Caveolin-1 (cav1) is a 22-kDa membrane protein essential to the formation of small invaginations in the plasma membrane, called caveolae. The cav1 gene is expressed primarily in adherent cells such as endothelial and smooth muscle cells and fibroblasts. Caveolae contain a variety of signaling receptors, and cav1 notably downregulates transforming growth factor (TGF)-β signal transduction. In pulmonary pathologies such as interstitial fibrosis or emphysema, altered mechanical properties of the lungs are often associated with abnormal ECM deposition. In this study, we examined the physiological functions and the deposition of ECM in cav1−/− mice at various ages (1–12 mo). Cav1−/− mice lack caveolae and by 3 mo of age have significant reduced lung compliance and increased elastance and airway resistance. Pulmonary extravasation of fluid, as part of the cav1−/− mouse phenotype, probably contributed to the alteration of compliance, which was compounded by a progressive increase in deposition of collagen fibrils in airways and parenchyma. We also found that the increased elastance was caused by abundant elastic fiber deposition primarily around airways in cav1−/− mice at least 3 mo old. These observed changes in the ECM composition probably also contribute to the increased airway resistance. The higher deposition of collagen and elastic fibers was associated with increased tropoelastin and col1α2 and col3α1 gene expression in lung tissues, which correlated tightly with increased TGF-β/Smad signal transduction. Our study illustrates that perturbation of cav1 function may contribute to several pulmonary pathologies as the result of the important role played by cav1, as part of the TGF-β signaling pathway, in the regulation of the pulmonary ECM.
- transforming growth factor-β
- extracellular matrix
- lung physiology
caveolin-1 (cav1) is a small 22-kDa protein essential to the formation of small omega-shaped invaginations of the plasma membrane, called caveolae. The cav1 protein is found primarily in adherent cells such as endothelial cells, smooth muscle cells (SMCs), type I pneumocytes, and fibroblasts. Within caveolae invaginations, cav1 interacts with a variety of growth factor receptors, including transforming growth factor (TGF)-β receptors. Cav1 regulates TGF-β signaling by modulating TGF-β receptor gene expression (24), preventing Smad2 phosphorylation (33), and/or mediating TGF-β receptor turnover (51). TGF-βs are multifunctional growth factors that regulate a large number of cellular activities that includes cell growth, differentiation, programmed cell death, and ECM production. Their signaling pathways are mediated by the associated TGF-β receptor complex and lead to the activation of Smad proteins or MAPKs that ultimately regulate the transcription of targeted genes. TGF-β signaling is tightly regulated at multiple levels via association with various ECM and membrane proteins including cav1 (6, 27).
The physiological importance of cav1 in lungs has recently been highlighted by the description of its dysregulation in pulmonary disorders such as interstitial fibrosis (48) and bronchiolization in lung fibrosis (29, 43) and in airway remodeling in an allergen-challenged mouse model (23). In these pathologies, decreased expression of the gene encoding cav1 (cav1) was associated with enhanced TGF-β signaling and increased collagen deposition. In addition, the lack of cav1 in knockout mice led to the development of a wide range of pulmonary abnormalities such as constricted alveolar spaces, thickening of the alveolar wall, and hypercellularity (22), which would likely impact lung mechanics and physiological functions. Indeed, pathologies marked by altered lung mechanical properties are often associated with altered levels of TGF-β and abnormal ECM deposition (3, 47).
Consequently, to further explore the role that cav1 plays in long term ECM remodeling, we quantified the deposition of selected ECM proteins and evaluated pulmonary mechanical properties and Smad2-related signaling in cav1 knockout mice (cav1−/−).
MATERIALS AND METHODS
Wild-type C57BL/6J (WT) and Cavtm1Mls/J (cav1−/−) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). The genetic background for the cav1−/− mice is a mixture of C57BL/6J and 129S6/SvEv. The resulting chimeric animals were then backcrossed into C57BL/6J background six times. As previously reported by The Jackson Laboratories (http://www.jax.org), the cav1−/− genotype in a predominantly C57BL/6J genetic background leads to lower fertility. Breeding was indeed less effective compared with WT mice, but the lower fertility of cav1−/− mice was addressed with a higher number of breeding pairs. In recent years, more consideration is given to the influence of genetic background of inbred strains on the development of any given phenotype (25). However, to our knowledge, there is little evidence from the literature that the 129S6/SvEv background presents significant difference in lung and vascular physiology, fibrosis development, or other phenotype with C57BL/6J (1, 15, 17, 35, 49). Euthanasia was performed with gradual filling of a chamber with CO2 on all mice used in this study. Bleeding from the nose, gross or histological evidence of CO2-induced pulmonary hemorrhage, was not observed under these conditions. Animal experiments were conducted using a protocol approved by the Institutional Animal Care and Use Committee of the University of Hawaii.
