ANGPTL8 links inflammation and poor differentiation, which are characteristics of malignant renal cell carcinoma

Figures

• 

  FIGURE 1. ANGPTL8 in clear cell renal cell carcinoma (ccRCC) cells correlates with poor RCC prognosis and inflammation (A) Correlation of ANGPTL8 and CRP expression in ccRCC based on the analysis of The Cancer Genome Atlas database (n = 354). Each value is presented on a log scale. Pearson's product–moment correlation analysis. (B) Cell proliferation of ANGPTL8‐overexpressing Caki‐1 cells and control Caki‐1 cells with normal ANGPTL8 expression cells (mock). (C) Ranking of the top 10 upregulated gene ontology (GO) terms in Caki‐1 cells following ANGPTL8 expression based on the GO analysis of the CDS expression data of RNA sequencing. (D) CXCL1 and CXCL2 expression in ANGPTL8‐overexpressing Caki‐1 cells by RNA‐sequence analysis. Each gene (n = 3) is presented as fragments per kilobase of exon million mapped reads (FPKM). **p < 0.01; unpaired Student's t test.

• 

  FIGURE 2. Knockout of ANGPTL8 in clear cell renal cell carcinoma (ccRCC) cells downregulates CXCLs and upregulates urogenital development‐related genes. (A) Cell proliferation assay of wild‐type and ANGPTL8 knockout (KO) Caki‐1 cells. (B) Ranking of the top 10 upregulated gene ontology (GO) terms in Caki‐1 cells following ANGPTL8 KO based on GO analysis of CDS expression data of RNA sequencing data. (C) CXCL1 and CXCL2 levels in ANGPTL8 KO Caki‐1 cells by RNA‐sequence analysis. Each gene (n = 3) is presented as fragments per kilobase of exon million mapped reads (FPKM). (D) Ranking of the top 10 downregulated GO terms in Caki‐1 cells following ANGPTL8 KO based on the GO analysis of the CDS expression data of RNA sequencing data. **p < 0.01; unpaired Student's t‐test.

• 

  FIGURE 3. Knockout of ANGPTL8 in clear cell renal cell carcinoma (ccRCC) cells promotes a differential state in proximal tubule‐like cells. (A) FGFR2, GATA2, and HOXD13 levels in ANGPTL8 knockout (KO) Caki‐1 cells by RNA‐sequence analysis. Each gene (n = 3) is presented as fragments per kilobase of exon million mapped reads (FPKM). (B) CD44 and MET levels in ANGPTL8 KO Caki‐1 cells by RNA‐sequence analysis. Each gene (n = 3) is presented as fragments per kilobase of exon million mapped reads (FPKM). (C) FGFR2, GATA2, HOXD13, and ACTB protein levels in ANGPTL8 KO Caki‐1 and KMRC‐1 cells using immunoblotting. (D) FACS analysis of cell surface CD44 and MET in wild‐type Caki‐1 cells and ANGPTL8 KO Caki‐1 cells. Green line, ANGPTL8 KO Caki‐1 cells. Purple area, wild‐type Caki‐1 cells. (E) FACS analysis of LTL staining in wild‐type Caki‐1 and ANGPTL8 KO Caki‐1 cell cultures. (F) Representative images of the colony formation assay in wild‐type and ANGPTL8 KO Caki‐1 cells. Scale bar, 1,000 μm. (G) Quantitative analysis of the colony formation assay shown in Figure 3F. (H) Representative image showing cell motility of the wild‐type and ANGPTL8 KO Caki‐1 cells at 0, 6, and 24 h. Left, wild‐type…

• 

  FIGURE 4. ANGPTL8 regulates CXCL1 and CXCL2 and attracts immune cells, such as MDSC and TAM. (A) Comparison of CXCL1 and CXCL2 levels in the culture medium of wild‐type, ANGPTL8 knockout (KO), and ANGPTL8 KO rescued by the ANGPTL8‐overexpressing vector in Caki‐1 cells. Data represents means ± SD from three experiments. (B) Correlation of the ANGPTL8 and CXCL1, and CXCL2 expression in clear cell renal cell carcinoma (ccRCC) by analysis of The Cancer Genome Atlas (TCGA) database (n = 354). Each value is presented in log scale. Pearson's product–moment correlation analysis. (C) Migration of THP‐1 cells co‐cultured with wild‐type or ANGPTL8 KO Caki‐1 cells (n = 3). (D) Correlation of ANGPTL8 and CD33 or CD68 expression in ccRCC by analysis of TCGA database (n = 354). Each value is presented on the log scale. Pearson's product–moment correlation analysis. *p < 0.05, **p < 0.01; unpaired two‐tailed Student's t‐test.

• 

  FIGURE 5. ANGPTL8 activates NF‐κB signaling pathway (A) Ranking of the top 14 upregulated transcription factor targets in ANGPTL8 overexpressing Caki‐1 cells by the enrichment analysis of the CDS expression data of the RNA sequencing. (B) Ranking of the top 20 downregulated transcription factor targets in ANGPTL8 knockout (KO) Caki‐1 cells by the enrichment analysis of the CDS expression data of the RNA sequencing. (C) Relative comparison of luciferase activity between wild‐type and ANGPTL8 KO Caki‐1 cells transfected with an NF‐κB expression plasmid. Data represent means ± SD from three experiments. (D) Representative images of the immunoblotting analysis of p65 and p‐p65 protein levels in wild‐type and ANGPTL8 KO Caki‐1 cells. **p < 0.01; unpaired two‐tailed Student's t‐test.

• 

  FIGURE 6. STAT3 induces ANGPTL8 expression (A) Ranking of the top 18 transcription factors that are highly correlated with ANGPTL8 by enrichment analysis of the top 3,000 genes in The Cancer Genome Atlas TCGA KIRC dataset. (B) Relative ANGPTL8 expression in Caki‐1 cells treated with PBS or human recombinant IL‐6. Data from the PBS‐treated cultures was set as 1. Data represent means ± SD from three experiments, p = 0.08; unpaired two‐tailed Student's t‐test. (C) Relative ANGPTL8 expression in Caki‐1 cells treated with DMSO or Stattic, which is a STAT3 inhibitor. Data from the DMSO‐treated cultures was set as 1. Data represent means ± SD from three experiments, **p < 0.01 unpaired two‐tailed Student's t‐test. (D) Reporter assay for the STAT3‐binding site in the promoter region of ANGPTL8. Comparison of STAT3‐binding sites with and without mutation. The lack of mutation is shown as mock. Data represent means ± SD from three experiments, **p < 0.01; unpaired two‐tailed Student's t‐test. (E) ANGPTL8 maintains the undifferentiated state in renal carcinoma and upregulates CXCL1 and CXCL2 to attract immune cells, which activates STAT3 signaling for ANGPTL8 induction in the tumor microenvironment.