Autophagy regulates hepatocyte identity and epithelial-to-mesenchymal and mesenchymal-to-epithelial transitions promoting Snail degradation

Abstract

Epithelial-to-mesenchymal transition (EMT) and the reverse process mesenchymal-to-epithelial transition (MET) are events involved in development, wound healing and stem cell behaviour and contribute pathologically to cancer progression. The identification of the molecular mechanisms underlying these phenotypic conversions in hepatocytes are fundamental to design specific therapeutic strategies aimed at optimising liver repair. The role of autophagy in EMT/MET processes of hepatocytes was investigated in liver-specific autophagy-deficient mice (Alb-Cre;ATG7(fl/fl)) and using the nontumorigenic immortalised hepatocytes cell line MMH. Autophagy deficiency in vivo reduces epithelial markers' expression and increases the levels of mesenchymal markers. These alterations are associated with an increased protein level of the EMT master regulator Snail, without transcriptional induction. Interestingly, we found that autophagy degrades Snail in a p62/SQSTM1 (Sequestosome-1)-dependent manner. Moreover, accordingly to a pro-epithelial function, we observed that autophagy stimulation strongly affects EMT progression, whereas it is necessary for MET. Finally, we found that the EMT induced by TGFβ affects the autophagy flux, indicating that these processes regulate each other. Overall, we found that autophagy regulates the phenotype plasticity of hepatocytes promoting their epithelial identity through the inhibition of the mesenchymal programme.

Figure 1

Liver-specific autophagy-deficient mice express decreased levels of epithelial genes and increased mesenchymal markers levels. (a and b) mRNA levels of the indicated genes were measured by quantitative real-time PCR (qPCR) in liver extracts of mice either autophagy-proficient (ATG7^fl/fl; white column; n=8) or autophagy-deficient (Alb-Cre+ATG7^fl/fl; black column; n=9). The values calculated by ΔΔCT method are relative to L34 mRNA levels and expressed as fold of change with respect to control mice. Data are expressed as mean±S.D. *P<0.01; #P<0.05; P-values were calculated by Mann–Whitney U-test. (c and d) Immunoblotting analysis for epithelial (c) and mesenchymal (d) markers of whole liver extracts of mice either autophagy-proficient (ATG7^fl/fl) or autophagy-deficient (Alb-Cre+ATG7^fl/fl). Tubulin and GAPDH were used for protein loading control

Figure 2

Lack of autophagy increases the mesenchymal markers expression. (a) Phase-contrast images of autophagy-deficient (siBECN1 or siATG7) or autophagy-proficient (siCTR) cells either untreated (NT) or treated with TGFβ (2 ng/ml) for 24 h. (b) mRNAs levels of the indicated genes were measured by qPCR in siCTR, siBECN1 and siATG7 cells either untreated (white columns) or treated with TGFβ (2 ng/ml) for 24 h (black columns). The values calculated by ΔΔCT method are relative to L34 mRNA levels and expressed as fold of change with respect to untreated siCTR cells. Data are shown as mean±S.D. of three independent experiments. *P<0.05; P-values were calculated by Mann–Whitney U-test. (c and d) The protein levels of epithelial and mesenchymal markers were measured in autophagy-proficient (siCTR) or autophagy-deficient (siBECN1; siATG7) cells either untreated or treated with TGFβ (2 ng/ml) for 24 h by western blotting. GAPDH (glyceraldehyde 3-phosphate dehydrogenase) was used for protein loading control. In panel (d), a higher amount of whole cell extracts proteins (30 versus 10 μg in panel (c)) were loaded to obtain a better Snail signal. (e) mRNA of Snail were measured as in panel (b) in untreated siCTR, siBECN1 and siATG7 cells. *P<0.05 versus siCTR

Figure 3

Autophagy degrades Snail in hepatocytes through the action of p62. (a) Colocalisation of LC3 with Snail was analysed by immunofluorescence staining for Snail (red) in hepatocytes expressing GFP-LC3 (green). Cells, either untreated or stimulated with TGFβ, were treated with BafA1 (5 ng/ml) for 3 h, fixed, permeabilised, stained with an anti-Snail antibody and analysed by confocal microscopy. In the magnified panels, arrows indicate the colocalisation of GFP-LC3 with Snail. (b) Immunoblotting analysis for Snail and p62 following immunoprecipitation of either p62 or using an isotype-matched control mAb from lysates of parental cells either untreated or stimulated with TGFβ (2 ng/ml) for 3 h. GAPDH (glyceraldehyde 3-phosphate dehydrogenase) was used for protein loading control. (c) Immunoblotting analysis for Snail and p62 following immunoprecipitation for HA (Snail) in parental cells or in hepatocytes overexpressing Snail-HA either in the presence or absence of p62-EGFP (48 h posttransfection). GAPDH was used for protein loading control. (d) Immunoblotting analysis of Snail and p62 following immunoprecipitation of either p62 or using an isotype-matched control mAb from lysates of autophagy-deficient (shBECN1) or control (shCTR) cells either untreated or stimulated with TGFβ (2 ng/ml) for 3 h. In whole-cell extracts (bottom panels), BECN1 and GAPDH were used for analysing the BECN1-silencing levels and as protein loading control, respectively. (e and f) Immunoblotting analysis for Snail and p62 of hepatocytes in which the expression of p62 was either inhibited by two specific siRNAs (e) or increased by transfecting a p62-EGFP-expressing vector (f). GAPDH levels were used for protein loading control

