedded in fibrotic lesions caused by TGF-α, changes in surfactant homeostasis may at least in part be influenced by tissue remodeling that results in decreased surfactant secretion. The number of nonembedded type II cells was decreased 30% when TGF-α was expressed during development and was increased threefold by TGF-α expression in adulthood, suggesting possible alteration of type II cells on surfactant metabolism in the adult lung. Abnormalities in surfactant function and decreased surfactant level in the airways may contribute to the pathophysiology induced by TGF-α in both the developing and adult lung.
- bronchopulmonary dysplasia
- acute respiratory distress syndrome
- pulmonary fibrosis
- surface activity of surfactant
transforming growth factor-α (TGF-α) is a member of the epidermal growth factor (EGF) family of polypeptides that is expressed in the developing and adult human lung airways and alveolar epithelial cells (46, 48). Increased expression of TGF-α and EGF receptor (EGFR) ligands was associated with a number of diseases characterized by lung remodeling. Levels of TGF-α in tracheal aspirates from infants with developing bronchopulmonary dysplasia (BPD) were increased compared with premature infants that did not develop BPD (46, 48). Likewise, TGF-α was increased in bronchoalveolar lavage fluid (BALF) of patients with acute respiratory distress syndrome (4, 31) and in alveolar macrophages, epithelial cells, and fibroblasts following lung injury (28–30, 51).
Remodeling of both vascular and alveolar compartments of the lung is associated with abnormal lung mechanics, including decreased compliance and increased airway resistance (8, 19, 36, 37). Increased surfactant synthesis and decreased surfactant secretion have been reported in a premature baboon model of BPD induced by ventilation (45). Saturated phosphatidylcholine (Sat PC) concentration in tracheal aspirates from infants dying of BPD was 60% lower in infants of similar age recovered from respiratory distress syndrome (7). Furthermore, surfactant protein content and surfactant function were altered in the samples from patients with BPD and lung injury (9, 14, 15, 23, 25, 37, 44, 52). Thus not only the morphological changes but also decreased surfactant secretion may contribute to altered pulmonary function in BPD. Whereas increased expression of TGF-α has been associated with BPD and other forms of lung injury, the effect of increased TGF-α on surfactant homeostasis is unknown.
To determine whether TGF-α influenced lung function, we previously generated transgenic mice in which TGF-α was expressed in epithelial cells of the peripheral lung under the control of the surfactant protein (SP)-C promoter (SP-C-TGF-α mice) (26). TGF-α caused dose-dependent lung remodeling, including disruption of postnatal alveolarization and pulmonary vascular morphogenesis (27), as well as pleural and parenchymal fibrosis (18). To determine the effects of TGF-α οn lung remodeling in the adult, TGF-α was conditionally expressed in respiratory epithelial cells under control of the Clara cell secretory protein promoter reverse tetracycline-responsive transactivator, CCSPrtTA+/−/(tetO)7TGF-α+/− mice (17). In these mice, TGF-α is induced by treating the mice with doxycycline (Dox). To understand the role of TGF-α in altered surfactant homeostasis in the remodeled lung, surfactant metabolism and function were studied in transgenic mice in which TGF-α was increased during development and adulthood.
Studies were approved by the Institutional Animal Care and Use Committee of the Cincinnati Children's Hospital Research Foundation. Mice were maintained in a barrier containment facility. All mice appeared healthy and were without evidence of infection. Sentinel mice in the colony were serologically negative for common murine pathogens.
Eight- to 10-wk-old FVB/N transgenic mice expressing human TGF-α under control of the human SP-C as promoter were studied (26). The lung content of TGF-α in line 28 SP-C-TGF-α mice used for this study was increased four- to fivefold to 2.45 ± 0.60 ng/lung of TGF-α compared with 0.58 ± 0.05 ng/lung littermate nontransgenic mice (18). In this model, TGF-α was expressed in respiratory epithelial cells throughout development.
Conditional expression of TGF-α.
