Caveolin-1 Deficiency Protects from Pulmonary Fibrosis by Modulating Epithelial Cell Senescence in Mice

Animals and Induction of Pulmonary Fibrosis

Animal experiments were conducted using a protocol approved by the Institutional Animal Care and Use Committee of the University of Hawaii. WT C57BL/6J (WT) and Cavtm1Mls/J (cav-1^−/−) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). The genetic background for the cav-1^−/− mice is a mixture of C57BL/6J and 129S6/SvEv. The resulting chimeric animals were then back-crossed into C57BL/6J background no more than six times due to a lower fertility. To induce pulmonary fibrosis, WT and cav-1^−/− mice were administered with 1.3 U/kg body weight of bleomycin intratracheally. Control animals from both cohorts received PBS intratracheally. Animals were killed at the indicated time points. Lungs and blood samples were collected. Dynamic compliance of the lungs was analyzed as previously described by Jourdan-Le Saux and colleagues (19) as a measure of lung mechanics using a Flexivent instrument (Scireq Inc., Montreal, Quebec, Canada) (21, 22).

Assessment of Pulmonary Fibrosis Development and Lung Injury

The extent of fibrosis indices in the lungs of bleomycin-challenged mice were assessed with established markers such as collagen and α-smooth muscle actin (α-SMA) expression using immunohistochemical staining. The level of expression of the α-2 chain of type I procollagen (procol1A2) and the α1 chain of type III procollagen (procol3A1) was detected by quantitative RT-PCR using commercially available TaqMan probes (Applied Biosystems, Foster City, CA). Total collagen was assessed by the Sircol Collagen Assay kit (Biodye Science, Westbury, NY) according to the manufacturer's instructions.

For the assessment of lung injury and repair, levels of CD44v10 were analyzed by immunohistochemistry (IHC), and expression of tenascin-C (tnc) was analyzed at mRNA level. Using the specific probes available commercially, expression levels of cav-1 and actin were also analyzed at mRNA levels to understand their relationship with the development of fibrosis.

Analyses of Senescence and Apoptosis

Lung cell senescence was measured by senescence-associated β-gal activity as described by Debacq-Chainiaux and colleagues (24). The levels of senescence markers p16 and pRB and the heterochromatin-associated protein, macrohistone 2A (MH2A) were analyzed by immunohistochemical and immunofluorescent staining (6, 25). Quantification of apoptotic index in whole-lung paraffin-embedded sections was performed using terminal deoxynucleotide transferase-mediated dUTP nick end labeling (TUNEL) in the in situ cell death detection kit (Roche Diagnostics, Indianapolis, IN) according to the manufacturer's protocol. Quantification of apoptosis in epithelial cells was performed by flow cytometric analysis using an Annexin V/PI labeling kit (BD Pharmingen, San Diego, CA) and E-cadherin labeling (Cell Signaling, Danvers, MA) (26). Analysis was done on the isolated total lung cell population from WT and cav1^−/− mice at 2, 4, and 6 days after bleomycin challenge. Results are represented as percent total and E-cadherin–gated cells population showing annexin-V positivity and PI negativity to derive the actual number of cells undergoing apoptosis as opposed to the total apoptotic cells quantified by TUNEL assay.

Statistical Analysis

Student's t test and two-way ANOVA were used to determine statistical significance between groups. Data entry, management, and statistical analysis were performed using Prism software (GraphPad Software, San Diego, CA).

Effects of Bleomycin Injury on Body Weight and Lung Compliance in cav-1^−/− Mice

Bleomycin-treated WT mice were more symptomatic of the fibrotic disease process than the cav-1^−/− mice. There was 50% attrition rate in the WT population over the 21-day observation period, with the majority of deaths occurring within Days 10 to 15, compared with a 100% survival rate in the cav-1^−/− mice (see Figure E1A in the online supplement). WT mice showed a steady approximately 20% decrease in body weight by Day 12 (Figure E1B) before slowly recovering toward their original weight by Day 20. The bleomycin-treated cav-1^−/− mice showed an initial rapid 10% decrease in body weight by Day 2, preceding a steady increase in weight to 96% of original at Day 11 and 100% of original at Day 21. There was no significant difference in lung compliance between cav-1^−/− and WT mice treated with PBS (Figure E1C). However, in bleomycin-injured groups, WT mice at Day 11 demonstrated a 46% loss in compliance compared with only a 26% loss for cav-1^−/− animals. Together, these data indicate that cav-1 deficiency protects the mice against bleomycin-induced loss of body weight and lung compliance.