Figure 4

TGFβ induces impaired autophagic flux in nontumorigenic hepatocytes. (a) Immunoblotting analysis of LC3, p62 and BECN1 in hepatocytes treated with TFGβ (2 ng/ml) for the indicated hours. GAPDH (glyceraldehyde 3-phosphate dehydrogenase) levels were used as loading control. (b) Fluorescence analysis of the localisation of LC3 or p62 in hepatocytes expressing either LC3-GFP or p62-GFP, respectively, following treatment with TGFβ (2 ng/ml) for 24 h. The blue is a ToPRO3 nuclear counterstaining. (c) Immunoblotting analysis of LC3 in hepatocytes treated with TFGβ (2 ng/ml; 9 h) either in the presence or absence of the autophagy inhibitor NH₄Cl (20 mM) for the last hour of treatment. GAPDH was used for protein loading control. A representative experiment out of four is shown. (d) Quantification of the ratio of LC3-II to GAPDH band intensities obtained by immunoblotting analysis as in panel (c), expressed as average ±S.D. (n=4)

Figure 5

Autophagy stimulation inhibits the TGFβ-induced EMT and promotes Snail degradation in hepatocytes. (a) Phase-contrast images of cells treated with TGFβ (2 ng/ml) in the absence (CTR) or presence (STV) of the starvation medium for 12 h. (b) Cell counts of hepatocytes cultivated either in normal medium alone (CTR) or with 2 ng/ml of TGFβ (TGFβ) or cultivated in the starvation medium alone (STV) or with 2 ng/ml of TGFβ (STV+TGFβ) for 12 or 24 h, as indicated. (c) Levels of mRNAs of the indicated genes were measured by qPCR in CTR or STV cells either untreated (white columns) or treated with TGFβ (2 ng/ml) (black columns) for 12 h. (d) Levels of Snail mRNA were measured by qPCR in CTR (white columns) or STV (black columns) cells treated with TGFβ (2 ng/ml) for different time periods as indicated. The values calculated by ΔΔCT method are relative to L34 mRNA levels and expressed as fold of change with respect to untreated cells. Data are expressed as mean±S.D. of three independent experiments. *P<0,05; P-values were calculated by Mann–Whitney U-test. (e–g) Immunoblotting analysis for Snail in hepatocytes treated with TFGβ (2ng/ml) for different time periods as indicated. One hour before TFGβ treatment, cells were left untreated or co-treated with starvation medium (STV) (e), Torin1 (1 μM) (f) or Trehalose (100 mM) (g). GAPDH (glyceraldehyde 3-phosphate dehydrogenase) was used as protein loading control

Figure 6

Lack of autophagy impairs the MET in hepatocytes. (a) Phase-contrast images and immunofluorescence analysis of the localisation of E-cadherin in hepatocytes. Cells were treated with TGFβ (2ng/ml) for 24 h, extensively washed and after 0, 24 or 48 h of culture in the absence of TGFβ cells were fixed and stained for E-cadherin. (b) Immunoblotting analysis of E-cadherin, HNF4α, Snail and BECN1 in shCTR or shBECN1 hepatocytes. Cells were lysed after 24 h of TGFβ (2 ng/ml) treatment or after 24 or 48 h of further culture in the absence of TGFβ. GAPDH (glyceraldehyde 3-phosphate dehydrogenase) was used as a loading control. (c) mRNA levels of the indicated genes were measured by qPCR in shCTR or shBECN1 hepatocytes during EMT (24 h of 2 ng/ml of TGFβ; white columns) or during MET (further cultured in the absence of TGFβ for 24 (grey columns) or 48 (black columns) hours). The values calculated by ΔΔCT method are relative to L34 mRNA levels. Data are expressed as mean±S.D. of three independent experiments. *P<0.05; P-values were calculated by Mann–Whitney U-test