Transgenic mice expressing the rtTA protein under control of the rat CCSP gene promoter (CCSPrtTA) (49) were bred with transgenic mice bearing (tetO)7TGF-α+/−. These bitransgenic mice have normal lung structure (17). In these bitransgenic mice, Dox induces the expression of TGF-α in the lung (17). In the present study, 6-wk-old mice were treated with Dox containing drinking water for 6 wk and studied at 12 wk of age. At this time, fibrosis and emphysema were observed (17). The control group consisted of 12-wk-old single transgenic CCSP-rtTA+/− littermates that were also treated in an identical manner with Dox.
The SP-C-EGFR-M transgenic mice express a truncated EGFR lacking a portion of the cytoplasmic domain in distal epithelial cells of the lung. This EGFR acts as a dominant negative inhibition of EGFR signaling. SP-C-EGFR-M transgenic mice were bred with SP-C-TGF-α mice, creating SP-C-TGF-α,SP-C-EGFR-M mice (16). TGF-α-induced fibrosis was reversed, and the extent of alveolar enlargement was significantly decreased compared with SP-C-TGF-α mice (16). Surfactant phospholipid metabolism was studied in 10-wk-old SP-C-TGF-α,EGFR-M mice to determine whether EGFR signaling in the respiratory epithelium mediated surfactant abnormalities in these mice.
Sat PC Pool Size
Mice were anesthetized with intraperitoneal pentobarbital sodium and exsanguinated by transection of the distal aorta. A 20-gauge blunt needle was tied into the trachea, the chest wall was opened, and 0.9% NaCl was flushed into the airways until the lungs were fully expanded (∼1 ml). The fluid was then withdrawn by syringe, and saline lavage was repeated five times. BALF was pooled and volume was measured. The lavaged lung tissue was homogenized in 0.9% NaCl. Aliquots of BALF and the lung homogenates were extracted with chloroform-methanol (2:1, by volume), Sat PC was isolated by OsO4 oxidation followed by column chromatography (32), and Sat PC was quantified by measuring phosphorus (3). Sat PC pool size measurements were in eight to fifteen mice for each genotype group.
Surfactant Phospholipid Precursor Incorporation Into Sat PC
Mice were given intraperitoneal injections of 10 μl/g body wt of 0.9% NaCl containing 0.5 μCi [3H]palmitic acid. Groups of seven to nine mice were killed at 8 h (and 16 h for SP-C-TGF-α mice) after injection of radiolabeled precursor. BALF was recovered from each animal, and lung tissue was homogenized in saline. Sat PC was isolated from BALF and lung homogenate as described above, and radioactivity was measured (20). The percent secretion of labeled Sat PC was calculated as the percent of radioactivity in Sat PC in BALF relative to the radioactivity in Sat PC in total lung (BALF + lung tissue).
Lungs were inflation fixed at 25 cmH2O with 4% paraformaldehyde. The volume of the lung was determined by the displacement method (41). Tissue was dehydrated through a graded series of alcohol and divided into five regions: left lobe, right cardiac lobe, right apical lobe, right diaphragmatic lobe, and accessory lobe. Each of the lobes was bisected perpendicularly to the lateral axis and embedded in paraffin. Ten paraffin sections were cut (5 μm) for each block and serially numbered. The following 20 sequential slices were discarded. This process was repeated five times to obtain five nonoverlapping regions of the lung. A random number generated from the Random Numbers Table was used to select the slide number (47) from each of the five levels for SP-C proprotein staining (53). An area measurement of each of the lung pieces was obtained using Metamorph Software (version 6.1; Universal Imaging, West Chester, PA) calibrated to 5.33 μm2 per pixel. A known area (92 mm2) generated by three random coordinate pairs was sampled in each lung piece to determine the number of alveolar type II cells (150 fields counted/mouse). Alveolar type II cells were identified by SP-C proprotein immunostaining from each field as cells possessing a cytoplasmic profile restricted to 18–72 μm2 (47). Additionally, a determination of lung parenchyma and fibrotic region in parenchyma was made on the tissue stained with hematoxylin and eosin using a 121-line test lattice grid superimposed onto each image to count the number of times the intercepts of the lines interacted with alveolar tissue. The fibrotic region recognized by light microscope was counted, and all the morphological analyses were done blindly by one experienced research assistant. Three to four mice were studied in each group. The numerical density (Nd) of type II cells at each level was determined by counting the number of cells in the given area and multiplying by the thickness (5 μm) of the tissue slice. The total number of alveolar type II cells in the lung was obtained by multiplying the averaged Nd for five levels by the volume of the lung minus a shrinkage factor of 25% due to paraffin processing (47) and multiplying by the average parenchymal tissue in the lung. The percent of type II cells embedded in fibrotic lesions relative to the total number of type II cells was determined in three fields from slices of five levels from the left lung and right lower lobe. Electron microscopy was performed on lung tissue from SP-C-TGF-α mice and nontransgenic control mice after inflation fixation at 25 cmH2O in glutaraldehyde.
