A Tumor-suppressive Molecular Axis EP300/circRERE/miR-6837-3p/MAVS Activates Type I IFN Pathway and Antitumor Immunity to Suppress Colorectal Cancer

Abstract

Purpose: The oncogenic role of circular RNAs (circRNA) has been well studied in cancers including colorectal cancer. However, tumor-suppressive circRNAs and the mechanism through which they exert their antitumor effects remain largely unknown. We aim to find out the critical tumor-suppressive circRNAs and their possibility to serve as gene therapy targets.

Experimental design: circRNA sequencing, gain-of-function and loss-of-function experiments, and transcriptomic analysis were performed to find tumor-suppressive and antitumor immunity effects of circRERE. Molecular biology experiments were conducted for mechanism exploration. Finally, we conducted adeno-associated virus (AAV) to deliver circRERE (circRERE-AAV) and evaluated circRERE-AAV alone and in combination with anti-PD-1 antibody in C57BL/6J mice bearing subcutaneous MC38 tumors.

Results: circRERE is lowly expressed in colorectal cancer. Overexpression of circRERE inhibits the malignant behaviors of colorectal cancer in vitro and in vivo, while knockdown exhibits the opposite effects. The expression of circRERE is regulated by EP300, a histone acetyltransferase downregulated in colorectal cancer as well. Mechanistically, circRERE acts as a competitive endogenous RNA to sponge miR-6837-3p to upregulate MAVS expression, thereby activating type I IFN signaling and promoting antitumor immunity. Delivery of circRERE-AAV elicits significant antitumor effects, and combination treatment with circRERE-AAV and anti-PD-1 antibody exhibits synergistic effects on tumor growth in preclinical models of colorectal cancer.

Conclusions: These results uncover modulatory axis constituting of EP300/circRERE/miR-6837-3p/MAVS and its essential roles in antitumor immunity, and demonstrate that circRERE-AAV might represent a new therapeutic avenue to prime immune responses and boost the effects of immunotherapy in clinic.

Figure 1.

circRERE is lowly expressed in colorectal cancer tissues and required for colorectal cancer tumorigenesis. A, Three pairs of colorectal cancer (T1, T2, T3) and the matched adjacent normal tissues (N1, N2, N3) were subjected to circRNA-seq, and the expression of circRNAs detected is shown by volcano plot. B, The expression of the upregulated and downregulated circRNAs in tumors as shown in A is shown by heatmap (P < 0.05, FC > 1.5). C, The flow chart for identifying clinically relevant cirRNAs in colorectal cancer is shown. D, The expression of RERE in normal (n = 41) and colorectal cancer (n = 480) tissues was analyzed on the basis of data from TCGA database. E, Kaplan–Meier survival analyses for overall survival of all patients with colorectal cancer using RERE as input is shown. F, The expression of RERE (left) and circRERE (right) in 75 pairs of colorectal cancer (T) and the adjacent normal tissues (N) was examined by qRT-PCR analysis. G, The expression of circRERE in HCT116, RKO, DLD1, SW480, and LoVo colorectal cancer cell lines was examined by qRT-PCR analysis. H–J, DLD1 and LoVo cells infected with control vector or vector expressing circRERE were subjected to qRT-PCR analysis to examine the expression of circRERE (H), cell proliferation assay (I), and colony formation assay (J). K, The quantification of the number of colonies in J. L–N, HCT116 (left) and RKO (right) cells transfected with control siRNA (siNC) or two individual siRNA specifically targeting circRERE (sicircRERE-1 and sicircRERE-2) were subjected to qRT-PCR analysis to examine the expression of circRERE (L), cell proliferation assay (M), and colony formation assay (N). O, The quantification of the number of colonies in M. P, MC38 cells stably expressing control vector or circRere were subjected to qRT-PCR analysis to examine the expression of circRere. Q, MC38 cells stably expressing control vector or circRere were injected into C57BL/6J mice for 3 weeks, and the tumors were collected and photographed. R, The weight of tumors as described in N is shown. S, HCT116 cells stably expressing control shRNA (shNC) or two individual shRNAs targeting circRERE (shcircRERE-1 and shcircRERE-2) were injected into Balb/c nude mice for 2 weeks, and the tumors were collected and photographed. T, The weight of tumors as described in P is shown. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 2.

circRERE positively regulates the type I IFN signaling pathway. A, HCT116 and RKO cells transfected with siNC or sicircRERE were subjected to RNA-seq analysis, and the genes positively and negatively regulated by circRERE in both cell lines are shown by volcano plot. B and C, The expression pattern of genes regulated by circRERE as described in A is shown by heatmap (B) and box plot (C; q < 0.05, FC > 1.2). D, Hallmark analysis for genes positively regulated by circRERE is shown. E, HCT116 cells transfected with siNC or sicircRERE were subjected to qRT-PCR analysis to examine the expression of genes as indicated. F, DLD1 cells infected with control vector or vector expressing circRERE (OE) were subjected to qRT-PCR analysis to examine the expression of genes as indicated. G and H, Tumor samples as shown in Fig. 1N (G) and Fig. 1P (H) were subjected to qRT-PCR analysis to examine the expression of genes as indicated. I, Tumor sections from Fig. 1N were subjected to immunofluorescence staining by anti-CD8A or anti-CD3 specific antibody. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 3.