Measurement of lung mechanics.
Dynamic compliance and elastance were measured using a flexiVent instrument (SCIREQ, Montréal, Québec, Canada; http://www.scireq.com/products/flexivent) as previously described (9, 10) according to the manufacturer's instructions. Briefly, mice were weighed before anesthesia with ketamine (100 mg/kg ip) and xylazine (10 mg/kg ip). Tracheae were exposed surgically, cannulated, and attached to the flexiVent for ventilation. The positive end-expiratory pressure (PEEP) was set at 2 cmH2O. The mice were mechanically ventilated at 150 breaths/min and a tidal volume of 7.5 ml/kg. Resistance, compliance, and elastance were measured 15 times over 45 s after an intraperitoneal injection of pancuronium bromide. The data were analyzed using the SCIREQ single compartment model. The measurements taken were categorized as forced maneuvers. After measurements, mice were disconnected from the ventilator, killed by CO2 inhalation, and exsanguinated by cardiac puncture. Bronchoalveolar lavage (BAL) was performed. The right lung was then excised for histology and electron microscopy (EM), whereas the left lung was reserved for RNA and protein extraction. The weight of total lungs was recorded.
Histology and collagen deposition.
Noninflated lung tissues were fixed in a solution of PBS containing 4% paraformaldehyde. Lungs were embedded in paraffin, sectioned, and stained with trichrome or picrosirius red stains. All samples were processed as in a single batch under identical conditions. For evaluation of active collagen deposition, half of the upper left lung was frozen, and collagen content was calculated relative to the total mass of the tissue segment using a Sircol Collagen Assay kit (Biocolor, Westbury, NY) according to the manufacturer's instructions. Briefly, 50 mg of lung tissue was homogenized in 500 μl of extraction buffer and incubated for 12 h at 4°C with stirring. Tissue homogenates were spun at 15,000 g for 60 min at 4°C. Aliquots of lung homogenate were then assayed for total lung collagen levels by comparison with a standard curve of collagen standard solutions included in the kit using optical density measurement read at 540 nm with a spectrophotometer. Collagen deposition was also determined by scoring trichrome-stained slides by investigators blinded to the coding of the samples. Two grading scales were established before the observation of the lung sections under ×200 power. The grading scale of peribronchial trichrome stain was: 1, marginal peribronchial and perivascular blue trichrome stain; 2, slight increase in peribronchial and/or perivascular trichrome stain; 3, increase in trichrome stain in all airways; 4, dramatically increased stain in all airways. The grading scale for parenchymal collagen deposition was: 0, no to marginal stain; 1, slight and consistent increase; 2, increased, uniform stain throughout the parenchyma; 3, dramatic increase. The final score for each sample reflected the grading using both scales. Note that although SMCs displayed a shade of blue with the trichrome stain method, these cells were easily discernable from collagen deposition and thus did not affect the final trichrome scores. To quantify collagen networks in lung parenchyma and around airways of WT and cav1−/− mice, the level of picrosirius red staining was assessed with ImagePro software and expressed as a percentage of the area of the image analyzed as described before (2).
Gene and protein expression.
