Idiopathic pulmonary fibrosis (IPF) is a progressive disorder of unknown etiology characterized by accumulation of fibroblasts/myofibroblasts and marked deposition of extracellular matrix components 
. Epithelial-mesenchymal transition (EMT), a process whereby epithelial cells lose their phenotypic characteristics and acquire mesenchymal features, has been suggested as a mechanism that may contribute to fibroproliferation in pulmonary fibrosis 
. Currently, there is no effective treatment to improve prognosis for IPF patients 
. Given the lack of treatment options and the possible contribution of EMT to the pathogenesis of IPF, pharmacologic inhibition of EMT may represent a novel therapeutic approach. Such inhibition could have the effect of slowing or reversing established fibrosis of the lung.
Cumulative evidence, both in vivo
and in vitro
, indicates that transforming growth factor (TGF)-β1 is a primary regulator of EMT. Development of strategies to inhibit active TGF-β1 and its associated activities appears to be an attractive approach to prevention of EMT and/or IPF. Recent investigations have revealed that ligands of peroxisome proliferator-activated receptor gamma (PPARγ) are capable of opposing profibrotic effects of TGF-β1 
. Additionally, in epithelial cells of the airways, such ligands serve to inhibit proinflammatory cytokine release and contribute to regulation of cellular differentiation 
, further implicating them in the fibrotic process. PPARγ ligands include endogenous agents such as the cyclopentenone prostaglandin 15-deoxy-Δ12,14-prostaglandin J2 (15d-PGJ2) and a group of synthetic compounds known as thiazolidinediones (TZDs) that are currently in clinical use for their anti-diabetic effects. Of note, certain biological actions of TZDs have been shown to occur independently of PPARγ 
In murine models, TZDs ameliorate bleomycin-induced lung fibrosis 
. Specifically, they have been shown to inhibit TGF-β1-induced differentiation of lung fibroblasts to myofibroblasts 
as evidenced by suppression of α-smooth muscle actin (α-SMA) upregulation, and effects appear to be mediated via both PPARγ-dependent 
and -independent mechanisms 
. In the context of EMT, recent studies in retinal pigment and renal proximal tubule epithelial cells have demonstrated that some PPARγ ligands inhibit EMT induced by either TGF-β1 or high glucose, respectively 
. In the lung, inhibitory effects of TZDs on EMT have been shown in a lung adenocarcinoma cell line (A549) 
to be PPARγ-independent. However, conflicting results with regard to Smad-dependence or -independence of inhibitory effects of TZDs emerged from these studies. It is not known if these results and underlying mechanisms can be extrapolated to non-transformed alveolar epithelial cells (AEC).
In the current study, we examined the effects of troglitazone, a synthetic PPARγ ligand, on TGF-β1-mediated EMT in both primary AEC and a non-transformed rat lung epithelial cell line, RLE-6TN 
. Results reveal that troglitazone attenuates transition of both primary AEC and RLE-6TN cells to myofibroblasts, effects that are independent of PPARγ. Troglitazone inhibited EMT-related phosphorylation of Akt, GSK-3β and Smad2/Smad3, and two key downstream events (β-catenin nuclear translocation and SNAI1 activation), suggesting that effects of troglitazone are mediated by β-catenin-dependent signaling downstream of TGF-β. Given the importance of EMT in IPF, our findings point to a potential therapeutic role for TZDs in this disorder.
Culture of RLE-6TN Cells
RLE-6TN cells, a rat alveolar epithelial type II (AT2) cell line, were purchased from American Type Culture Collection (Manassas, VA). Cells were maintained in Dulbecco’s Modified Eagle’s medium, nutrient mixture F-12 Ham supplemented with 10% fetal bovine serum, 40 mmol/L HEPES, 100 U/ml penicillin G and 100 µg/ml streptomycin. For EMT studies, cells were allowed to attach overnight in media alone. For the majority of experiments, cells were maintained in either media alone or media supplemented with 2.5 ng/ml TGF-β1 (R&D Systems, Minneapolis, MN) with or without 10 µM troglitazone (Cayman Chemical, Ann Arbor, MI) for 3 days. Dose response effects of troglitazone (or rosiglitazone) were investigated at concentrations from 0 to 20 µM (or from 10-40 µM), respectively. Cultures were maintained in a humidified 5% CO2 incubator at 37°C, and all media and additives were replaced every other day, starting on day 2.
Primary AEC Isolation and Culture
AT2 cells were isolated from adult male Sprague-Dawley rats by elastase disaggregation (2.0–2.5 U/ml) and panning on rat IgG-coated bacteriological plates 
. All animals were treated in accordance with the guidelines and approval of the University of Southern California Institutional Animal Care and Use Committee. AT2 cells were resuspended in minimal defined serum-free medium (MDSF) 
. Cells were seeded into 1.1-cm2
tissue culture-treated polycarbonate (Nuclepore) filter cups (Transwell; Corning Costar, Cambridge, MA). Media were supplemented with 100 µg/ml cis
-OH-proline (Sigma, St. Louis, MO) for the first 24 to 48 hours of culture to selectively eliminate fibroblasts 
. Cells were subsequently maintained in MDSF or in MDSF supplemented with 2.5 ng/ml TGF-ß1 (R&D Systems) with or without 10 µM troglitazone in both apical and basolateral compartments for up to 12 additional days (for a total of 14 days). Equivalent amounts of vehicle for each supplement (4 mM HCl containing 1 mg/ml of bovine serum albumin (BSA) in the case of TGF-ß1 and dimethyl sulfoxide (DMSO) in the case of troglitazone) were added to control cultures. Cultures were maintained in a humidified 5% CO2
incubator at 37°C. Media were changed every other day. Cell viability (>95%) was measured by trypan blue dye exclusion. In studies investigating the impact of GW-9662 (Sigma), an irreversible PPARγ antagonist, cells were treated with TGF-β1 (2.5 ng/ml) ± troglitazone (10 µM) ± GW9662 (1.0–7.5 µM).
