Focal adhesion kinase signaling determines the fate of lung epithelial cells in response to TGF-β

Abstract

Alveolar epithelial cell (AEC) injury and apoptosis are prominent pathological features of idiopathic pulmonary fibrosis (IPF). There is evidence of AEC plasticity in lung injury repair response and in IPF. In this report, we explore the role of focal adhesion kinase (FAK) signaling in determining the fate of lung epithelial cells in response to transforming growth factor-β1 (TGF-β1). Rat type II alveolar epithelial cells (RLE-6TN) were treated with or without TGF-β1, and the expressions of mesenchymal markers, phenotype, and function were analyzed. Pharmacological protein kinase inhibitors were utilized to screen for SMAD-dependent and -independent pathways. SMAD and FAK signaling was analyzed using siRNA knockdown, inhibitors, and expression of a mutant construct of FAK. Apoptosis was measured using cleaved caspase-3 and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining. TGF-β1 induced the acquisition of mesenchymal markers, including α-smooth muscle actin, in RLE-6TN cells and enhanced the contraction of three-dimensional collagen gels. This phenotypical transition or plasticity, epithelial-myofibroblast plasticity (EMP), is dependent on SMAD3 and FAK signaling. FAK activation was found to be dependent on ALK5/SMAD3 signaling. We observed that TGF-β1 induces both EMP and apoptosis in the same cell culture system but not in the same cell. While blockade of SMAD signaling inhibited EMP, it had a minimal effect on apoptosis; in contrast, inhibition of FAK signaling markedly shifted to an apoptotic fate. The data support that FAK activation determines whether AECs undergo EMP vs. apoptosis in response to TGF-β1 stimulation. TGF-β1-induced EMP is FAK- dependent, whereas TGF-β1-induced apoptosis is favored when FAK signaling is inhibited.

Keywords: FAK; TGF-β1; alveolar epithelial cell; apoptosis; differentiation.

Fig. 1.

Transforming growth factor- β1 (TGF-β1) can induce epithelial-myofibroblast plasticity (EMP) and functional phenotypical change in RLE-6TN cells. RLE-6TN cells were serum starved for 24 h then treated without or with TGF-β1 (2.5 ng/ml) for 48 h (A and B) or indicated time (C and D). We confirm that RLE-6TN cells undergo TGF-β1-induced EMP by both protein analysis and immunohistochemistry staining. A: Western blotting demonstrates the loss of the epithelial cell marker E-cadherin and increase in mesenchymal markers N-cadherin and vimentin and the myofibroblast marker α-smooth muscle actin (α-SMA). B: immunohistochemical staining demonstrates a zona occludens-1 (ZO-1) downregulation, increased vimentin expression, increased α-SMA expression, and the development of intermediate filaments. C: RLE-6TN cells were grown in a 3-dimensional (3-D) collagen gel matrix and treated with TGF-β1 (2.5 ng/ml). Immediately after treatment the sides of the gel were detached from the plate and contraction was measured at 24, 48, 72, and 96 h. TGF-β1 treatment of RLE-6TN cells induces significant collagen contraction at indicated time points demonstrating differentiation into a functional mesenchymal phenotype. Pictorial representation of TGF-β1 induced collagen contraction at each time point. D: percent gel contraction relative to time 0 h in the control and TGF-β1-treated cells. All experiments were repeated 3-4 times, and representative images are shown. #P < 0.01 and *P < 0.001 for TGF-β1-treated cells vs. controls at indicated time points.

Fig. 2.

TGF-β1-induced EMP can be inhibited by blocking the TGF-β receptor 1 (Alk5) and FAK pathways in RLE-6TN cells. A: RLE-6TN cells were grown to 50–60% confluence and treated with TGF-β1 alone or with the following protein kinase inhibitors (in mM): 20 PD98059, 6 SB203580, 1 SB600125, 15 Y27632, 10 SB431542, and 10 SU6656. Cells were then lysed at 48 h after treatment and Western blot analysis was performed for α-SMA. B: densitometry analysis of experiments in A. #P < 0.05 and *P < 0.01. C: RLE-6TN cells were grown as A, treated with TGF-β1 alone or with PF573228 (FAK inhibitor) at indicated concentration (µM). Cells were then lysed at 48 h after treatment and Western blot analysis was performed for α-SMA. D: densitometry analysis of experiments in C. *P < 0.01. E: RLE-6TN cells were grown to 50% confluence and were then transfected with plasmids overexpressing pcDNA (control), wild-type (WT) FAK, and Y397F-FAK mutant in a dominant negative fashion. Stably transfected cell lines were then serum starved then grown in the presence and absence of TGF-β1 (2.5 ng/ml) for 48 h. Western blot analysis for α-SMA demonstrates cells transfected with Y397F-FAK transfects had decreased α-SMA expression following TGF-β1 treatment. F: RLE-6TN cells grown in a 3-D collagen gel matrix in the presence and absence of TGF-β1 (2.5 ng/ml) for 48 h. Cell lines transfected with Y397F-FAK mutant were unable to contract the collagen gel in the presence of TGF-β1. G: quantification of collagen gel matrix contraction. All experiments were repeated 3-4 times, and representative images are shown. *P < 0.01.