Development of Bleomycin-Induced Pulmonary Fibrosis Is Attenuated in cav-1^−/− Mice

Blind histological scoring of paraffin sections stained for collagen type I indicated that WT mice had a progressive and statistically significant increase in collagen deposition from Day 6 through Day 21 (Table 1). In contrast, in cav-1^−/− mice, we measured no significant increase in collagen deposition at any of the time points evaluated. For α-SMA levels, WT mice showed a progressive, statistically significant increase in the presence of α-SMA from Day 6 to that measured at Days 11 21. Conversely, cav-1^−/− mice showed a lack of response at Day 6 followed by a significant increase in α-SMA at Day 11 that returned to baseline levels by Day 21 (Table 1). Analysis of picrosirius red–stained tissue sections (Figure 1A) confirmed the data obtained through blind scoring of collagen type 1 deposition. In WT mice, the level of picrosirius red staining in lung parenchyma increased through Day 21. This increase was significant at all three measured time points compared with PBS-treated mice. However, in cav-1^−/− mice, there was no significant increase in collagen deposition. Reminiscent of the scored samples, we observed an initial significant increase from PBS mice at Day 11, which was lost by Day 21. As previously described (19), we measured significantly more collagen in the lungs of PBS-treated cav-1^−/− mice than in the lungs of their WT counterparts, indicating a higher background level of collagen in the cav-1^−/− mice. Thereafter, WT mice showed significantly more collagen than cav-1^−/− mice, whose collagen deposition had returned to baseline levels by Day 21 (Figure 1A). In addition, quantification of collagen was performed by SIRCOL assay and confirmed the reduced levels of collagen in bleomycin-challenged cav-1^−/− mice (0.18 ± 0.05 in the WT versus 0.04 ± 0.002 in cav-1^−/− mice on Day 21 after bleomycin treatment) (Figure 1B). Furthermore, to calculate whether the increase in collagen deposition correlated with a change in gene expression, we analyzed levels of procol1a2 and procol3a1 mRNA extracted from whole lung tissue (Figures 1C and 1D). Although not statistically significant, at Days 6 and 11 we detected a general increasing trend in procol1a2 mRNA expression in WT mice. This progressed into a statistically significant doubling (1.9 ± 1.4) of procol1a2 at Day 21 compared with PBS-treated mice. There was also an insignificant increase in procol3a1 expression at Day 21 in cav-1^−/− mice, where we measured a 1.5 ± 0.5-fold increase above averaged controls, compared with a 2.0 ± 0.8-fold increase in WT mice. These data suggested a correlation between the increased collagen depositions measured in bleomycin-treated WT mice and increased gene expression for procol1a2 and procol3a1.

Reduced Injury and Repair Response in the Lungs of cav-1^−/− Mice after Bleomycin Challenge

We measured expression levels of the matrix protein tnc mRNA as a marker of tissue repair (27–29). Tnc showed a marked increase in WT mice from Day 6 through Day 21 after bleomycin challenge. In addition, at Day 21, levels of tnc found in WT mice were significantly higher than those found in cav-1^−/− mice. A significant but smaller increase in tnc was measured in cav-1^−/− mice through Day 21 (Figure 2A). These data suggested the possibility that repair mechanisms are not triggered in cav-1^−/− mice after bleomycin challenge. To analyze the level of epithelial injury, we assessed levels of CD44v10 by IHC. CD44v10 is a marker of alveolar epithelial type II cells that also stains the bronchial epithelium in mouse lungs. During fibrogenesis, epithelial alterations can be monitored with epithelial isoforms of CD44 (30). In most cases, alveolar epithelial type II cell morphology changes toward an (intermediate) alveolar epithelial type I cell–like cell type. In bleomycin-injured WT mouse lungs, CD44v10 staining was prominent, compared with that of PBS-treated WT controls, which showed no change over time. However, cav-1^−/− mice showed mild positivity of CD44v10 upon bleomycin challenge, whereas no apparent difference was observed in the PBS-cav-1^−/− controls. More importantly, the positive correlation observed between epithelial injury and repair process in bleomycin-injured WT mouse lungs versus cav-1^−/− indicated that cav-1^−/− mice might have experienced less injury and therefore less repair process (Figure 2B). Injury and repair profiles are reduced in bleomycin-challenged cav-1^−/− mouse lungs.

It has been reported that cav-1 expression is decreased with the development of pulmonary fibrosis in mice and human IPF (15). To reconcile our findings with these previous published data, we determined that in our animal model, cav-1 expression levels were significantly decreased by over 2-fold in the lungs of WT mice after bleomycin challenge only starting at Day 11 through Day 21 (Figure 2C). Indeed, at Day 6 after injury, cav-1 expression was not affected when compared with PBS control samples. This decrease in cav-1 was a time-dependent process concomitant with the progression of fibrosis.

