ANGPTL2 ‐mediated epigenetic repression of MHC‐I in tumor cells accelerates tumor immune evasion

Fig. 1. CKO tRCC mice exhibit enhanced infiltration of kidney tissues by activated CD8+ T cells. (A) Representative hematoxylin–eosin staining images of kidney tissues from Control (Rosa LSL‐PRCC‐TFE3/+ ; Angptl2 Flox/Flox ), tRCC (Cdh16‐Cre; Rosa26 LSL‐PRCC‐TFE3/+), and CKO tRCC (Cdh16‐Cre; Rosa LSL‐PRCC‐TFE3/+ ; Angptl2 Flox/Flox ) mice at 45–49 weeks of age. Scale bar, 2 mm (top) and 100 μm (bottom). Shown is a representative of three independent experiments. Arrowheads and asterisks indicate respective tumors and dilated renal tubules. T, tumor lesion. (B) Kaplan–Meier survival curves of control (n = 20), tRCC (n = 18), and CKO tRCC (n = 20) mice. Statistical significance was determined by log‐rank test. **P < 0.01; ***P < 0.001. (C) Immunofluorescent staining for CD8 (red) in kidney tissues from Control, tRCC, and CKO tRCC mice at 40 weeks of age. Shown is a representative of two independent experiments. Nuclei are counterstained with DAPI (blue). Arrowheads indicate CD8+ T cells. Scale bar, 100 μm. (D–G) Flow cytometry analysis of infiltrating lymphocytes in kidney tissues from tRCC and CKO tRCC mice at 45–50 weeks of age (n = 5 per group). The number of CD45+ cells (D) is shown…

Fig. 2. Angptl2 deficiency accelerates T‐cell‐mediated tumor cell eradication by increasing IFNγ‐induced MHC‐I expression in tumor cells. (A) Representative immunoblotting of ANGPTL2 in control KO (Control) and Angptl2 KO tRCC cell lines. Shown is a representative of two independent experiments. Hsc70 served as a loading control. (B) Relative expression of mRNAs encoding MHC‐I (H2‐K1 and H2‐D1) and antigen presentation machinery‐related molecules (B2m, Tap2, and Tapbp) in IFNγ‐treated or untreated control KO (Control) and Angptl2 KO (Angptl2 KO1 and KO2) cells (n = 4 per group). Levels in IFNγ‐untreated control cells were set to 1. Data are means ± SD. Statistical significance was determined by two‐way ANOVA with Tukey's post hoc test. *P < 0.05; **P < 0.01; ***P < 0.001. (C) Representative histograms showing cell surface expression of H‐2Kb and H‐2Db in IFNγ‐treated or untreated control KO (Control) and Angptl2 KO (Angptl2 KO1 and KO2) cells. Shown is a representative of three independent experiments. (D) Representative histograms showing cell surface expression of SIINFEKL‐bound H‐2Kb in IFNγ‐treated or untreated Control‐OVA, Angptl2 KO‐OVA, and Angptl2 KO cells. Shown is a…

Fig. 3. ANGPTL2 attenuates IFNγ‐induced MHC‐I expression via α5β1 integrin. (A) Representative histograms showing cell surface expression of α5 integrin, β1 integrin, α5β1 integrin, and PIR‐B in mouse tRCC cells. Shown is a representative of three independent experiments. (B) Representative histograms showing cell surface expression of α5 integrin in control KO (Control) and Itgα5 KO (Itgα5 KO1 and KO2) cells. Shown is a representative of three independent experiments. (C) Relative expression of transcripts encoding MHC‐I (H2‐K1 and H2‐D1) and antigen presentation machinery‐related molecules (B2m, Tap2, and Tapbp) in IFNγ‐treated or untreated control KO (Control) and Itgα5 KO (Itgα5 KO1 and KO2) cells (n = 4 per group). Levels in IFNγ‐untreated control cells were set to 1. Data are means ± SD. Statistical significance was determined by two‐way ANOVA with Tukey's post hoc test. *P < 0.05; **P < 0.01; ***P < 0.001. (D) Representative histograms showing cell surface expression of H‐2Kb and H‐2Db in control KO (Control) and Itgα5 KO (Itgα5 KO1 and KO2) cells. Shown is a representative of three independent experiments.