Total RNA was extracted from cell or tissue samples using the RNeasy kit (QIAGEN, Valencia, CA). RNA samples were treated with 0.05 U/ml DNase I (QIAGEN) at 20°C for 15 min. Total RNA (5 μg) was converted into first-strand cDNA using random hexamers (SuperScript First-Strand Synthesis System for RT-PCR; Invitrogen, Carlsbad, CA). The level of expression of α2-chain of type I collagen (col1α2) and tropoelastin (eln) mRNA was detected by quantitative RT-PCR using commercially available TaqMan probes (Applied Biosystems, Foster City, CA). Total protein was extracted by homogenizing 0.5 g of frozen lung tissues on ice in 10 ml of Cell Lytic MT buffer (Sigma St. Louis MO) containing 1 mM DTT, 1× protease inhibitor cocktail (Calbiochem, San Diego, CA), and 5 mM EDTA. Homogenates were centrifuged at 16,000 g, and the protein concentration in the supernatant was determined by the Bradford assay (Bradford reagent; Bio-Rad, Hercules, CA). Total protein (60 μg) was combined with reducing Laemmli buffer, heated at 95°C for 10 min, cooled on ice, and loaded into wells of a 10% polyacrylamide gel (Bio-Rad). Protein was transferred to nitrocellulose membrane and blotted with primary antibodies including anti-Smad2 (cat. no. 51-1300; Zymed Laboratories, San Francisco, CA), anti-phosphorylated Smad2 (anti-pSmad2; cat. no. 3104; Cell Signaling Technology, Danvers, MA), anti-ERK1/2 (cat. no. 9102; Cell Signaling Technology), and anti-pERK1/2 (cat. no. 9101; Cell Signaling Technology). Appropriate ECL-peroxidase-linked secondary antibodies were detected using ECL Plus (Amersham, Piscataway, NJ). For densitometry, digital images of autoradiographic film were captured using Gel Logic 200 and Kodak MI software (Kodak Scientific Imaging Systems, Rochester, NY). The net intensity of the target bands was normalized to that of the β-actin band to obtain a relative level of proteins of interest on an unsaturated exposure.
Lung vascular leakage.
To evaluate extravasation of proteins in pulmonary tissues of cav1−/− and control mice, 1 ml of PBS was injected in the trachea of anesthetized mice. The BAL fluid was aspirated from the lungs and centrifuged at 1,000 g at 20°C for 10 min. The resulting supernatant was stored at −80°C until it was tested for protein content and albumin levels. Total protein content was measured by standard Bradford assay (Bradford reagent; Bio-Rad) as described by the manufacturer's instructions. Albumin levels were measured in BAL fluid diluted to 1:10,000 using Mouse Albumin ELISA Quantitation Kit (Bethyl Laboratories, Montgomery, TX) according to instructions. Lung permeability was also evaluated using Evans blue injection. Three-month-old mice were anesthetized by intraperitoneal injection of ketamine (50 mg/kg) and xylazine (10 mg/kg). Evans blue (20 mg/kg) was injected into the tail vein at a concentration of 5 mg/ml. Injected volumes ranged from 50 to 150 μl depending on the weight of the mouse. One hour after injection, mice were killed. The chest cavity was then opened, and the mice were perfused with 5 ml of PBS through the left ventricle. Whole lungs were removed, and measurements of the amount of Evans blue that diffused into the lung tissue were performed as previously described (46). Briefly, lungs were homogenized in PBS and 15 volumes of 10% formamide and incubated at 60°C for 12–18 h. After centrifugation at 5,000 g for 30 min, the supernatant was collected, and absorbance was measured at 620 and 720 nm. Evans blue content (micrograms of Evans blue per gram of lung) was determined after adjusting values at 620 nm for the presence of heme pigments as follows: corrected absorbance at 620 nm = A620 − (1.426 × A720 + 0.030). Results were compared with a standard curve of Evans blue in formamide/PBS. Finally, the wet weight of whole lungs from separate groups of eight cav1−/− and eight WT mice was measured before the lungs were placed in an oven at 110°C for 4 h. The weight of the dried lung was then recorded and normalized to animal weight.
Cellular content and cytokine levels in BAL.
After an intratracheal injection of 1 ml of PBS, BAL fluid was collected from each mouse and centrifuged at 1,000 g at 20°C for 10 min. The cell pellet was resuspended in 100 μl of PBS to determine cell counts. Cytospin preparations were made with a maximum of 2.5 × 105 cells per slide using a Shandon Cytospin centrifuge (Shandon/Lipshaw, Pittsburgh, PA). Total cell count was performed using a standard hemocytometer slide. Cytokines were measured in BAL fluid using Mouse Th1/Th2 Cytokine Kit Cytometric Bead Array (BD Biosciences Pharmingen) according to the manufacturer's instructions (n = 5–10).
Characterization of the elastic fiber deposition.
