Deletion of platelet-derived growth factor receptor β suppresses tumorigenesis in metabolic dysfunction-associated steatohepatitis (MASH) mice with diabetes

Abstract

The platelet-derived growth factor (PDGF) family contributes to the progression of steatohepatitis; however, changes in and the characteristics of isoform-specific expression remain unclear. Since diabetes is a major driver of metabolic dysfunction-associated steatohepatitis (MASH), we characterized the mouse model of diabetic MASH (dMASH) by focusing on PDGF signaling. Pdgfa-d expression was markedly higher in hepatic stellate cells among flow-sorted cells in control mice and also increased in dMASH. In contrast, a reanalysis of human single-cell RNA-Seq data showed the distinct distribution of each PDGF isoform with disease progression. Furthermore, inflammation and fibrosis in the liver were less severe in diabetic MASH using tamoxifen-induced PDGF receptor β (PDGFRβ)-deficient mice (KO) than in control dMASH using floxed mice (FL) at 12 weeks old. Despite the absence of tumors, the expression of tumor-related genes was lower in KO than in FL. Tumorigenesis was significantly lower in 20-week-old KO. An Ingenuity Pathway Analysis of differentially expressed miRNA between FL and KO identified functional networks associated with hepatotoxicity and cancer. Therefore, PDGFRβ signals play important roles in the progression of steatohepatitis and tumorigenesis in MASH, with the modulation of miRNA expression posited as a potential underlying mechanism.

Keywords: Hepatic stellate cells (HSC); Metabolic dysfunction-associated steatohepatitis (MASH); MicroRNA; Platelet-derived growth factor (PDGF).

Fig. 1

Changes in cell populations and expression profiles of thePdgf family in diabetic MASH mice on the C57BL/6J background. (A) Schema of the protocol for generating diabetic MASH (dMASH) in C57BL/6j mice. (B) Gating of liver non-parenchymal cells. CD45⁻PDGFRβ⁺ cells, CD45⁻CD31⁺ cells, CD45⁺F4/80^loCD11^hi cells, and CD45⁺F4/80^hiCD11b^lo cells in liver non-parenchymal cells were isolated as hepatic stellate cells (HSC), endothelial cells (EC), Kupffer cells (KC), and infiltrated macrophages (MC), respectively. (C) The ratios and absolute cell numbers of each cell in control and dMASH livers. (D-I) mRNA levels of Pdgf and the Pdgfr family in each cell. Data are presented as the mean ± SEM (Control, n = 10; dMASH, n = 7). *p < 0.05 and **p < 0.01.

Fig. 2

Expression profiles of the PDGF family in healthy subjects and cirrhosis patients. Violin plot of PDGF-related gene expression by cell type reanalyzed using public scRNA-Seq data. The left and right columns show the results for healthy subjects and cirrhosis patients, respectively.

Fig. 3

Impact of the deletion of PDGFRβ on steatohepatitis in dMASH at 12 weeks old. (A) Schema of the protocol for producing PDGFRβ-deficient dMASH. KO and FL mice were orally administered tamoxifen for 5 consecutive days at 6 weeks old to achieve the knockout of pdgfrb. Mice were analyzed at 12 weeks old. (B) Representative photomicrographs of the liver in H&E and Sirius-Red staining, and a histological evaluation by NAS (FL, n = 13; KO, n = 9). Scale bar = 200 μm. (C) Serum AST and ALT levels (n = 5–6). (D-G) mRNA levels of Pdgf isoform-, inflammation-, fibrosis-, and tumorigenesis-related genes. Data are presented as the mean ± SEM (FL, n = 21; KO, n = 12). *p < 0.05, **p < 0.01, and ***p < 0.001.

Fig. 4

Impact of the deletion of PDGFRβ on tumorigenesis and progression in dMASH at 20 weeks old. (A) Schema of the protocol for producing PDGFRβ-deficient dMASH. Mice were analyzed at 20 weeks old. (B) Body and liver weights of mice (FL, n = 12; KO, n = 8). (C) Serum glucose (FL, n = 12; KO, n = 8) and AST and ALT levels (FL, n = 7; KO, n = 5). (D) Macroscopic appearance of the whole liver and a representative photomicrograph of a liver section in H&E staining. Arrows indicate the edges of tumors. Scale bar = 1 mm. (E) Total numbers of tumors and numbers of tumors by size in the entire left lateral, left median, and caudate lobes of FL and KO livers (FL, n = 8; KO, n = 4). Data are presented as the mean ± SEM. *p < 0.05 and **p < 0.01. (F) The top 10 up- and down-regulated miRNAs in the non-tumor regions of the livers of 20-week-old KO relative to FL (FL, n = 5; KO = 5). (G) Ingenuity Pathway Analysis (IPA) of miRNAs showing expression changes in KO livers. Significant functions identified in KO are listed according to their p-value.