Surfactant Protein and mRNA Analysis
The contents of SP-A, -B, -C, and -D in BALF were analyzed by Western blot. Samples containing 1.25 nmol of Sat PC were used for SP-A and SP-D, 0.22 nmol of Sat PC were used for analysis of SP-B, and 0.65 nmol of Sat PC were used for SP-C. Antisera for SP-B and SP-D were from Chemicon (Temecula, CA). Guinea pig anti-rat SP-A and rabbit anti-recombinant human SP-C were used for SP-A and SP-C, and immunoreactive bands were quantitated by densitometric analyses as described previously (22, 39). S1 nuclease protection assays were performed as described previously (2) using lung tissues frozen in liquid nitrogen. mRNA encoding SP-B were quantified with ribosomal protein L32 as an internal control.
Large aggregate surfactant was isolated from pooled BALF by centrifugation at 40,000 g for 15 min over 0.8 M sucrose in 0.9% NaCl cushion. Surface activity was measured with a captive bubble surfactometer (21, 42). The concentration of each sample was adjusted to 7 nmol of phospholipid/μl, and 3 μl of the sample was applied to the air-water interface of a 25-μl volume bubble by microsyringe (n = 3 pool, 2 mice/pool). Surface tension was measured every 10 s for 300 s to establish equilibrium surface tension, and then bubble pulsation was started. The minimum surface tension after 80% bubble volume reduction was measured, which did not change significantly after the third pulsation. The minimum surface tension at the fifth pulsation was reported.
Effect of TGF-α on Surfactant Secretion by Cultured Type II Cells
To determine whether TGF-α directly affects type II cell surfactant secretion, type II cells were isolated from normal FVB/N mice and cultured as previously reported (39). Combined type II cells isolated from four mice were seeded into 12 wells and cultured on Matrigel matrix in bronchial epithelial cell growth media (BEGM) containing 5% charcoal-stripped fetal bovine serum and 10 ng/ml of keratinocyte growth factor (39). The purity of type II cell preparations was >90% as assessed by modified PAP stain (10). It was previously shown that under these conditions, synthesis and secretion of surfactant by murine type II cells decline on day 1 of culture but recover by day 3, reaching levels comparable to or exceeding freshly isolated cells by day 5 (39). Effects of 5, 50, and 250 ng/well of TGF-α on [3H]PC secretion were assessed as previously described. The recombinant human TGF-α (R&D Systems, Minneapolis, MN) had low endotoxin levels (<0.1 ng/1 μg TGF-α) and was used within 3 wk after purchase. Murine cells were labeled with 1 μCi/ml of [3H]choline for 48 h before assay on 7 days in culture. Cells were washed three times to remove free labels, and TGF-α was added. Media were removed 3 h after incubation with TGF-α, and the cells were rinsed. Media samples were combined, and the cells were removed by centrifugation (130 g for 8 min). Phospholipids from the total cell and media samples were extracted with methanol and chloroform, and radioactivities were determined. The percent PC secretion was calculated as radioactivity in media divided by total radioactivity in media + cells × 100.