To measure desmosine, lung samples were subjected to hydrolysis in 6 N HCl at 100°C for 24 h. The hydrolysates were evaporated and redissolved in 100 μl of water. Desmosine was determined by radioimmunoassay and normalized to the total protein in the hydrolysate as described by Starcher et al. (39, 40). For histology and immunohistochemistry procedures, paraffin-embedded formalin-fixed lung samples were cut in 5-μm sections. The sections were then deparaffinized and rehydrated. Hart's staining was performed by immersing sections in a solution of working concentration Resorcin-Fuchsin solution (Poly Scientific, Bay Shore, NY) overnight. After washing in water, sections were counterstained with tartrazine, dehydrated in ethanol, cleared in xylene, and mounted. For immunocytochemistry, sections were blocked with 5% normal goat serum in PBS for 1 h. Sections were subsequently incubated with elastin primary antibody (PR387; Elastin Product, Owensville, MO) at 4°C overnight. After several washes in PBS and 0.05% Tween-20, sections were incubated with a secondary antibody conjugated with an Alexa Fluor 568 dye. Nuclei were counterstained using 4,6-diamidino-2-phenylindole (DAPI). Sections were mounted and visualized with an Axioskop 2+ fluorescent microscope (Zeiss). Images of immunohistochemistry and histology were processed using Adobe Photoshop. For the morphometric analysis of elastin deposition (Fig. 5), ImageJ software (http://rsb.info.nih.gov/ij/) was used following a previously described method (12). For EM, 1.5-mm3 pieces of lung were fixed in 3% glutaraldehyde in 0.1 M sodium cacodylate, stained sequentially en bloc in OsO4, tannic acid, and uranyl acetate, and then dehydrated and embedded in Epon (4). Thin sections (60 nm) were placed on Formvar-coated grids and counterstained with 7% methanolic uranyl acetate followed by lead citrate. Sections were viewed using a Tecnai 12 transmission electron microscope at 120 kV, and images were digitally captured.
The results were expressed as means ± SD. Student's t-test or ANOVA were used to determine statistical significance between compared groups, as applicable. A P value <0.05 was considered significant. The data entry, management, and statistical analysis were performed using Prism software (GraphPad Software, San Diego, CA).
Cav1−/− mice present altered lung mechanics.
Lung compliance, elastance, and airway resistance in cav1−/− and WT mice were evaluated using a flexiVent instrument at five time points between 1 and 12 mo of age at baseline with no other stimulation. We observed that from 3 mo on, cav1−/− mice presented significantly decreased lung compliance compared with WT mice (Fig. 1A). Conversely, we found a sustained increase in lung elastance that was statistically significant at 3 mo and older ages (Fig. 1B). Finally, airway resistance was evaluated at the same time points. The results were similar, showing that from 3 mo on, cav1−/− animals had a significant increased airway resistance (Fig. 1C). These data were suggestive of a stiffening of lung tissue, including the airways and increased pulmonary recoil that is likely to result from increased collagen deposition, increased elastic fiber content, and perhaps lung fluid accumulation.
Increased collagen deposition in cav1−/− mice pulmonary tissues.
There were two probable causes for the observed decreased in lung compliance. The increased stiffness of the lungs could have resulted from the accumulation of fibrous tissue in the lungs and/or the buildup of fluid in the alveolar spaces (10). Indeed, we found that an increase in airway resistance in cav1−/− mice derived from the thickening of the subepithelium basement membrane caused by an overabundance of collagen deposition (unpublished observations). To determine whether a similar increased collagen deposition might be contributing to the progressive decrease in lung compliance in cav1−/− mice, we first performed histological analysis using trichrome staining on paraffin sections of lung tissues at 1, 3, 6, and 12 mo of age. A representative image of airways in 6-mo-old mice is shown in Fig. 2A. The cumulative collagen deposition in airways and parenchyma was scored and compared between age-matched cav1−/− and WT lung sections (Fig. 2B). The data showed a greater level of collagen accumulation in the lung parenchyma and also at the periphery of airways in cav1−/− than in WT mice. Cav1−/− mice presented significantly higher scores at all ages despite raised WT values at 12 mo of age. To further evaluate collagen deposition, we used picrosirius red as an alternate collagen stain. Figure 2C illustrates the higher levels of picrosirius red staining in the parenchyma of cav1−/− mice lungs. We quantified picrosirius red staining in airways and parenchyma separately (Fig. 2, D and E). In the parenchyma, we observed no significance difference at 1 and 3 mo of ages between cav1−/− and WT mice. However, between 3 and 6 mo, picrosirius red levels rose sharply to become significantly elevated at 12 mo (Fig. 2D). The results obtained with picrosirius red quantification in airways presented a similar profile (Fig. 2E), which suggested a relatively uniform and late rise in collagen deposition cav1−/− mice lungs. Because an overall increase in collagen synthesis may contribute to this accumulation of fibrotic material in the lungs of cav1−/− mice, we quantified newly deposited and soluble collagen using the Sircol collagen assay. This method also showed elevated level of collagen synthesis at all ages (1–12 mo) (Fig. 2F).
