Chemical regulators of epithelial plasticity reveal a nuclear receptor pathway controlling myofibroblast differentiation

Abstract

Plasticity in epithelial tissues relates to processes of embryonic development, tissue fibrosis and cancer progression. Pharmacological modulation of epithelial transitions during disease progression may thus be clinically useful. Using human keratinocytes and a robotic high-content imaging platform, we screened for chemical compounds that reverse transforming growth factor β (TGF-β)-induced epithelial-mesenchymal transition. In addition to TGF-β receptor kinase inhibitors, we identified small molecule epithelial plasticity modulators including a naturally occurring hydroxysterol agonist of the liver X receptors (LXRs), members of the nuclear receptor transcription factor family. Endogenous and synthetic LXR agonists tested in diverse cell models blocked α-smooth muscle actin expression, myofibroblast differentiation and function. Agonist-dependent LXR activity or LXR overexpression in the absence of ligand counteracted TGF-β-mediated myofibroblast terminal differentiation and collagen contraction. The protective effect of LXR agonists against TGF-β-induced pro-fibrotic activity raises the possibility that anti-lipidogenic therapy may be relevant in fibrotic disorders and advanced cancer.

Figure 1. Overview of the high-content screen.

(A) Immunofluorescence microscopy with the indicated antibodies after 72 h stimulation of HaCaT cells with 5 ng/ml TGF-β. A dotted frame marks fibronectin expression. (B) HaCaT cell images taken with ArrayScan. Nuclear distribution (blue) and fibronectin expression (green) changes upon TGF-β stimulation. (C) HaCaT cell image analysis using BioApplications. Displayed are changes in nuclear distribution (first row); nuclear spreading (second row) with two algorithms segmenting single nuclei (red spots) and groups of cells as colonies (green lines); fibronectin expression (third row) and fibronectin analysed as intensity and texture measurements within single cells and colonies (fourth row, green lines). White bars indicate 10 μm. (D) Schematic overview of data acquisition and analysis. 18 parameters selected (Fig. S3) from two individual image analysis algorithms (segmentation of nuclei as single cells and segmentation of grouped nuclei as colonies) were normalised plate-by-plate using the robust percentage-of-control (.poc) method and then normalised per parameter using the Z score (.zscore) method within the open-source software KNIME (HCS-Tools extensions). Clustering was performed with the k-means algorithm using k = 5 after normalisation of parameters using the Euclidian distance (phenotypic strength). The cluster containing EMT inhibitors was filtered for phenotypic strength greater than 15 and finally the hit list was condensed to 93 unique compounds that were hits in the primary screen. To remove compounds interfering with TGF-β signalling, a counter screen was performed using the p-Smad2 assay (see Fig. S4C,D) and 16 candidate EMT inhibitors were taken into further validation.

Figure 2. Optimisation of HCS assay using GW6604 and compound hits.

(A) Dose-response curves of TGF-β type I receptor kinase inhibitors LY-364947 (blue curves) and GW6604 (red curves) for four different parameters. The median IC₅₀ value of GW6604 was 1.9 μM and of LY-364947 was 0.2 μM. The parameters describe the texture of fibronectin signal within a colony (group of cells) (EntropyInten), the minimum distance to a neighboring nucleus (NeighborMinDist), the percentage of colonies with lower texture measurement of fibronectin signal within a colony than the threshold derived from a control population of colonies (ContrastCoocInten), and the ratio of convex hull to perimeter of a colony (ObjectConvexHullPerimeterRatio). (B) The scatter plot of two repeated runs of 7 plates (n = 2576) shows very good linear correlation with a Pearson coefficient of 0.927 for a single texture parameter (fibronectin signal). Negative control wells and inactive compounds scatter around the zero coordinate, whereas controls show phenotype strength-dependent scattering along the diagonal. (C) Table of false positive rate (FPR) and true positive rate (TPR) for the entire screen. Using the strict threshold of Euclidian distance >15 for the EMT cluster, the strong EMT inhibitor controls (GW6604: 3 and 10 μM) were identified as hits with a TPR of 98.3% and the negative control wells (DMSO) appeared as hits with a FPR of 0.21%. From the library 0.17% of wells were identified as hits. (D) Chemical structures of 16 compound hits. Simple heterocyclic and carboxylic acid analogs, steroidal and alkyl analogs, and extended multi-ring structures. Numbers correspond to the EPM IDs of each compound.

Figure 3. Analysis of 13 selected compounds in keratinocytes.

(A–C) Protein expression analysis in HaCaT total cell lysates stimulated (+) or not (−) with 5 ng/ml TGF-β for 96 h. In (A) cells were co-treated with DMSO or specific EPMs (10 μM). In (B) cells were co-treated with DMSO (0 μM) or EPM-1 and EPM-13 at the indicated concentrations. In (C) cells were co-treated with DMSO (0 μM) or EPM-1 and EPM-13 (10 μM) at the indicated time points following TGF-β stimulation. Immunoblots for the indicated proteins and Gapdh, the protein loading control, are shown. (D) E-cadherin immunofluorescence microscopy of HaCaT cells stimulated with vehicle or 5 ng/ml TGF-β for 96 h in the presence of DMSO or 10 μM of EPM-13 or 3.3 μM GW6604 (GW). (E) Actin microfilament direct fluorescence microscopy of HaCaT cells treated as in panel D. White bars indicate 10 μm.