To evaluate whether TGF-α influences SP-B secretion, type II cells were used on day 7 of culture. Cells were washed three times with BEGM to remove extracellular surfactant, and TGF-α (0, 5, 50, and 250 ng/well) was added. Media were removed after 3 h, and cells were rinsed with fresh media. The media samples were combined, and surfactant was isolated by centrifugation as in previous studies (39). The level of SP-B in isolated surfactant from each well media was determined by Western blotting. SP-B secretion was quantitated by scanning densitometry and expressed relative to control wells lacking TGF-α.
Results are presented as means ± SE. Two group comparisons were carried out using unpaired Student's t-tests. The comparisons for in vitro cultured type II cell study was by one-way ANOVA and Tukey-Kramer multiple comparisons tests. Significance was accepted at P < 0.05.
Effects of constitutive expression of TGF-α in developing and adult lungs: SP-C-TGF-α mice.
In the SP-C-TGF-α mice, TGF-α is constitutively expressed throughout lung development to adulthood. In adult SP-C-TGF-α mice, Sat PC levels in BALF were significantly lower than in BALF from their littermate nontransgenic mice. In contrast, Sat PC content in lung tissue was increased 1.5-fold in SP-C-TGF-α mice (Fig. 1A). Secreted Sat PC (in BALF) relative to Sat PC in the total lung (BALF + tissue) was decreased by 60% in SP-C-TGF-α mice (Fig. 1B), and radiolabeled Sat PC incorporation in BALF was decreased 8 and 16 h after radiolabeled precursor injection (P < 0.05; Fig. 2, A and B). Expression of TGF-α increased radiolabeled Sat PC in lung tissue by approximately twofold (Fig. 2, A and B) and decreased [3H]Sat PC secretion into the alveolus by 60% 8 and 16 h after injection (Fig. 2, C and D).
Conditional expression of TGF-α in adult lung.
To avoid the possible effects of TGF-α during lung morphogenesis, we assessed the effects of increased TGF-α on surfactant metabolism in the adult lung in which TGF-α was expressed in a conditional manner. Surfactant Sat PC pool sizes, surfactant protein composition, and surfactant secretion were determined in CCSPrtTA+/−/(tetO)7TGF-α+/− mice in which TGF-α is conditionally expressed only when the mouse is exposed to Dox. Surfactant pool size and metabolism were determined in CCSPrtTA+/−/(tetO)7TGF-α+/− mice treated with Dox from 6–12 wk of age. Sat PC levels in BALF were decreased, whereas total content of Sat PC in the lung was increased in CCSPrtTA+/−/(tetO)7TGF-α+/− mice (Fig. 3A). As seen in SP-C-TGF-α mice, induction of TGF-α expression caused decreased secretion of Sat PC (Sat PC in BALF relative to total lung) by 40% (Fig. 3C). Eight hours after [3H]palmitic acid injection, [3H]Sat PC incorporated from [3H]palmitic acid in lung tissue was increased (Fig. 3B) and percent secretion of [3H]Sat PC was markedly decreased in TGF-α-expressing mice (Fig. 3D). Thus TGF-α expression during lung development (SP-C-TGF-α) or in adult lungs [CCSPrtTA+/−/(tetO)7TGF-α+/−] led to similar alterations in surfactant content.
Effects of TGF-α on Lung Morphology
As previously reported (17, 18, 26), pleural and peribronchioalveolar perivascular fibrosis and enlarged alveoli were observed in SP-C-TGF-α transgenic mice. Numerous type II cells, identified by pro-SP-C immunostaining, were embedded within fibrotic lesions (Fig. 4B), an observation that was confirmed by electron microscopy (Fig. 5). Although type II cell ultrastructure was preserved, the cells that were embedded within fibrotic lesions thus had limiting access to the alveoli.
Alveolar enlargement, as well as pleural and perivascular fibrosis, also was observed in CCSPrtTA+/−/(tetO)7TGF-α+/− mice after treatment with Dox from 6 to 12 wk of age. As in SP-C-TGF-α mice, type II cells were embedded in fibrotic lesions (Fig. 6B). Qualitatively, the fibrotic area within which type II cells were embedded was not as extensive as that characteristic of SP-C-TGF-α mice.