circRERE acts as a ceRNA to sponge miR-6837-3p to regulate the expression of MAVS. A and B, HCT116 (A) or RKO (B) cells transfected with siNC or sicircRERE were subjected to qRT-PCR analysis to examine the expression of genes as indicated. C, DLD1 (left) and LoVo (right) cells infected with control vector or vector expressing circRERE (OE) were subjected to qRT-PCR analysis to examine the expression of genes as indicated. D, HCT116 (left) or RKO (right) cells transfected with siNC or sicircRERE were subjected to immunoblotting analysis using antibodies as indicated. E, DLD1 (left) and LoVo (right) cells infected with control vector or vector expressing circRERE (OE) were subjected to immunoblotting analysis using antibodies as indicated. F and G, HCT116 cells overexpressed with circRERE were infected with control shRNA (shNC) or shRNA specifically targeting MAVS (shMAVS), followed by qRT-PCR to examine the expression of genes as indicated. H and I, HCT116 (H) or RKO (I) cells were subjected to RIP assay using control IgG or anti-AGO2 antibody, followed by qRT-PCR to detect circRERE enrichment. J and K, The sequence match between miR-6837-3p and the linear sequence of circRERE [circRERE (WT)] (J) and the 3′ UTR of MAVS [3′ UTR MAVS (WT)] (K) as well as the mutant form with the potential miR-6837-3p binding site mutated [circRERE (MT) or 3′ UTR MAVS (MT)]. L, The copy number of circRERE and miR-6837-3p in HCT116 (left) and RKO (right) cells is shown. M, HCT116 cells were subjected to ChIRP assay using biotin-labeled control probe or probe specifically targeting circRERE, followed by qRT-PCR analysis to detect the enrichment of miR-6837-3p and U6 snRNA. N and O, circRERE (WT), circRERE (MT), 3′ UTR MAVS (WT), and 3′ UTR MAVS (MT) sequences were cloned into psiCHECK2 vectors, which were then transfected into HCT116 and RKO cells with control or miR-6837-3p mimics, followed by luciferase activity measurement. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 4.

The regulation of MAVS by circRERE is dependent on miR-6837–3p. A and B, HCT116 and RKO cells transfected with siNC or sicircRERE in the presence or absence of miR-6837-3p inhibitors were subjected to qRT-PCR (A) or immunoblotting (B) analysis to examine the expression of MAVS. C, The representative densitometry graphic of B. D–G, HCT116 and RKO cells as described in A and B were subjected to qRT-PCR analysis to determine the expression of genes as indicated (D and E), cell proliferation assay (F), and colony formation assay (G). H, The quantification of the number of colonies in G. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 5.

EP300 regulates the expression of circRERE. A, The expression of EP300 in normal (n = 41) and colorectal cancer (n = 480) tissues was analyzed on the basis of the data from TCGA database. B, The correlation between the expression of EP300 and RERE in colorectal cancer tissues (n = 480) was analyzed based on the data from TCGA database. C and D, HCT116 and RKO cells transfected with control siRNA (siNC) or siRNA specifically targeting EP300 (siEP300) were subjected to RNA extraction and qRT-PCR analysis to examine the expression of EP300 (C) and circRERE (D). E and F, HCT116 and RKO cells infected with lentivirus expressing control sgRNA or sgRNA targeting EP300 gene promoter fused to dCas9-VP64 transcriptional activator were subjected to RNA extraction and qRT-PCR analysis to examine the expression of EP300 (E) and circRERE (F). G, HCT116 and RKO cells were subjected to ChIP analysis with control IgG or anti-EP300 antibody, followed by qRT-PCR analysis using two primer sets (P1 and P2) targeting the two EP300 binding sites predicted in the RERE promoter region. H, HCT116 and RKO cells were transfected with siNC or siEP300 followed by ChIP-qPCR analysis as described in G. I, The RERE promoter sequence [RERE promoter (WT)] and the mutant form with the predicted EP300 binding sites mutated [RERE promoter (MT)] were cloned into pGL3-basic luciferase reporter vector, which were then transfected into HEK293T cells in the presence or absence of control vector or vector expressing EP300, followed by luciferase activity measurement. J and L, HCT116 and RKO cells as described in C and D were subjected to qRT-PCR (J) analysis and immunoblotting analysis (L) as indicated to examine the expression of MAVS. K and M, HCT116 and RKO cells as described in E and F were subjected to qRT-PCR (K) analysis and immunoblotting analysis (M) to examine the expression of MAVS. N, HCT116 cells stably expressing shNC or shcircRERE were infected with lentivirus expressing control sgRNA or sgRNA targeting EP300 gene promoter fused to dCas9-VP64 transcriptional activator, followed by RNA extraction and qRT-PCR analysis to examine the expression of circRERE, EP300, and MAVS. O, HCT116 cells as described in N were subjected to immunoblotting analysis using antibodies as indicated. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 6.

circRERE-AAV is potent in inhibiting colorectal cancer tumor growth and capable of improving the therapeutic effects of anti-PD-1 antibodies. A, Balb/c nude mice were subcutaneously implanted with HCT116 cells, and treated with AAV-expressing negative control vector (NC-AAV) or circRERE (circRERE-AAV) for 2 weeks (once per week), and the tumors were collected and photographed. B, The weight of tumors as described in A is shown. C, The growth curve of tumors as described in A is shown. D, The body weight of mice as described in A is shown. E, The tumors as described in A were subjected to RNA extraction and qRT-PCR analysis to detect the expression of genes as indicated. F, C57BL/6J mice were subcutaneously implanted with MC38 cells, and treated with NC-AAV or circRere-AAV (once per week) in the presence or absence of anti-PD-L1 antibody for 2 weeks (twice per week), and the tumors were collected and photographed. G, The weight of tumors as described in F is shown. H, The growth curve of tumors as described in F is shown. I, The body weight of mice as described in F is shown. J, Tumor sections from F were subjected to immunofluorescence staining by anti-CD8A or anti-CD3 specific antibodies. *, P < 0.05; **, P < 0.01; ***, P < 0.001.