Region-specific role for Pten in maintenance of epithelial phenotype and integrity

Abstract

Previous studies have demonstrated resistance to naphthalene-induced injury in proximal airways of mice with lung epithelial-specific deletion of the tumor-suppressor gene Pten, attributed to increased proliferation of airway progenitors. We tested effects of Pten loss following bleomycin injury, a model typically used to study distal lung epithelial injury, in conditional Pten^SFTPC-cre knockout mice. Pten-deficient airway epithelium exhibited marked hyperplasia, particularly in small bronchioles and at bronchoalveolar duct junctions, with reduced E-cadherin and β-catenin expression between cells toward the luminal aspect of the hyperplastic epithelium. Bronchiolar epithelial and alveolar epithelial type II (AT2) cells in Pten^SFTPC-cre mice showed decreased expression of epithelial markers and increased expression of mesenchymal markers, suggesting at least partial epithelial-mesenchymal transition at baseline. Surprisingly, and in contrast to previous studies, mutant mice were exquisitely sensitive to bleomycin, manifesting rapid weight loss, respiratory distress, increased early mortality (by day 5), and reduced dynamic lung compliance. This was accompanied by sloughing of the hyperplastic airway epithelium with occlusion of small bronchioles by cellular debris, without evidence of increased parenchymal lung injury. Increased airway epithelial cell apoptosis due to loss of antioxidant defenses, reflected by decreased expression of superoxide dismutase 3, in combination with deficient intercellular adhesion, likely predisposed to airway sloughing in knockout mice. These findings demonstrate an important role for Pten in maintenance of airway epithelial phenotype integrity and indicate that responses to Pten deletion in respiratory epithelium following acute lung injury are highly context-dependent and region-specific.

Keywords: adherens junctions; alveolar epithelium; bleomycin; cell adhesion; reactive oxygen species.

Fig. 1.

Analysis of SFTPC-cre activity and verification of Pten deletion in pulmonary epithelium. Analysis of Cre/loxP mediated green fluorescent protein (GFP) reporter activation in SFTPC-cre;ROSA^mT/mG double-transgenic mice. A: GFP reporter expression is evident in the entire alveolar epithelium, although the signal is much stronger in alveolar epithelial type II (AT2) cells (yellow arrow) compared with AT1 cells (red arrow). Blue is 4′,6-diamidino-2-phenylindole (DAPI) nuclear staining, red fluorescence represents Tomato-positive cells in which no Cre/loxP recombination has occurred, mainly in endothelial cells (barely visible). In SFTPC-cre;ROSA^mT/mG lung, GFP reporter expression is also detected in larger airways (B), smaller bronchioles (C), and at bronchoalveolar duct junctions (BADJs) (D). E: analysis of Pten expression in whole lung of Pten^SFTPC-cre knockout (KO) mice by quantitative RT-PCR. Pten mRNA levels were significantly (***P < 0.001) reduced to 48 ± 6% (means ± SE) of Pten^F/F control (F/F) levels (n = 8 of both genotypes). F: verification of Pten deletion in Pten^SFTPC-cre KO mice by Western blot analysis of freshly isolated type II (AT2) cells. G: PTEN protein expression in Pten^SFTPC-cre AT2 cells (KO, n = 3) was substantially reduced (~90%) compared with Pten^F/F control (F/F, n = 3) levels, as verified by quantitative analysis (***P < 0.001). H: increased levels of phosphorylated Akt (right) in KO lung compared with F/F controls. Total Akt levels (left) were similar between genotypes.

Fig. 2.

Hematoxylin and eosin (H&E) staining of lung sections. Top: sections of larger bronchioles (A), smaller bronchioles (B), and BADJ (C) from Pten^F/F mouse lung (F/F). Corresponding sections shown in D, E, and F demonstrate hyperplasia in the airways of Pten^SFTPC-cre KO mice. Immunofluorescent (DyLight 488, green) staining for club cell-specific protein (CCSP) (G–L). Most cells in the airways of Pten^F/F lung stain positive for CCSP as observed in larger bronchioles (G), smaller bronchioles (H), and BADJ (I), as does the corresponding hyperplastic airway epithelium in Pten^SFTPC-cre KO lung (J–L). H&E staining of floxed (M) and KO (P) distal lung does not reveal apparent morphological differences. Original magnification ×2. Immunofluorescent (Cy3, red) staining for surfactant protein C (SP-C) (N and Q) is not different between genotypes in terms of SP-C-positive cells as a percentage of total number of DAPI-positive cells. Antibody staining for aquaporin 5 (AQP5) expression in AT1 cells in the distal lung epithelium is similar in distribution and levels in F/F control and KO lungs (O and R).

Fig. 3.