Monolayer Transepithelial Electrical Resistance (Rt)
Rt (KΩ·cm2) was measured using a rapid screening device (Millicell-ERS; Millipore, Bedford, MA). Effects of TGF-β1 supplementation (in the presence or absence of troglitazone) on Rt were evaluated on days 3, 5, 7, 9, and 10 following plating.
Cells were lysed in 2% sodium dodecylsulfate (SDS) lysis buffer (62.5 mM Tris-HCl, 2% SDS and 10% glycerol) on ice for 30 min and briefly sonicated. Protein sample concentrations were determined using a standard protein concentration assay (Bio-Rad, Hercules, CA). Samples were separated by SDS-polyacrylamide gel electrophoresis and transferred to Immuno-Blot polyvinylidene fluoride membranes (Bio-Rad). Membranes were blocked in 5% nonfat dry milk in Tris-buffered saline with Tween (TBS-T; pH 7.4) for 1 h at room temperature (RT). Incubation with primary antibodies was carried out overnight at 4°C, and with horseradish peroxidase-conjugated secondary antibodies at RT for 1 h. Primary antibodies for α-SMA, FLAG and β-catenin were obtained from Sigma and ZO-1 antibody was purchased from Invitrogen (Carlsbad, CA). Phospho-Akt (Ser473), total Akt, phospho-Smad2, total Smad2, phospho-Smad3, total Smad3, phospho-GSK-3β and total GSK-3β antibodies were purchased from Cell Signaling (Danvers, MA), and all secondary antibodies were obtained from Promega (Madison, WI). Peroxidase activity was detected with Super Signal (Pierce, Rockford, IL) and images analyzed using a FluorChem imager (Alpha Innotech, San Leandro, CA). To ensure equal loading, protein levels were normalized to the levels of lamin A/C, glyceraldehyde 3-phosphate dehydrogenase (GAPDH) or β-actin detected using anti-lamin A/C polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA), anti-GAPDH monoclonal antibody (Abcam, Cambridge, MA) or anti-β-actin monoclonal antibody (Sigma), respectively.
Production of Lentivirus in 293T Cells
PPARγ dominant negative expression plasmid, LV-PPARγ-DN (human PPAR LV-PPARγ-DN 1-L466A/E469A mutant cloned in pCDH1-MCS1-EF1-copGFP vector) was kindly provided by R.P. Phipps (University of Rochester, Rochester, NY). Infectious lentivirus was created by cotransfection of LV-PPARγ-DN or LV-control (pCDH1-MCS1-EF1-copGFP) with pCMVΔR8.91 and pMD.G into human 293T cells. Virus was harvested after 48 hours, filtered through 0.45 µm filters, concentrated with PEG-it virus precipitation solution (System Biosciences, Mountain View, CA ) and titered with HIV p24 ELISA (Cell Biolabs, San Diego, CA).
Overexpression of PPARγ-DN in RLE-6TN Cells
RLE-6TN cells were seeded at a density of 40,000/well in 24-well-plates and transduced with virus expressing PPARγ-DN (LV-PPARγ-DN) or LV-control at MOI
2 on day 1 postseeding, followed by TGF-β (2.5 ng/ml) ± troglitazone (10 µM) treatment 16 hours after transduction. Protein was harvested for Western analysis of α-SMA and expression of FLAG-tagged PPARγ-DN after 4 days of treatment.
Rat AEC grown as monolayers on polycarbonate filters and RLE-6TN cells grown on chamber slides were fixed in 4% paraformaldehyde for 15 min and blocked in CAS Block (Invitrogen) for 1 h at RT. Filters and slides were incubated with primary antibodies overnight at 4°C and incubated with Alexa Fluor 488 conjugated secondary antibodies (Invitrogen) at RT for up to 2 h. Slides were mounted using Vectashield antifade mounting medium with 4′,6-diamidino-2-phenylindole (DAPI) or propidium iodide (PI) (Vector, Burlingame, CA) for nuclear staining. Slides were viewed with an Olympus BX60 microscope equipped with epifluorescence optics (Olympus, Melville, NY).
Data are shown as mean ± SE (standard error of the mean). Significance (P<0.05) for more than or equal to 3 group means was determined by one-way analysis of variance followed by post hoc procedures based on Student-Newman-Keuls approaches. Where applicable, two group means were compared for significance using Student's t-tests. Z-tests were used to determine if ratiometric data (i.e., normalized) were different from control.