Absence of cav-1 Protects against Bleomycin-Induced Epithelial Cell Senescence and Apoptosis

Cellular senescence was evaluated using four well accepted senescence markers: the expression of SA-β gal activity, p16, pRB, and MH2A. The injury and repair processes showed distinct responses in the two bleomycin-challenged animal groups. Therefore, we focused our analyses of epithelial cell senescence at a midpoint on Day 11 after bleomycin challenge. Bleomycin-injured WT lung tissues presented an increase of SA-β gal activity after Day 11 (Figure 3A). In contrast, no evidence of the activation of this enzyme was detected in samples from bleomycin-injured cav-1^−/− lung tissue. Examination of nuclear localization of pRB by immunofluorescence staining also depicted a similar phenotype (Figure 3B). We then analyzed the profile of epithelial cell senescence on Days 6, 11, and 21 in bleomycin-injured lungs by IHC analysis of the senescence markers p16, pRB, and MH2A. The senescence nuclear localization of p16, pRB, and MH2A was not yet observed on Day 6 after injury on WT and cav1^−/− mice but showed a clear cytosolic staining in the alveolar lining and bronchial epithelial cells (Figure E2). On Days 11 and 21, challenged WT lung sections showed prominent nuclear localization of p16, pRB, and MH2A along with alveolar wall thickening and fibroblasts recruitment (Figure 3C and Figure E2). In contrast, in cav1^−/− mice, nuclear localization of senescence markers was reduced (Figure 3C and Figure E2).

Because senescence is decreased in the lungs of bleomycin-challenged cav-1^−/− mice, we assessed the impact of cav-1 during bleomycin-induced apoptosis. To understand if the apoptotic process was prominent during the early response to the bleomycin-induced injury to the lungs, we performed a TUNEL assay on lung tissue sections at 2, 4, and 6 days after bleomycin challenge. Two days after injury, we counted a near 3-fold increase in TUNEL-positive cells in bleomycin-challenged WT mice (15.6 ± 1.9%) compared with PBS-treated WT mice (5.3 ± 1.6%). At Day 4 and Day 6, apoptosis was still increased in bleomycin-challenged WT mice at 12 ± 2.3% and 7.8 ± 2.2%, respectively (Figure 4A). There was also a moderate increase in apoptosis at 2 days (7.0 ± 1.2%) and 4 days (5.7 ± 1.5%) after bleomycin treatment as compared with that of PBS-treated cav-1^−/− mice (2.6 ± 1.5%). By Day 6, the level of apoptosis in bleomycin-challenged cav-1^−/− mice had returned to baseline levels. At Day 4, approximately 50 to 60% of TUNEL-positive cells also stained positive for the epithelial marker K19 (data not shown) in bleomycin-challenged mice from both WT and cav-1^−/− groups.

To quantify the rate of apoptosis in epithelial cells from WT and cav1^−/− mouse lungs, we performed a flow cytometric Annexin-V/PI labeling assay by sorting the cells with the epithelial cell marker E-cadherin on Day 2, 4, and 6 after bleomycin-induced injury. A significant decrease in E-cadherin–sorted apoptotic cells was measured in the cav1^−/− samples (8.80 ± 0.92% in WT versus 4.42 ± 0.75% in cav-1^−/− mice; P = 0.002), with no significant difference between the two strains on Day 2 or 6.

Cav-1^−/− Mice Exhibit a Low Profile of SASP

We further examined the secretory phenotype of senescent cells (SASP) in bleomycin-challenged WT and cav-1^−/− mice by analyzing the levels of two of the most characteristic proteins of SASP, MMP-2 and MMP-9 (6, 7). The marked increase in MMP-2 levels in the perivascular and alveolar spaces observed in the bleomycin WT lungs sections was attenuated in cav-1^−/− counterparts at Day 11 after injury (Figure 5A). MMP-9 levels were more concentrated in the peribronchial areas and alveolar spaces in the WT bleomycin lung sections, and there was a reduced staining in the lung sections of cav-1^−/− mice (Figure 5A). MMP-2 and MMP-9 levels were only increased over time in WT lungs after bleomycin injury (Figure 5B). Tissue IL-6 levels were similarly increased in bleomycin-challenged WT mice (P ≤ 0.001) and cav-1^−/− mice (P = 0.002). In contrast, no significant difference was observed in MIP-2 levels in the lung homogenates of both strains after challenge with bleomycin (Figure 5C).