Expression of TGF-α in SP-C-TGF-α mice resulted in lower body weight (Table 1). Whereas total lung volumes were not significantly changed, lung volume relative to body weight was significantly greater in TGF-α-expressing mice. Parenchymal volume was significantly decreased in SP-C-TGF-α mice, reflecting enlarged air spaces. Whereas percent parenchyma was unchanged in adult mice with conditionally expressed TGF-α, lung volume was increased 1.5-fold, reflecting an increase in air space. Enlarged air spaces were observed in both SP-C-TGF-α mice and CCSPrtTA+/−/(tetO)7TGF-α+/− on Dox mice; however, the air spaces were not as large in the CCSPrtTA+/−/(tetO)7TGF-α+/− as in the SP-C-TGF-α mice. Although type II cell numbers for total lung and per body weight were not affected in SP-C-TGF-α mice, a threefold increase in type II cell numbers (per total lung and per body weight) was observed after 6 wk of Dox treatment in the CCSPrtTA+/−/(tetO)7TGF-α+/− mice. Eighteen percent of type II cells were embedded in the fibrotic region in SP-C-TGF-α mice (Table 2), and embedded type II cells in the CCSPrtTA+/−/(tetO)7TGF-α+/− mice were 10%. Therefore, the number of nonembedded type II cells in the total lung was decreased 30% in the SP-C-TGF-α mice and increased threefold in the CCSPrtTA+/−/(tetO)7TGF-α+/− mice compared with the number of type II cells in the control mice group. There were no fibrotic regions in the nontransgenic control mice lungs. There were more fibrotic regions in SP-C-TGF-α mice than in CCSPrtTA+/−/(tetO)7TGF-α+/− mice, resulting in a higher % of type II cells that were found embedded in matrix in SP-C-TGF-α mice.
Surfactant Proteins and Surfactant Function
The content of SP-A, -B, -C, and -D in BALF containing the same amount of Sat PC was analyzed by Western blot (Fig. 7A). From the total volume of BALF and the aliquot volume used for Western blot, surfactant protein content in total BALF was determined (Fig. 7B). Expression of TGF-α decreased content of all the surfactant proteins in BALF, whereas surfactant protein mRNAs were unchanged (Fig. 7C). Equilibrium surface tension of large aggregate surfactants adjusted to the same concentration of phospholipid were similarly low for both groups, but initial time to reach equilibrium surface tension was significantly slower (P < 0.05) for surfactant from SP-C-TGF-α mice (Fig. 8A). Large aggregate surfactant isolated from SP-C-TGF-α mice had a significantly higher minimum surface tension (Fig. 8B).
The content of SP-A and SP-B in BALF was decreased and SP-C and SP-D were unchanged in CCSPrtTA+/−/(tetO)7TGF-α+/− mice (Fig. 9A). Surfactant protein mRNAs (Fig. 9B) were unchanged in CCSPrtTA+/−/(tetO)7TGF-α+/− mice. Surface activity of isolated large aggregate surfactant from CCSPrtTA+/−/(tetO)7TGF-α+/− mice was altered, with mean minimum surface tension of 13 mN/m (Fig. 8, C and D). Altered surfactant protein content, especially decreased SP-B in SP-C-TGF-α mice and CCSPrtTA+/−/(tetO)7TGF-α+/− mice BALF, may have influenced surfactant function.
Expression of a Dominant Negative EGFR Corrects Surfactant Metabolism and Lung Morphology
As previously reported, enlarged air spaces and fibrosis seen in SP-C-TGF-α mice were substantially corrected, and type II cells shown by pro-SP-C immunostaining were distributed normally in SP-C-TGF-α/EGFR-M mice (Fig. 10D) (16). Sat PC level and % secretion in SP-C-TGF-α/EGFR-M mice were similar to nontransgenic mice (Fig. 10, A and B). Incorporation of [3H]palmitic acid into Sat PC and % [3H]Sat PC secreted in BALF at 8 h after precursor injection was similar to that of nontransgenic mice (Fig. 10C). Surfactant pool sizes and metabolism were similar to controls in the SP-C-TGF-α/EGFR-M mice lung. Thus expression of the dominant negative EGFR in respiratory epithelial cells blocked effects of TGF-α on lung morphology and surfactant homeostasis.