Because all three collagen-specific staining methods indicated active collagen deposition, we speculated that this accumulation might derive from increased gene expression. We quantified col1α2 and col3α1 expression, as these genes encode the predominant type of collagen fibrils in pulmonary tissues. In whole lung RNA extract, we found increased col1α2 mRNA levels between 3 and 6 mo of age in cav1−/− mice compared with WT lung (Fig. 3A). Similarly, we found that col3α1 was expressed at higher levels in the lungs of cav1−/− mice but only peaked above WT levels at 3 mo (Fig. 3B). These expression data suggested that elevated collagen deposition resulted, at least in part, from the upregulation of collagen type I and III gene expression.
Extravasation of albumin and Evans blue in the lungs of cav1−/−.
Because the loss of cav1 and caveolae structures results in enhanced vascular permeability (5, 28, 32), we evaluated the possibility that vascular leakage could lead to the accumulation of fluid in lung tissues in cav1−/− mice over time, which, in turn, may contribute to the altered compliance we observed. To evaluate excess pulmonary fluid extravasation in cav1−/− mice, we quantitatively measured albumin levels in BAL and Evans blue accumulation in the lung parenchyma. We also determined cell quantity and cytokines in BAL. As previously shown, we did not observed significant differences in the lung wet-to-dry weight ratios (13, 14) or total body weight between age-matched cav1−/− and WT mice (data not shown). Cellular content and cytokines levels in BAL were similar between the two strains of aging mice (Tables 1 and 2), indicating that no inflammation was concurrent to the other phenotypic changes of cav1−/− mice. In contrast, an accumulation of proteins (data not shown), especially albumin, was observed in BAL of cav1−/− mice from 1 to 9 mo of age (Fig. 4A). By tracking Evans blue bound to albumin in the blood stream, we determined the level of diffusion of albumin into pulmonary tissues after injecting the dye in the tail vein of 3-mo-old cav1−/− mice (albumin leakage peaks at 3 mo). Significant amounts of Evans blue dye were found in lung tissues of cav1−/− mice indicating the likely circulatory origin of albumin found in BAL (Fig. 4B). These results confirmed the increased permeability of the pulmonary endothelial barrier to plasma proteins that was previously reported (28, 36) but also indicate that vascular leakage contributes to the presence of pulmonary fluid in cav1−/− mice.
Tropoelastin gene expression and elastic fiber deposition.
