P53 deficiency affects cholesterol esterification to exacerbate hepatocarcinogenesis

Abstract

Background and aims: Cholesterol ester (CE) biosynthesis and homeostasis play critical roles in many cancers, including HCC, but their exact mechanistic contributions to HCC disease development require further study.

Approach and results: Here, we report on a proposed role of tumor suppressor P53 in its repressing ubiquitin-specific peptidase 19 (USP19) and sterol O-acyltransferase (SOAT) 1, which maintains CE homeostasis. USP19 enhances cholesterol esterification and contributes to hepatocarcinogenesis (HCG) by deubiquitinating and stabilizing SOAT1. Loss of either SOAT1 or USP19 dramatically attenuates cholesterol esterification and HCG in P53-deficient mice fed with either a normal chow diet or a high-cholesterol, high-fat diet (HCHFD). SOAT1 inhibitor avasimibe has more inhibitory effect on HCC progression in HCHFD-maintained P53-deficient mice when compared to the inhibitors of de novo cholesterol synthesis. Consistent with our findings in the mouse model, the P53-USP19-SOAT1 signaling axis is also dysregulated in human HCCs.

Conclusions: Collectively, our findings demonstrate that SOAT1 participates in HCG by increasing cholesterol esterification, thus indicating that SOAT1 is a potential biomarker and therapeutic target in P53-deficient HCC.

FIGURE 1

Cholesterol esterification mediates HCG induced by P53 deficiency. (A) Liver images from DEN-treated P53^fl/fl and P53^Δhep mice fed with NCD, HFD, and HCHFD, respectively. (B,C) Liver tumor nodules (B) and tumor size (C) from DEN-treated P53^fl/fl and P53^Δhep mice fed with NCD, HFD, and HCHFD. (D) ALT levels in DEN-treated P53^fl/fl and P53^Δhep mice fed with NCD, HFD, and HCHFD. (E) Heatmap showing CE species containing different lengths of fatty acids in HCC tissues from DEN-treated P53^Δhep mice fed with the indicated diets HCHFD. (F) Relative Soat1, Soat2, and Nceh1 transcript levels in DEN-treated P53^fl/fl and P53^Δhep mice fed with the indicated diets. (G) Western blottings of SOAT1, SOAT2, and P53 protein expression in HCC tissues from DEN-treated P53^fl/fl and P53^Δhep mice fed with NCD, HFD, and HCHFD, respectively. (H) SOAT1 activity in HCCs from (A). (I,J) Hepatic total cholesterol (I) and CE (J) levels from tissues in (A). (K) CE levels in HCCs from P53^fl/fl and P53^Δhep mice with indicated AAV8 virus treatment. Data are presented as mean ± SEM in (B), (C), (D), (F), (H), (I), (J), and (K). Statistical significance was calculated by using an unpaired t test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

FIGURE 2

Pharmacological inhibition of SOAT1 by avasimibe abrogates HCHFD-induced HCC. (A) Liver tumor nodules from P53^fl/fl and P53^Δhep mice fed with NCD (A, left) and HCHFD (A, right) treated as the indicated inhibitor. (B,C) Liver tumor size from P53^fl/fl and P53^Δhep mice fed with NCD (B) and HCHFD (C) treated as the indicated inhibitor. (D,E) ALT levels from P53^fl/fl and P53^Δhep mice fed with NCD (D) and HCHFD (E) following the indicated treatment. (F) Hepatic total cholesterol (CHOL) levels from P53^fl/fl and P53^Δhep mice fed with NCD (F, left) and HCHFD (F, right) following the indicated treatments. (G) Hepatic CE levels from P53^fl/fl and P53^Δhep mice fed with NCD (left) and HCHFD (right) following the indicated inhibitor treatments. (H) Hepatic SOAT1 activity in HCCs from P53^fl/fl and P53^Δhep mice fed with NCD (H, left) and HCHFD (H, right) following the indicated treatments. Data are presented as mean ± SEM in (A), (B), (C), (D), (E), (F), (G), and (H). Statistical significance was calculated by using an unpaired t test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

FIGURE 3

USP19 mediates SOAT1 deubiquitination and stabilization. (A) Screening for the deubiquitinating enzyme of SOAT1. The indicated DUBs were knocked down with shRNA in Huh7 cells, which were then assayed for SOAT1-luciferase 24 h later. Two shRNAs were used for each DUB independently, and the average value of the two shRNAs is shown. (B) Endogenous interaction of USP19 and SOAT1 in Huh7 cells. (C) Western blotting of SOAT1 expression in Huh7 cells following overexpression of the indicated vectors. (D) SOAT1 half-life in Huh7 cells ± USP19 depletion following CHX block. (E) USP19 inhibition promotes SOAT1 ubiquitination that can be repressed by coexpression of USP19 WT, but not C506S. (F) USP19 deubiquitinated SOAT1 at K120. Both PhosphoSite Plus and GPS websites were used to predict the possible ubiquitination sites of SOAT1 protein and mutate these sites from lysine to arginine. The URL of the PhosphoSite Plus website is: https://www.phosphosite.org/proteinAction.action?id=11918&showAllSites=true. The URL for the Group-based Prediction System website is: http://gps.biocuckoo.cn/. (G) CHX chase assay showing half-lives of SOAT1 WT and K120R following overexpression in HEK293 cells (upper). The intensity of SOAT1 expression for each time point was quantified by densitometry, with β-Actin as a normalizer (below). (H) CHX chase assay showing half-lives of SOAT1 WT and K120N mutant overexpression in HEK293 cells (top). The intensity of SOAT1 expression for each time point was quantified by densitometry, with β-Actin as a normalizer (bottom). (E,F) HEK293 cells were transfected with indicated vectors. IP and IB with indicated antibodies. Data are presented as mean ± SEM in (A). Data are presented as mean ± SD in (G) and (H). Statistical significance was calculated by using an unpaired t test. *p < 0.05; **p < 0.01; ***p < 0.001. HA, hemagglutinin; IB, immunoblotting; WCL, whole cell lysates.

