TGFβ Impairs HNF1α Functional Activity in Epithelial-to-Mesenchymal Transition Interfering With the Recruitment of CBP/p300 Acetyltransferases

Abstract

The cytokine transforming growth factor β (TGFβ) plays a crucial role in the induction of both epithelial-to-mesenchymal transition (EMT) program and fibro-cirrhotic process in the liver, where it contributes also to organ inflammation following several chronic injuries. All these pathological situations greatly increase the risk of hepatocellular carcinoma (HCC) and contribute to tumor progression. In particular, late-stage HCCs are characterized by constitutive activation of TGFβ pathway and by an EMT molecular signature leading to the acquisition of invasive and metastatic properties. In these pathological conditions, the cytokine has been shown to induce the transcriptional downregulation of HNF1α, a master regulator of the epithelial/hepatocyte differentiation and of the EMT reverse process, the mesenchymal-to-epithelial transition (MET). Therefore, the restoration of HNF1α expression/activity has been proposed as targeted therapeutic strategy for liver fibro-cirrhosis and late-stage HCCs. In this study, TGFβ is found to trigger an early functional inactivation of HNF1α during EMT process that anticipates the effects of the transcriptional downregulation of its own gene. Mechanistically, the cytokine, while not affecting the HNF1α DNA-binding capacity, impaired its ability to recruit CBP/p300 acetyltransferases on target gene promoters and, consequently, its transactivating function. The loss of HNF1α capacity to bind to CBP/p300 and HNF1α functional inactivation have been found to correlate with a change of its posttranslational modification profile. Collectively, the results obtained in this work unveil a new level of HNF1α functional inactivation by TGFβ and contribute to shed light on the early events triggering EMT in hepatocytes. Moreover, these data suggest that the use of HNF1α as anti-EMT tool in a TGFβ-containing microenvironment may require the design of new therapeutic strategies overcoming the TGFβ-induced HNF1α inactivation.

Keywords: CBP/p300; EMT; HCC; HNF1α; TGFβ; fibrosis; histone acetylation.

Figure 1

Analysis of HNF1α transcriptional activity in hepatocytes following transforming growth factor β (TGFβ) treatment. (A) Gene expression analysis by RT-qPCR of HNF1α and the indicated target genes in hepatocytes in a time-course experiment following TGFβ treatment. qPCR data, obtained in triplicate and normalized to the ribosomal RNA 18S, are expressed as fold change in gene expression in TGFβ-treated versus untreated cells (arbitrary value = 1) (^#p <0.06). The mean ± SEM of three independent experiments is shown. The observed differences in gene expression are statistically significant (*p < 0.05; **p < 0.01; ***p < 0.001). (B) Western blot analysis for HNF1α in protein extracts from one of the experiments shown in (A). α-Glyceraldehyde 3-phosphate dehydrogenase was used as loading control. Densitometric analysis of WB data from three independent experiments is shown.

Figure 2

TGFβ overrides HNF1α constitutive expression. (A) Analysis of constitutively expressed HNF1α activity following TGFβ treatment. RT-qPCR analysis for the indicated genes in hepatocytes, transiently transfected with pLCPX-HNF1α^Myc (HNF1) or the empty vector (CTR), treated with 4 ng/ml TGFβ for 24 h or left untreated (NT). qPCR data, obtained in triplicate and normalized to the housekeeping gene RPL34, are expressed as relative gene expression. The mean ± SEM of three independent experiments is shown. Statistically significant differences are reported (*p < 0.05; ns, not significant). (B) Immunofluorescence analysis of hepatocytes transfected and treated as in (A). Cells were stained with anti-MycTag antibody (red) and 4’,6-diamidino-2-phenylindole (nuclei, blue). Magnification 10×. (C) Analysis of EMT-related gene expression by RT-qPCR in hepatocytes, transiently transfected with pLCPX-HNF1αMyc (HNF1) or the empty vector (CTR), treated with TGFβ for 24 h or left untreated (NT). qPCR data, obtained in triplicate and normalized to the housekeeping gene RPL34, are expressed as relative gene expression. The mean ± SEM of three independent experiments is shown. Statistically significant differences are reported (*p < 0.05; ns, not significant). (D) Immunofluorescence analysis of hepatocytes transfected and treated as in (C). Cells were stained with anti-E-cadherin (green) or anti-HNF1α antibody (red). Magnification 20×.

