Farnesoid X receptor via Notch1 directs asymmetric cell division of Sox9+ cells to prevent the development of liver cancer in a mouse model

Abstract

Background: Asymmetrical cell division (ACD) maintains the proper number of stem cells to ensure self-renewal. The rate of symmetric division increases as more cancer stem cells (CSCs) become malignant; however, the signaling pathway network involved in CSC division remains elusive. FXR (Farnesoid X receptor), a ligand-activated transcription factor, has several anti-tumor effects and has been shown to target CSCs. Here, we aimed at evaluating the role of FXR in the regulation of the cell division of CSCs.

Methods: The FXR target gene and downstream molecular mechanisms were confirmed by qRT-PCR, Western blot, luciferase reporter assay, EMAS, Chip, and IF analyses. Pulse-chase BrdU labeling and paired-cell experiments were used to detect the cell division of liver CSCs. Gain- and loss-of-function experiments in Huh7 cells and mouse models were performed to support findings and elucidate the function and underlying mechanisms of FXR-Notch1 in liver CSC division.

Results: We demonstrated that activation of Notch1 was significantly elevated in the livers of hepatocellular carcinoma (HCC) in Farnesoid X receptor-knockout (FXR-KO) mice and that FXR expression negatively correlated with Notch1 level during chronic liver injury. Activation of FXR induced the asymmetric divisions of Sox9⁺ liver CSCs and ameliorated liver injury. Mechanistically, FXR directs Sox9⁺ liver CSCs from symmetry to asymmetry via inhibition of Notch1 expression and activity. Deletion of FXR signaling or over-expression of Notch1 greatly increased Notch1 expression and activity along with ACD reduction. FXR inhibited Notch1 expression by directly binding to its promoter FXRE. FXR also positively regulated Numb expression, contributing to a feedback circuit, which decreased Notch1 activity and directed ACD.

Conclusion: Our findings suggest that FXR represses Notch1 expression and directs ACD of Sox9⁺ cells to prevent the development of liver cancer.

Keywords: FXR; Liver cancer stem cell; Notch1; Sox9; Symmetric cell division.

Fig. 1

Notch1 level is elevated in the livers of FXR-KO mice and inversely correlate with FXR level in the DDC-treated mouse livers. a Expression of NICD1 and Notch1 was elevated in liver tumor samples from 12-month-old FXR-KO mice, normalized to LaminB or GAPDH. b Representative photomicrographs shown expression of FXR and Notch1 in 12-month-old WT and FXR-KO mouse liver tissues as assessed by an IF assay. Scale bars: 100 μm. c Expression of FXR and Notch1 mRNAs in the livers of WT and DDC-treated mice. d Expression of FXR and Notch1 in WT and DDC-treated mice was examined by western blotting, normalized to GAPDH. e The liver of DDC-treated mouse samples was co-stained with anti-FXR and anti-Notch1 antibodies, then counterstained with DAPI for confocal microscopy. Scale bars: 10 μm. Data were presented as mean ± SEM of three independent experiments. *p < 0.05

Fig. 2

FXR inhibits Notch1 expression in HCC cells and directs the asymmetric division of Sox9⁺ liver CSCs. a Huh7 cells were treated with CDCA (100 μM) or DMSO for 24 h, and the expression of SHP and Notch1 was examined by quantitative real-time PCR. b Huh7 cells were treated with GW4064 (10 μM) or DMSO for 24 h, and the expression of SHP and Notch1 was assayed by quantitative real-time PCR. GAPDH was used as an internal control for the examination of SHP and Notch1. c Effects of FXR agonists (CDCA or GW4064) on Notch1 protein expression. Huh7 cells were treated with CDCA (100 μM) or GW4064 (10 μM) for 36 h. Total protein samples were collected and subjected to Western blotting analysis to detect Notch1 protein expression, normalized to GAPDH. d the schematic depiction shown the BrdU pulse-chase and paired-cell assays. e Left: Representative images of the paired-cell assay in CD133⁺ liver CSCs from Huh7 cells. BrdU, red; DNA, blue; Sox9, green. Scale bars: 5 μm. Right: Quantification of BrdU asymmetry or symmetry in Sox9⁺ liver CSCs maintaining in 10% stripped FBS-containing medium supplemented with GW4064 (10 μM) or DMSO for 24 h. f Western blotting was used to detect the protein levels of NICD1 and Numb in Huh7 cells maintaining in 10% stripped FBS-containing medium supplemented with GW4064 (10 μM) or DMSO for 36 h. Data were presented as mean ± SEM of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 3