FIGURE 4

P53 transcriptionally represses SOAT1 and USP19 expression. (A) Schematic representation of the SOAT1 promoter (−10 to +5 kb of the transcription start site [TSS]). P53 binding sites are indicated at the top. (B) Transcript levels of P53, SOAT1, SOAT2, and USP19 in FHCC98 cells in response to shRNA-mediated suppression of P53. (C) Immunoblottings of SOAT1 and USP19 in FHCC98 cells in response to shRNA-mediated suppression of P53. (D,E) mRNA (D) and protein (E) levels of SOAT1 and USP19 in FHCC98 cells following Nutlin-3a treatment for the indicated times. (F) ChIP-qPCR analysis of the SOAT1 promoter with IgG and P53 antibodies in FHCC98 cells. (G) ChIP-qPCR analysis of the indicated USP19 promoter regions from Figure S4A with IgG and P53 antibodies in FHCC98 cells. (H) Relative luciferase activity of SOAT1 promoter in FHCC98 cells in response to shRNA-mediated suppression of P53. M1 and M2 are the mutants of p53 binding sites (313 bp from TSS) and (321 bp from TSS) in the SOAT1 promoter region, respectively. DM is the double mutant of these two sites. (I) Relative luciferase activity of USP19 promoter in FHCC98 cells in response to shRNA-mediated suppression of P53. (J) Relative SOAT1 (left) and USP19 (right) mRNA levels in P53 WT and MUT (mutant) liver tumors from the TCGA LIHC data sets using the CAMOIP platform (https://www.camoip.net/). Data are presented as mean ± SEM in (B), (D), (F), (G), (H), (I), and (J). Statistical significance was calculated by using an unpaired t test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

FIGURE 5

SOAT1 accelerates HCC growth in response to P53 loss. (A) Liver images from DEN-treated fl/fl, P53^Δhep, SOAT1^Δhep, and SOAT1^ΔhepP53^Δhep mice fed with NCD or HCHFD. (B,C) Liver tumor nodules (B) and tumor size (C) from (A). (D) SOAT1 activity from HCCs shown in (A). (E) Heatmap of CEs containing different lengths of fatty acids in HCC tissues from (A). (F,G) Liver tumor nodules (F) and tumor sizes (G) from WT mice treated with the indicated AAV8 virus. (H-J) SOAT1 activity (H), total cholesterol (I), and CE (J) levels from WT mice with the indicated treatment. Data are presented as mean ± SEM in (B), (C), (D), (F), (G), (H), (I), and (J). Statistical significance was calculated by using an unpaired t test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. DKO, double knockout.

FIGURE 6

USP19 loss attenuates HCC growth in response to P53 loss. (A) Liver images from DEN-treated fl/fl, P53^Δhep, USP19^−/−, and P53^Δhep USP19^−/− mice fed with NCD or HCHFD. (B) Liver tumor nodules from the indicated mice fed with NCD (B, top) and HCHFD (B, bottom) from (A). (C) Liver tumor size from the indicated mice fed with NCD (C, top) and HCHFD (C, bottom) from (A). (D) SOAT1 activity in HCCs from (A). (E) Heatmap of CEs containing different lengths of fatty acids in HCC tissues from WT and Usp19^−/− mice. (F-J) Liver tumor nodules (F), tumor size (G), SOAT1 activity (H), CE (I), and hepatic total cholesterol (J) levels from WT and Usp19^−/− mice treated with the indicated AAV8 virus. Data are presented as mean ± SEM in (B), (C), (D), (F), (G), (H), (I), and (J). Statistical significance was calculated by using an unpaired t test. **p < 0.01; ***p < 0.001; ****p < 0.0001. DKO, double knockout.

FIGURE 7

P53-USP19-SOAT1 signaling axis is dysregulated in human HCCs. (A,B) P53 (A), USP19 (B, left), and SOAT1 (B, right) protein levels in HCC samples were compared with normal livers. (C-E) Hepatic total CHOL (C), hepatic free CHOL (D), and hepatic CE (E) in human HCC tissues compared with normal tissues. (F) Correlation of SOAT1 and USP19 protein levels in human liver tissues. (G-I) Correlations between P53 (G), USP19 (H), and SOAT1 (I) protein levels and hepatic CE levels in human liver tissues. In (F-I), each circle is an individual sample. R, Spearman correlation coefficient. p values are indicated. Significance was performed using the Wilcoxon signed-rank test. Horizontal lines in the box plots represent the median, boxes represent the interquartile range, and whiskers represent the minimal and maximal values. (J) Schematic diagram. P53 deficiency and P53 mutation transcriptionally activate USP19 and SOAT1 expression. USP19 deubiquitinates and stabilizes SOAT1 to promote cholesterol esterification. Avasimibe largely represses HCG. CHOL, cholesterol.