Analysis of Pten^F/F control (F/F) and Pten^SFTPC-cre KO mice after bleomycin injury. A: percent survival of F/F and KO mice (n = 8 of each genotype) after intratracheal administration of bleomycin (1 U/kg body weight). There were no surviving KO mice on day 5, whereas all control mice were alive. B: body weight of F/F and KO mice during the first 4 days after treatment with bleomycin, shown as percent of body weight at day 0 (immediately before treatment). Number of animals was ≥8 for all time points. Four days after injury, body weight loss was greater (23.4 ± 0.72%, n = 17) in KO compared with F/F mice (7.61 ± 1.01%, n = 16) (P < 0.0001). C: arterial oxygen saturation levels were not significantly (NS) different between control and KO mice 3 days after injury. On day 4, oxygen saturation was significantly lower (*P < 0.05) in KO (74.3 ± 3.6%, n = 6) compared with F/F mice (86.3 ± 3.7%, n = 3). D: dynamic compliance (C_dyn) measured on day 4 was significantly lower (**P < 0.01) in KO (0.0266 ± 0.0014 ml/cmH₂O) than F/F (0.0344 ± 0.0006 ml/cmH₂O) mice.

Fig. 4.

Analysis of inflammation and injury on day 4 after bleomycin injury. A: average number of cells in bronchoalveolar lavage (BAL) fluid of bleomycin-treated Pten^SFTPC-cre KO mice (n = 13) was significantly higher (*P < 0.05) than that of bleomycin-treated Pten^F/F control (F/F) mice (n = 10). B: percent neutrophils in BAL was significantly higher (**P < 0.01) in bleomycin-treated KO mice (n = 18) than in bleomycin-treated F/F mice (n = 13). C: total protein concentration in BAL was significantly higher (*P < 0.05) in KO (n = 13) than F/F (n = 10) mice. D: semiquantitative analysis of distal lung injury in H&E-stained lung sections based on alveolar neutrophil infiltration, deposition of debris, and septal wall thickening. All mice were evaluated at three different depths per lung (details in materials and methods). The acute lung injury scores of F/F and KO mice were 26.9 ± 2.93 (n = 6) and 27.9 ± 3.95 (n = 4), respectively (NS). E and F: H&E staining of F/F (E) and KO (F) mouse lung sections showed no apparent differences in distal lung injury, interstitial inflammation, or perivascular cell infiltration between genotypes.

Fig. 5.

H&E and terminal deoxynucleotidyl transferase (TdT) dUTP nick-end labeling (TUNEL) staining of lung sections following bleomycin administration. Pten^SFTPC-cre KO mice instilled with PBS feature extensive airway epithelial hyperplasia (A), whereas the airway epithelium of KO mice treated with bleomycin (B) show sloughing of a substantial portion of the hyperplastic epithelial cells on day 2. Four days after bleomycin treatment, the airways of Pten^F/F controls (F/F) are clear (C), whereas smaller airways in lungs of KO mice are obstructed (D). Comparison of lungs of F/F (E) and KO (F) mice at 6 h after bleomycin administration demonstrates a higher number of apoptotic cells (green) in bronchioles of KO mice. DAPI (blue) stains nuclei of all cells. Quantification (G) shows 2.9 ± 0.6 TUNEL-positive cells per bronchiole in F/F mice and 13.4 ± 1.6 TUNEL-positive cells per bronchiole in KO mice (n = 3 of both genotypes, 10 bronchioles per lung). TUNEL staining 2 days after bleomycin injury (H and I) reveals a higher percentage of apoptotic cells in distal lung epithelium of KO (8.6 ± 1.7%, n = 3) compared with F/F (1.7 ± 0.8%, n = 3) mice, graphically represented in J (**P < 0.01).

Fig. 6.

Sod3 mRNA expression in Pten KO mice compared with floxed controls (F/F). Analysis of Sod3 expression was performed by quantitative RT-PCR in whole lung. Relative expression levels are shown in naïve animals (A) and on day 4 after bleomycin administration (B), where Sod3 expression in F/F lung is set to 1.0. Expression of Sod3 in KO lung is decreased at baseline and following injury compared with control lungs (**P < 0.01) (n = 3 of each genotype).

Fig. 7.

A–F: expression of E-cadherin, β-catenin, and integrin β4 in lung. Immunofluorescent (Cy3, red) staining is robust between epithelial cells in uninjured Pten^F/F control (F/F) lungs for E-cadherin (A) and β-catenin (C), whereas uninjured Pten^SFTPC-cre KO lungs display aberrant staining for both E-cadherin (B) and β-catenin (D), with loss of staining particularly in hyperplastic epithelial cells closest to the airway lumen. In lungs from bleomycin-injured animals, integrin β4 staining is similar in control (E) compared with KO mice (F). Blue represents nuclear staining by DAPI. Differential interference contrast (DIC) images were merged with Cy3 and DAPI images to enable outlining of lung and cell morphology.

Fig. 8.

A–H: expression of N-cadherin and fibroblast specific protein-1 (FSP-1) in naïve Pten^F/F (F/F) and Pten^SFTPC-cre KO lungs. In distal lung (A–D, top), positive staining (immunohistochemistry, using Vector Red as chromogen) is evident for both mesenchymal markers in AT2 cells in KO lungs (C and D), but not in control lungs (A and B). In conducting airways (E–H, bottom), distinct mesenchymal marker staining is present in most epithelial cells in the hyperplastic epithelium in bronchioles of KO (G and H) but not in the epithelium of F/F (E and F) mice. Nuclei were stained with hematoxylin.