The overexpression of Notch1 reverses the FXR agonist-induced asymmetric cell division in liver CSCs. a Huh7 cells were transfected with Notch1 expression plasmid or vector plasmid, and quantitative real-time PCR confirmed Notch1 overexpression. b Huh7 cells were transfected with Notch1 expression plasmid or vector plasmid, Western blottings confirmed Notch1 overexpression, normalized to GAPDH. c After transfected with Notch1 expression plasmid or vector plasmid for 7 h, Huh7 cells were treated with GW4064 (10 μM) for another 24 h, and quantitative real-time PCR was assayed for SHP and Notch1 expression. d After transfected with Notch1 expression plasmid or vector plasmid for 6 h, Huh7 cells were treated with GW4064 (10 μM) for another 36 h, and Western blotting was assayed for Notch1 and NICD1 expression, normalized to LaminB or GAPDH. e Left: Representative images of the paired-cell assay of liver CSCs from Huh7 cells. BrdU, red; DNA, blue; Sox9, green. Scale bars: 5 μm. Right: Quantification of BrdU asymmetry or symmetry in liver CSCs after Notch1 overexpression expression maintained treated with GW4064 (10 μM) medium. Data were means of three separated experiments ± SEM, *p < 0.05, **p < 0.01

Fig. 4

FXR promotes asymmetric division of liver CSCs through regulation of Notch1. a The expression of FXR was examined by quantitative real-time PCR and Western blotting, normalized to GAPDH. b After interfering FXR expression for 12 h, Huh7 cells were treated with GW4064 (10 μM) or control DMSO for another 24 h, and quantitative real-time PCR was assayed for SHP and Notch1 expression. c After interfering FXR expression for 12 h, Huh7 cells were treated with GW4064 (10 μM) or Control DMSO for another 36 h, and Western blotting was assayed for Notch1 and NICD1 expression, normalized to LaminB or GAPDH. d Left: Representative images of the paired-cell assay of liver CSCs from Huh7 cells. BrdU, red; DNA, blue; CD133, green. Scale bars: 5 μm. Right: Quantification of BrdU asymmetry or symmetry in liver CSCs after interfering FXR expression maintained treated with GW4064 (10 μM) medium. e Left: Representative images of the paired-cell assay of liver CSCs from Huh7 cells. BrdU, red; DNA, blue; Notch1, Green. Scale bars: 5 μm. Right: The percentage of BrdU and Notch1 co-expression or inverse expression in daughter cells of CD133⁺ liver CSCs from Huh7 cells undergoing division. Data were means of three separated experiments ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 5

FXR binds to FXRE in Notch1 promoter region and suppresses Notch1 transcriptional activity. a Potential FXRE/DR7 in the Notch1 promoter region was predicted that the site of Notch1 was indicated by red letter, and the FXRE/DR7 was underlined. b pGL3-Notch1 FXRE-wt and pGL3-Notch1 FXRE-mut were separately co-transfected with the Renilla luciferase expression vector pRL-TK into HepG2 cells. After 6 h of incubation, the cells were treated with DMSO or GW4064 (10 μM) for 36 h. c The sequences of the DR7 probe and mutant DR7 probe were shown. d EMSA analysis of the binding of FXR proteins to the DR7 was performed. The position of the up-shifted FXR/RXR a complex was indicated. e ChIP assays were performed using chromatin isolated from GW4064-treated HepG2 cells. Data were means of three separated experiments ± SEM, *p < 0.05

Fig. 6

FXR activation inhibits Notch1 and promotes Sox9⁺ cells asymmetric division. a Serum ALT (left) and AST (right) activities were determined. b H&E staining of liver sections from WT and FXR-KO mice. Scale bars: 100 μm. c Liver tissues were subjected to quantitative real-time PCR assay for determination of SHP (left) and Notch1 (right) mRNA levels. d Liver tissues were applied to Western blotting analysis of NICD1, Notch1 and Numb protein levels, normalized to LaminB or GAPDH. e, f Representative images e and quantification f the asymmetric division frequency of Sox9⁺ cells in DDC fed WT and FXR-KO mice with either (DDC/control) or (DDC/GW4064) treatment. Scale bars: 10 μm. g A model for the regulation of asymmetric division by FXR-Notch1. Data represented the mean ± SEM (N = 4–6). Statistical significance of differences between each treatment and control group (*p < 0.05; **p < 0.01) were determined