Fig. 4. ANGPTL2 promotes H3K27me3‐mediated repression of MHC‐I expression in tumor cells. (A) Relative expression of Irf1 and Nlrac5 mRNAs in IFNγ‐treated or untreated control KO (Control), Angptl2 KO, and Itgα5 KO cells (n = 4 per group). Levels in IFNγ‐untreated control cells were set to 1. Data are means ± SD. Statistical significance was determined by two‐way ANOVA with Tukey's post hoc test. **P < 0.01; ***P < 0.001; n.s, not significant. (B) Representative immunoblotting of phosphorylated STAT1 (Y701), total STAT1, and total IRF1 in control KO (Control), Angptl2 KO, and Itgα5 KO cells treated with indicated concentrations of IFNγ. Shown is a representative of three independent experiments. Hsc70 served as a loading control. (C) ChIP assay of H3K27me3 and H3K4me3 at the H2‐K1 and Tap1 promoters in IFNγ‐treated control KO (Control), Angptl2 KO, and Itgα5 KO cells (n = 4 per group). Data are means ± SD. H3K27me3 and H3K4me3 levels at each promoter are presented as a percentage of input. Statistical significance was determined by two‐way ANOVA with Dunnett's post hoc test. *P < 0.05; **P < 0.01; ***P < 0.001. (D) Relative expression of Ezh2, Eed, Suz12, Mtf2, and Jarid2 mRNAs in…

Fig. 5. ANGPTL2 knockdown increases MHC‐I expression in human tRCC cells. (A) Relative expression of ANGPTL2 mRNA in control (siScramble) and ANGPTL2‐knockdown (siANGPTL2) UOK120 cells (n = 4 per group). Levels in control cells were set to 1. Data are means ± SD. Statistical significance was determined by two‐sided unpaired Student's t‐test. ***P < 0.001. (B) Representative immunoblotting of ANGPTL2 in control (siScramble) and ANGPTL2‐knockdown (siANGPTL2) UOK120 cells. Shown is a representative of two independent experiments. Hsc70 served as a loading control. (C) Representative histograms showing cell surface expression of HLA class I molecules (HLA‐A, HLA‐B, and HLA‐C) in control (siScramble), ANGPTL2‐knockdown (siANGPTL2), and IFNγ‐treated (IFNγ) UOK120 cells (left) and quantification of fluorescent intensity in each group (right) (n = 5 per group). Mean fluorescent intensity (MFI) seen in control cells was set to 1. Data are means ± SD. Statistical significance was determined by one‐way ANOVA with Tukey's post hoc test. **P < 0.01; n.s, not significant. (D) Relative expression of HLA‐A, HLA‐B, HLA‐C, IRF1, and NLRC5 mRNAs in IFNγ‐treated or untreated control (siScramble) and…

Fig. 6. Activation of ANGPTL2 signaling is negatively correlated with expression of HLA class I in human cancer cells. (A) Scatter plots showing correlations between ANGPTL2, HLA‐B, HLA‐C, and JARID2 mRNA expression levels in metastatic lesion‐derived cell lines. Shown are Spearman's correlation coefficients r (Rho). (B–D) Scatter plots showing correlations between ANGPTL2 (B), ITGA5 (C), or JARID2 (D) mRNA levels and those of indicated genes in tumor samples from 417 patients with ccRCC. Shown are Spearman's correlation coefficients r (Rho).

Fig. 7. Proposed mechanism underlying ANGPTL2‐mediated evasion of immune eradication by tumor cells. (left) ANGPTL2 activates the α5β1 integrin pathway in tumor cells to induce JARID2 expression and accelerate repressive PRC2‐mediated H3K27me3 deposition at the MHC‐I promoter, thereby repressing IFNγ‐induced MHC‐I expression in tumor cells. As a result, tumor cells show reduced susceptibility to CD8+ T‐cell‐mediated anti‐tumor immunity, allowing tumor progression. (right) By contrast, tumor cells lacking ANGPTL2 expression show enhanced responsiveness to IFNγ and upregulation of MHC‐I expression due to decreased JARID2 expression. As a result, ANGPTL2‐deficient tumor cells are more readily eliminated by anti‐tumor CD8+ T‐cell‐mediated immunity.