The increased lung elastance that we observed in cav1−/− mice was indicative of greater pulmonary elastic recoil during passive expiration. In elastic tissues, this indispensable mechanical property is conferred by elastic fibers for which the major component is insoluble elastin. Proper elastic recoil depends primarily on the correct expression of the tropoelastin gene (eln), the formation of lysine-dependent cross-links, and appropriate assembly with other essential elastic fiber proteins in airways and alveolar walls (18). Therefore, we first examined the developmental profile of eln expression. In normal developing tissues that contain elastic fibers, maximal eln expression occurs during the 2nd wk of life, dramatically decreases toward the end of the 1st mo, and remains relatively low thereafter (45). Thus we included this early postnatal period into the developmental profile of eln expression. In the lungs of cav1−/− mice, the eln expression profile was overall similar to WT lungs but with two notable distinctions. The peak of expression was shifted to a perinatal period (2 days of age) instead of during the 2nd wk of postnatal growth for WT mice (Fig. 5A). The largest difference in expression was observed at 2 wk of age. Thereafter, the pattern of expression in cav1−/− mice followed that of the WT mice, but the level was significantly higher. To determine whether increased eln expression correlated with more abundant deposition of elastin, we determined the levels of desmosine cross-links (a predominant lysine cross-link specific to mature elastin) as an index directly proportional to the amount of deposited elastic fiber of the lungs. The results showed an increasing trend starting with low desmosine levels early in life followed by a slow and steady rise with age (Fig. 5B). The level of desmosine in cav1−/− mice increased to levels significantly higher than control animals at 3 mo of age, suggesting a continuous synthesis and accumulation of elastic fibers. Because the measurements of gene expression and elastin deposition gave no indication as to the spatial distribution and ultrastructural quality of the elastic fibers, we used histological and immunological stains to determine whether the elastic fibers in the cav1−/− lungs presented qualitative alterations. Localization of elastin notably revealed more elastic fibers in proximity and around airways and arteries with no apparent increase in the parenchyma or alveolar walls (Fig. 5C). Morphometric analysis and statistical evaluation of immunodetected elastin on paraffin-embedded sections of 6-mo-old cav1−/− mice showed an increase of 60% of elastin deposits (Fig. 5D), which was comparable with the increase in desmosine (+40%) at the same age (Fig. 4B).
To better evaluate the ultrastructural quality of these overabundant elastic fibers, both light and EM observations were performed. Using Hart's stain specific for elastin, we found thicker layers of elastic fibers primarily around airways and arteries (Fig. 6A). Analysis by EM showed that elastic fibers in cav1−/− mouse lungs were ultrastructurally normal in appearance and distribution (Fig. 6B) and well-integrated with the elastic fiber-related microfibrils indicative of a proper assembly process (Fig. 6C).
TGF-β signaling is enhanced in cav1−/− mice.
One of the main consequences of the lack of caveolae in cav1−/− mice is an increase in the signaling mediated by TGF-β (23, 33), which is a major regulator of the transcriptional activity of ECM genes. Because we observed an upregulation of the expression of three essential ECM genes (col1α2, col3α1, and eln) in the lung of cav1−/− mice, we hypothesized that Smad2 as the main downstream effector of TGF-β signaling is activated (phosphorylated) in accordance. We performed Western blot assays with lung tissues, and indeed we found that the ratio of pSmad2 over total Smad2 (pSmad2/Smad2) was significantly higher (P = 0.01) between 3 and 6 mo in cav1−/− mice compared with age-matched WT animals (Fig. 7A). Interestingly, the profile of Smad2 activation mirrors quite accurately col1α2, col3α1, and eln expression. In addition, we investigated similar ratios of phosphorylated vs. total ERK1 and ERK2 as these signaling proteins have been reported to play a role in TGF-β1-induced ECM production in pulmonary fibrosis (48), are known to localize in caveolae, and directly interact with cav1. These ratios were not elevated at 3 mo, whereas a moderate increase was seen at 6 mo of age (Fig. 7B).
Life span and lung pathology.
The life span of cav1−/− mice was previously studied and was dramatically reduced compared with WT mice (30). Although our studies were not designed to replicate these results, we noted that our mouse colony suffered no unexpected loss of cav1−/− mice over the first 12 mo of life. Pathological examination of cav1−/− animals revealed no significant emphysema, whereas mild perivascular lymphoid infiltrates were found in old mice (>9 mo).
Cav1 is a small membrane protein with a hairpin-loop conformation that exposes both NH2 and COOH termini to the cytoplasm. Cav1 provides a scaffolding structure essential for the formation of caveolae, which are small membrane invaginations containing a variety of signaling receptors. Cav1 is involved in the regulation of many signaling cascade, and among them we focused our attention on TGF-β signaling because of its critical role in ECM deposition, pulmonary development, and related pathologies (23). In this study, we specifically examined the relationship between pulmonary function, TGF-β signaling, and the deposition of selected ECM proteins in cav1−/− mice over the first 12 mo of life to determine the mechanism by which cav1 influences lung physiology and structure.
During normal aging, the pulmonary function of WT mice evolve toward increased compliance, reduced elastance, and airway resistance. As we characterized these functions of cav1−/− mice and compared the results with age-matched control mice, we found similar but exaggerated profiles. Compliance, elastance, and airway resistance in cav1−/− mice diverged from WT animals relatively early in adulthood (<3 mo) and reached a plateau at 6 mo. These altered lung functions of cav1−/− mice are probably the reason behind the previously reported exercise intolerance of these mice (32). Although these age-related pulmonary characteristics of cav1−/− mice are consistent with other studies (5, 30) we found one notable difference in that we did not observe an increase in mortality rate (30).