Effects of TGF-α on Surfactant Secretion In Vitro
Percent secretion of surfactant phospholipid from type II cells was not affected by 3-h exposure to a wide range of 5–250 ng of TGF-α/well (Table 3). Secreted SP-B from type II cells was similar (by ANOVA) after 3-h exposure to 0–250 ng of TGF-α/well. There was no direct influence of TGF-α on surfactant phospholipid and SP-B secretion by type II cells in culture.
Increased expression of TGF-α in epithelial cells of the lung decreased surfactant secretion and increased surfactant content in the tissue. Enlarged hypoplastic alveoli and fibrotic lesions seen in the TGF-α-expressing mice were similar to findings in infants and animal models with BPD (8, 19). TGF-α caused tissue remodeling with resultant embedment of type II cells within the fibrotic lesions. Abnormalities in surfactant secretion in TGF-α mice are probably related, at least in part, to tissue remodeling that limits surfactant secretion.
The presence of EGFR in rat type II cells in culture and its binding specificity were previously shown (38). Isolated type II cells from adult rabbit lung and H441 cells, a pulmonary epithelial cell line, in culture proliferate in response to TGF-α via EGFR (5). Furthermore, we have confirmed the presence of EGFR in mouse type II cells in culture (on days 5 and 7) by immunostaining using EGFR primary antibody (Upstate Biotechnology, Lake Placid, NY) and second antibody, Alexa Fluor 568 (Molecular Probes, Eugene, OR). Immunohistochemical staining of EGFR on cultured mice type II cells was strong, whereas control with only a second antibody did not show any staining (data not shown). TGF-α did not affect surfactant secretion when added directly to the media of cultured type II cells. We used 5–250 ng of TGF-α/well, and 100 ng/well was the highest amount of TGF-α used by others for the studies with cultured cells (17, 24, 40). With the use of a similar method of cultured type II cells, a twofold increase in surfactant secretion was previously shown to be induced by phorbol 12-myristate 13-acetate (39), demonstrating the utility of the cultures. The lack of a direct effect of TGF-α on surfactant secretion by type II cells suggests that altered surfactant secretion may be a consequence of TGF-α-induced alterations in lung structure in vivo. Despite high tissue levels of TGF-α in SP-C-TGF-α/EGFR-M mice, surfactant pool size and secretion were normalized by expression of a dominant negative EGFR in respiratory epithelial cells. The inhibition of TGF-α-induced signaling through the EGFR in distal epithelial cells corrected TGF-α-induced fibrosis (16) and surfactant secretion.
In the present study, the decrease in SP-B in BALF from SP-C-TGF-α mice ranged between 36 and 51% of normal mouse level, a level similar to that in SP-B(+/−) mouse lung. Heterozygous SP-B(+/−) mice with ∼50% of normal SP-B levels are known to survive without respiratory distress when housed in conditions of no stress. However, SP-B(+/−) mice are more susceptible to stress caused by hyperoxia (50). Lung inflammation induced by intratracheal injection of endotoxin was less severe in mice with higher SP-B levels than SP-B(+/−) mice (11). SP-B is the singularly required critical component to form a stable high surface-active pulmonary surfactant film (6, 35). The conditional SP-B(−/−) mice had severe respiratory distress and died when SP-B was decreased to 25% of normal levels (34). In the present study, decreased SP-B was associated with higher minimum surface tension of isolated surfactant from SP-C-TGF-α mice. Decreased levels of mature SP-B in SP-C-TGF-α mice were not due to an SP-B processing problem because pro-SP-B was not detected in BALF by Western blot (data not shown). The mechanisms by which TGF-α decreased SP-B is presently unknown. Because SP-B mRNA levels were normal, TGF-α might also perturb SP-B secretion and access to the alveoli. In SP-C-TGF-α mice, SP-A and SP-D levels in BALF were also decreased more than that for nontransgenic mice. SP-A and SP-D play roles in host defense of the lung, including clearance and killing of both bacteria and viruses and regulation of the inflammatory responses in the lung (33). Lower levels of SP-A and SP-D seen in the alveoli of the TGF-α-expressing mice might render the mice susceptible to lung inflammation. A higher incidence of lung infection and prolonged lung inflammation in patients with pulmonary fibrosis and emphysema might, at least partially, be related to changes in concentrations of surfactant proteins associated with lung remodeling.