Figure 4. Compounds that could suppress EMT transcription factor expression in HaCaT cells.

(A,B) mRNA expression analysis of Snail1 (A) and Snail2 (B) in HaCaT cells in the absence or presence of 5 ng/ml TGF-β for 72 h, and in the presence of DMSO or specific compound (EPM-1, EPM-10 and EPM-13), analysed by real-time RT-PCR and normalised against the housekeeping Gapdh mRNA. The data are expressed as bar graphs of average determinations with corresponding standard errors from triplicate determinations. Stars indicate significant difference at p < 0.05. (C,D) Number (C) and size (D) of hepatospheres grown in hanging drops using Insphero assays in the presence of control, DMSO or 10 μM EPM-1. The data are expressed as bar graphs of average determinations with corresponding standard errors from triplicate determinations. Stars indicate significant difference at p < 0.05. (E) Representative phase contrast images of hepatospheres grown in hanging drops using Insphero assays in the presence of control, DMSO or 10 μM EPM-1. (F) Protein expression analysis in the hepatospheres treated with DMSO or 10 μM EPM-1 under the same conditions as in panels C-E. Immunoblots for the indicated proteins and Gapdh, the protein loading control, are shown.

Figure 5. Analysis of 13 selected compounds in HTERT fibroblasts.

(A,B) Protein expression analysis in HTERT total cell lysates stimulated (+) or not (−) with 5 ng/ml TGF-β for 72 h and co-treated with (A) DMSO or specific EPMs (10 μM); (B) DMSO (0 μM) or EPM-6 and EPM-1 at the indicated concentrations. Immunoblots for the indicated proteins and for Gapdh, the protein loading control, are shown. (C) Actin microfilament direct fluorescence microscopy of HTERT cells stimulated with vehicle or 5 ng/ml TGF-β for 72 h in the presence of DMSO or 10 μM of EPM-6 and EPM-1. White bar indicates 10 μm. (D) Protein expression analysis in HTERT total cell lysates stimulated (+) or not (−) with 5 ng/ml TGF-β for 72 h and co-treated with DMSO or specific compounds (10 μM). Immunoblot for the indicated proteins and for Gapdh, the protein loading control, are shown.

Figure 6. LXR agonists block myofibroblast differentiation.

(A) Protein expression analysis in HTERT and AG1523 total cell lysates stimulated (+) or not (−) with 5 ng/ml TGF-β for 72 h and co-treated with DMSO (0 μM) or specific T0901317 (T0) compound at the indicated concentrations. Immunoblots for the indicated proteins and for Gapdh, the protein loading control, are shown. (B) αSMA microfilament immunofluorescence microscopy in AG1523 cells stimulated with vehicle or 5 ng/ml TGF-β for 72 h in the presence of DMSO or 10 μM of T0901317 and 3.3 μM GW6604 (GW). Images stained blue for DAPI-positive nuclei, red for F-actin microfilaments and green for αSMA microfilaments. White bar indicates 10 μm. (C) Protein expression analysis in AG1523 total cell lysates stimulated (+) or not (−) with 5 ng/ml TGF-β for 72 h in the presence of DMSO (control) or the indicated LXR agonists. β-Actin serves as a loading control. (D) Collagen gel contraction assay performed on AG1523 cells stimulated (+) or not (−) with 5 ng/ml TGF-β for 72 h in the presence of LXR agonist T0901317 or DMSO (control). A representative image and corresponding quantification of contracted gels graphed as average of 5 repeats with associated standard deviation. A star indicates statistically significant difference at p < 0.05. (E) Protein expression analysis in AG1523 total cell lysates stimulated (+) or not (−) with 5 ng/ml TGF-β for 24 h and co-treated with DMSO (0 μM) or specific T0901317 (T0) compound at the indicated concentrations. Immunoblots for the indicated proteins and for Gapdh, the protein loading control, are shown.

Figure 7. LXRα inhibits TGF-β-induced myofibroblast differentiation.

(A–D) Protein expression analysis in AG1523 (A–C) and MEF (D) total cell lysates stimulated (+) or not (−) with 5 ng/ml TGF-β for 72 h (A–D) or 24 h (C). In (A,B) AG1523 cells were stimulated with the LXR agonist T0901317 for 24 h in order to stabilise LXRα and LXRβ, thus providing evidence for the specificity of the detected protein band. In (A, B) AG1523 cells were also transfected with shRNA vectors targeting LXRα (A) and LXRβ (B) in order to show specificity of the detected protein band. In (C) AG1523 were transfected with the indicated siRNAs. In (D) MEFs were transiently transfected with LXRα cDNA vector. Immunoblots for the indicated proteins and for Gapdh, the protein loading control, are shown. Arrows mark the specific protein bands. (E) Direct and immunofluorescence microscopy of MEFs transfected and stimulated as in panel C and analysed for total F-actin and αSMA microfilaments. White bar indicates 10 μm. (F) Collagen gel contraction assay of MEFs transfected with LXRα cDNA vector (or control vector) and stimulated (+) or not (−) with 5 ng/ml TGF-β for 72 h. Quantification of the surface area of contracted gels is presented as in Fig. 6D (star: statistically significant difference at p < 0.05).