The absence or decrease in cav1 levels results in an increase in TGF-β signaling (23, 33). Several studies have demonstrated that the addition of exogenous TGF-β to in vitro cultures of dermal or lung fibroblasts leads to an increase in the abundance of several ECM gene transcripts (16, 21, 44, 50). Because of the detrimental changes in lungs mechanics that we observed in cav1-deficient mice, we investigated whether the changes in TGF-β signaling also affected the deposition of lung ECM proteins. We specifically focused on collagen type I and III and elastic fibers because these fibers play a direct role in determining lung compliance in high and low air pressure, respectively (26). Indeed, compliance is negatively affected by an increased deposition of collagen fibrils in the parenchyma but also by smooth muscle constriction and/or peripheral airway inhomogeneities. An increase in elastic fiber deposition will also contribute to a reduced compliance and to greater elastic recoil (elastance) (10). The situation is reversed in decorin-deficient mice, which show improper collagen deposition. In these animals, lung compliance was increased, but tissue elastance was not altered since elastic fiber deposition was unchanged (9). Furthermore, heterozygous eln mice (eln+/−) displayed increased pulmonary compliance and reduced elastance, which is consistent, and inversely correlates, with our data (38).
Using different but complementary staining methods specific for collagens with associated quantifications, we found a continuous accumulation of newly deposited collagen at all ages in whole lungs. Both methods used (Sircol and trichrome) showed a positive tendency toward the accumulation of collagen that appeared more pronounced at older ages (>6 mo). When collagen in airways and parenchyma was separately quantified using picrosirius red, we found marked accumulation toward older ages (12 mo) that most likely contributed to changes in the lung mechanical properties at and around 12 mo. To determine the origin of the net increase in pulmonary collagen, we examined the expression of col1α2 and col3α1. The transcription of these genes was significantly upregulated between 3 and 6 mo of age, whereas the expression at 12 mo was equivalent to WT levels. This spike in col1α2 and col3α1 expression has certainly contributed to elevated collagen deposition but only during a few months interval, 3–6 mo of age. It is possible that other factors that control the turnover rate of collagens, such as matrix metalloproteinases and associated inhibitors, could have played a preponderant role in collagen increases in cav1−/− mice at later time points (between 6 and 12 mo of age) as seen with picrosirius red and trichrome stains. However, one should note the remarkable correlation between the peak of col1α2 and col3α1 expression at 3 and 6 mo (the latter only for col1α2) and the spike of Smad2 phosphorylation at the same ages. This indicated clearly that collagen accumulation in the lungs of cav1−/− mice is in part due to a significant rise in Smad2 signal transduction resulting from the lack of caveolae-dependent inhibition of TGF-β. There is little doubt that the remodeling of lung ECM we characterized contributed to the previously described thickening of the alveolar wall (5) and played an important role in altering lung functions of cav1−/− mice.
The altered lung compliance in cav1−/− mice might also be explained by a cumulative effect and persistent increased synthesis of ECM and pulmonary vascular permeability. In our study, we observed abnormally elevated levels of albumin in the BAL of cav1−/− mice and leakage of Evans blue in pulmonary tissues because of increased endothelium permeability in these animals (5, 28, 32). In capillary endothelial cells, 30% of the cell surface area is occupied by caveolae (11). As a consequence, cav1−/− mice displayed elevated pulmonary vascular leakage, which was particularly elevated around the same 3–6 mo period when TGF-β/Smad2 signaling is maximal. It is unclear why the wet-to-dry weight ratio we measured in cav1−/− mice is similar to WT animals, but one possible explanation may be that the dry lung weight augments proportionally as a result of hypercellularity (22) and increased lung ECM. Here, we speculate that two biological pathways could possibly be responsible for the increased lung permeability we observed in our samples. Cav1 negatively regulates nitric oxide (NO) synthesis, and it was recently proposed that the increase in permeability is a NO-dependent process (8). Because TGF-β indirectly upregulates the synthesis of NO synthase (eNOS) in normal human endothelial cells (37), TGF-β has been proposed as a key regulator of pulmonary edema by directly increasing vascular permeability and also by altering cell-to-cell contacts in the endothelium (31). As cav1 negatively regulates TGF-β signaling, we suggest that the presence of vascular-derived fluid in the pulmonary tissues of the cav1−/− mice resulted from cumulative effects of increased TGF-β signaling and, potentially, from excessive NO synthesis.