Whereas effects of TGF-α on surfactant phospholipid secretion were similar in SP-C-TGF-α mice and CCSPrtTA+/−/(tetO)7TGF-α+/− mice, lung morphology was distinct for each model. In SP-C-TGF-α mice, lung volume was not significantly increased but percent parenchyma was significantly decreased. The larger alveolar air space was the result of a lack of secondary septation during postnatal alveologenesis. In contrast, in the CCSPrtTA+/−/(tetO)7TGF-α+/− mice, the lung volume was increased 1.5-fold. The functional area of parenchyma in CCSPrtTA+/−/(tetO)7TGF-α+/− mice was the same as control and resulted in increased alveolar air space. Numbers of type II cells were measured threefold in CCSPrtTA+/−/(tetO)7TGF-α+/− mice. In contrast, type II cell numbers in SP-C-TGF-α mice were similar to nontransgenic mice. A higher percentage of parenchyma was fibrotic, and more type II cells were embedded in fibrotic lesions in SP-C-TGF-α mice compared with CCSPrtTA+/−/(tetO)7TGF-α+/− mice. The degree of decrease in surfactant protein contents in BALF was also higher in SP-C-TGF-α mice than CCSPrtTA+/−/(tetO)7TGF-α+/− mice. These differences in surfactant protein contents are associated with higher surface tension of isolated surfactant from SP-C-TGF-α mice compared with that from CCSPrtTA+/−/(tetO)7TGF-α+/− mice. These differences in the two models are likely related to the time and/or duration of TGF-α expression. Lung remodeling, decreased surfactant proteins in BALF, and altered surface activity of surfactant were more severe when TGF-α was expressed throughout lung development compared with 6 wk in adults.
In patients with lung injury, extensive damage to the alveolar epithelium and vascular endothelial cells can cause increased protein permeability with resultant extravasation of protein-rich edema fluid into the air spaces that can inhibit surfactant activity and alter pulmonary function. Increased inflammatory cells synthesize and secrete various cytokines, including IL-6 and IL-1β, which are associated with increased EGF and TGF-α. IL-1β induced alveolar epithelial wound repair by EGF- or TGF-α-dependent mechanisms in vitro (13, 24), demonstrating a role of TGF-α in repair of lung injury. The various cytokines and growth factors, including TGF-α, activate STAT-3, which mediates a wide variety of cellular and organ responses to inflammation (1). Activated STAT-3 is also known to increase expression of TGF-α in many organs and malignant cells (12, 43), whereas the signaling mechanisms for induction of TGF-α in injured lung are unknown.
The present study demonstrates that expression of TGF-α in the developing and adult lung altered surfactant phospholipid and SP-B secretion in vivo. Abnormalities in surfactant homeostasis in clinical conditions, such as BPD and lung recovery from acute respiratory distress syndrome, may be related in part to TGF-α-induced tissue remodeling that limits surfactant secretion. Although multiple factors are likely to play roles in the pathogenesis of BPD and other forms of chronic lung injury, the present study supports the concept that TGF-α may play a role in altered surfactant protein homeostasis and surfactant function that are often associated with lung injury and remodeling.
This work was supported by National Heart, Lung, and Blood Institute Grants HL-56387 (J. A. Whitsett, T. R. Korfhagen, M. Ikegami), HL-63329 (M. Ikegami, J. A. Whitsett, T. R. Korfhagen), HL-72894 (T. D. Le Cras), HL-58795 (T. R. Korfhagen), and K08-HL-04172 (W. D. Hardie) and American Lung Association Career Investigator Award CI-31-N (T. D. Le Cras).
We thank Dr. Kent E. Pinkerton (Univ. of California, Davis) and Prithy C. Martis (Univ. of Cincinnati) for suggestions on morphological analysis.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society