Fig. 3.

Time course of SMAD3 and FAK activation after TGF-β1 stimulation. RLE-6TN cells were grown to 50–60% confluence and serum starved for 24 h. They were then treated with TGF-β1 (2.5 ng/ml), and Western blot analysis was performed for phophorylated SMAD3 (A) and phophorylated Y397 (C) of FAK for various time points up to 72 h. B and D: densitometric ratio for pSMA3 (B) and pFAK (D). Phosphorylated SMAD3 begins to increase at immediately after treatment, peaks around 1 h posttreatment, and returns to baseline by 6 h. Phosphorylated FAK does not begin to increase until ~1 h after treatment, peaks at 6 h posttreatment, and does not return to baseline until 24 h after treatment. All experiments were repeated 3 times, and representative images are shown.

Fig. 4.

TGF-β-induced FAK activation is SMAD3 dependent. RLE-6TN cells were grown to 50–60% confluence and serum starved for 24 h. A: cells were then treated with TGF-β1 (2.5 ng/ml) in the presence of pharmacological inhibitors of SMAD (SB431542) and FAK (PF573228) signaling for 48 h. Control includes the same concentration vehicle. Western blots were performed for SMAD3 and FAK autophosphorylation. B: relative densitometry was used to examine the TGF-β1-induced FAK activation in A. The data represent the relative level of phospho-397-FAK vs. total FAK ratio (relative to without TGF-β1). C: cells were transfected with either nontargeting siRNA or SMAD3 siRNA and treated without or with TGF-β1 for 48 h. Western blot analysis was done for phosphorylated SMAD3 and FAK. D: relative densitometry was used to examine the TGF-β1-induced FAK activation in C. SB431542 or SMAD3 siRNA blocked both SMAD3 autophosphorylation as well as FAK autophosphorylation in TGF-β1-treated cells. All experiments were repeated 3 times, and representative images are shown. *P < 0.01.

Fig. 5.

TGF-β1 induces apoptosis and EMP in RLE-6TN cells but not in the same cell at the same time. A: TGF-β1 induces apoptosis in RLE-6TN cells. RLE-6TN cells were grown to 50–60% confluence and serum starved for 24 h. They were then treated with TGF-β1 (2.5 ng/ml) for 48 h, and Western blot analysis for cleaved caspase-3 was performed at various time points. TGF-β1 treatment induced cleaved caspase-3, which peaked at 48 h posttreatment. B: RLE-6TN cells were grown to 50–60% confluence in 35-mm tissue culture dishes. They were then treated with TGFβ-1 (2.5 ng/ml) in serum-free media for 72 h. They were then stained with antibodies to α-SMA and with terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining. BA–BD: representative fields with control (BA), TUNEL staining (BB), α-SMA staining (BC), and merged images (BD). C: then, 5 high power magnification fields (hpf) were examined for the number of α-SMA-positive cells as described in B; the number of TUNEL-positive cells and the number of cells that were positive for both. In an average of five hpf fields, there were 73 α-SMA(−)/TUNEL(+) cells and only 1 TUNEL(+)/α-SMA(+) cell. *P < 0.01. D: cells were treated as described in B. Immunofluorescent costain of α-SMA and pY397-FAK (pFAK) (DA–DD) showed that FAK was activated and located on prominent focal adhesions (arrowheads) in cells containing α-SMA-positive fibers. Immunofluorescent costain of TUNEL and phospho-FAK (DA–DD) showed that TUNEL-positive cells did not have the active FAK and focal adhesions. All experiments were repeated 3-4 times, and representative images are shown.

Fig. 6.

Inhibition of FAK signaling increases TGFβ-1-induced apoptosis. RLE-6TN cells were grown to 50–60% confluence in 35-mm tissue culture dishes. They were then treated with TGFβ-1 (2.5 ng/ml) in serum-free media for 72 h, in the presence and absence of pharmacological inhibitors to FAK (PF573228; 5 µM) signaling. A: Western blot analysis for total and cleaved caspase-3 was performed. B: densitometric measures (of A) support that inhibition of SMAD signaling has no effect on TGFβ-1-induced caspase-3 cleavage, but inhibition of FAK signaling increases TGFβ-1-induced cleaved caspase-3 by ~3-fold (data corrected for GAPDH and represented as the ratio of cleaved caspase-3:total caspase 3). C: cells in A were subjected to TUNEL staining, and TUNEL-positive cells were quantified as in Fig. 5. All experiments were repeated 3-4, times and representative images are shown. *P < 0.01.

Fig. 7.

Schematic of the fates of alveolar epithelial cells (AECs) in response to TGF-β1. The data suggest that TGFβ-1 signaling through FAK is SMAD3 dependent. There is likely a feedback loop between FAK and SMAD3 signaling. There are likely other SMAD3-independent pathways that lead to TGFβ-1-induced FAK activation that are beyond the scope of this article. When FAK is autophosphorylated, it induces a phenotypical change in AECs to a more mesenchymal phenotype. When FAK signaling is inhibited, then there is a switch to a more apoptotic phenotype for TGFβ-1-treated AECs. This is likely through SMAD3-independent pathways that lead to apoptosis.