Because altered pulmonary elastance in cav1−/− mice suggested increased lung elasticity, we quantified elastic fiber deposition and eln expression. The effect of TGF-β on eln expression has been described many years ago (17). It is mediated by an increase in the rate of transcription or a combination of increased transcriptional activity and mRNA stabilization. Several studies have indeed shown that eln transcripts and protein synthesis is primarily increased through mRNA stabilization in lung fibroblasts (human and murine) (20, 21, 42). A more recent work complemented these findings by showing that in addition to stabilizing eln mRNA, TGF-β induces a short burst of transcriptional activity ahead of stabilization (19). Various elements of the TGF-β downstream signaling pathway are critical for eln mRNA increase and posttranscriptional stabilization including the Smads, protein kinase Cδ, and p38 MAPK (19). As an illustration, Smad3 is a one of the early components of the TGF-β signal transduction pathway, and its deficiency in mice leads to the repression of eln expression in lung and subsequent emphysema (3). In fact, elastin levels in lungs are a critical determinant of lung physiology in normal and developmental pathologies (38). These facts correlate well with our findings in cav1−/− mice in which TGF-β-related signaling is increased. We indeed observed in adult cav1−/− mice an upregulation of eln mRNA levels and elastic fiber deposition in lung airways mirroring Smad2 activation (3–6 mo). Interestingly, we observed a temporal shift of the peak of eln expression during the early postnatal period (2–30 days). Since it was not possible to reliably determine pulmonary function in newborn and preadult mice, it is unclear whether this early change in eln expression produced any adverse effects. However, the increase in elastin synthesis and assembly into functional fibers in adult animals (1–12 mo old) certainly played a role in the increased pulmonary elastance of cav1−/− mice.
Taken together, our data showed that in pulmonary tissues of mice lacking cav1, col1α2, col3α1, and eln expression along with vascular leakage correlated with the profile of TGF-β/Smad2 signaling. It is noteworthy that the ECM gene expression and vascular leakage peaked between 3 and 6 mo of age and gradually returned to near normal values thereafter, although collagen and elastic fiber deposition continued to rise with little apparent turnover. The pSmad2 is the very likely effector that determined col1α2, col3α1, and eln expression as well as vascular leakage; however, it is unclear why TGF-β/Smad2 signaling progressively returns to WT levels over time. One possible explanation could reside with a progressive retention of TGF-β in its latent form associated with latent TGF-β binding protein complex (LTBP) in lungs as part of a structural component of the increasingly abundant elastic fibers (34, 41).
The understanding of cav1 function has clinical relevance because the pulmonary phenotype developed by cav1−/− mice (pulmonary collagen accumulation, vascular permeability, and the presence of pulmonary fluid) overlaps with characteristics of acute lung injury, a very severe lung syndrome associated with high mortality. Many research studies have addressed the molecular mechanisms regulating acute lung injury, and the role of TGF-β in the onset of this pathology is now well established (7). Because cav1 functions as a part of the TGF-β signaling pathway, it plays a significant role in lung development and integrity. This protein could therefore potentially constitute a novel therapeutic target for severe acute lung injury.
This study was supported by National Institutes of Health Grant RR-16453 and Hawaii Community Foundation Grant 20060395 to O. Le Saux, an American Lung Association of Hawaii Grant to E. K. Tam and C. Jourdan-Le Saux, National Institutes of Health Grant P20-RR-016467 to C. Jourdan-Le Saux, and Canadian Institutes of Health Research Grant MOP-57663 to E. C. Davis. E. C. Davis is a Canada Research Chair.
We thank Drs. Sheppard and Huang (University of California, San Francisco) for their assistance in setting up the lung function facility at University of Hawaii and Jeannie Mui for excellent electron microscope technical assistance.
↵* O. Le Saux and K. Teeters contributed equally to this study.
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 © 2008 the American Physiological Society