Figure 3

(A) HNF1α DNA binding activity after TGFβ treatment. qPCR analysis of chromatin immunoprecipitated from hepatocytes with anti-HNF1α antibody was performed. HNF1α consensus regions embedded in the indicated HNF1α target gene promoters were analyzed. A HNF1α nonbound region of Neurogenin 1 promoter was utilized as negative control. Data are normalized to total chromatin input and background (control immunoprecipitation with IgG) and expressed as % input. Mean ± SEM of qPCR data obtained in triplicate from three independent experiments are reported. (B) Analysis of acetyl histone H3 by chromatin immunoprecipitation assay in untreated and TGFβ-treated hepatocytes. qPCR analysis of chromatin immunoprecipitated from hepatocytes with anti-acetyl H3 antibody was performed. HNF1α consensus regions embedded in the indicated HNF1α target gene promoters were analyzed. A HNF1α nonbound region of Neurogenin 1 promoter was utilized as negative control. Data are normalized to total chromatin input and background (control immunoprecipitation with IgG) and expressed as (Ip/IgG) % input. The mean ± SEM of qPCR data obtained in triplicate from five independent experiments are reported. *p <0.05, **p <0.01. (C) Electrophoretic mobility shift assay (EMSA) assays. Nuclear extracts from untreated (NT) or TGFβ-treated (for 5 or 24 h) parental (left, upper panel) or HNF1α-overexpressing hepatocytes (right, upper panel) were analyzed for the binding to HNF1α consensus site within the murine HNF4α promoter. The presence of the HNF1α protein in the protein/DNA complexes was revealed by the band supershift obtained with the addition of two different anti-HNF1α antibodies. A 200-fold excess of unlabeled oligonucleotide and the antitubulin antibody were added to the untreated extracts to test the binding specificity. As control, HNF4α DNA binding activity to its consensus site within the ApoC3 promoter was analyzed by EMSA in the same extracts (lower panel).

Figure 4

(A) Analysis of CBP/p300 DNA binding activity after TGFβ treatment. qPCR analysis of chromatin immunoprecipitated with anti-CBP/p300 antibody from hepatocytes untreated or treated with TGFβ (for 5 h) was performed. HNF1α consensus regions embedded in the indicated target gene promoters were analyzed. A HNF1α nonbound region of Neurogenin 1 promoter was utilized as negative control. Data are normalized to total chromatin input and background (control immunoprecipitation with IgG) and expressed as % input. Mean ± SEM of qPCR data obtained in triplicate from three independent experiments are reported. (B) In vivo coimmunoprecipitation of HNF1α and CBP/p300 in hepatocytes untreated or treated with TGFβ (for 5 h). Cells were lysed, immunoprecipitated with anti-HNF1α antibody, and then analyzed by Western blotting with the indicated antibodies. As control, the immunoprecipitation with normal goat IgG was performed. TCE, total cell extracts. Densitometric analysis of WB data from three independent experiments is shown. (C) Analysis of HNF1α protein PTMs following TGFβ treatment. Nuclear extracts from control hepatocytes (upper panel) and HNF1α^Myc-overexpressing hepatocytes treated for 3 h with TGFβ (lower panel) or left untreated (middle panel). Samples were separated by two-dimensional gel electrophoresis followed by Western blotting with anti-Myc-Tag antibody. HNF1α-specific spots are indicated by the arrow. The appearance of multiple spots in TGFβ-treated hepatocytes can be observed. **p <0.05, **p<0.01.

Figure 5

Proposed mechanisms of HNF1α and HNF4α inactivation by TGFβ in EMT. In hepatocytes, TGFβ induces the impairment of HNF4α DNA binding ability through the loss of activating phosphorylations of the protein and the impairment of HNF1α functional activity through the displacement of CBP/p300 acetyltransferases from target gene promoters, possibly mediated by change in PTM profile.