Long Noncoding RNA MIR4435-2HG Suppresses Colorectal Cancer Initiation and Progression By Reprogramming Neutrophils

Abstract

MIR4435-2HG, also known as LINC00978, has previously been described as an oncogenic long noncoding RNA (lncRNA). However, we show here that Mir4435-2hg depletion promoted colorectal tumorigenesis and progression in in vivo models of colitis-associated colorectal cancer, spontaneous intestinal adenomatous polyposis, and subcutaneous tumors. Alteration of MIR4435-2HG in colorectal cancer cells did not change the potential for cell proliferation, migration, or invasion in vitro. RNAscope assays showed that most MIR4435-2HG was located in the tumor stroma, which caused high expression of MIR4435-2HG in colorectal cancer tumor tissue. Transcriptome analysis of colorectal cancer tissues from wild-type and Mir4435-2hg-deficient mice revealed Mir4435-2hg as a tumor suppressor gene that regulated the immune microenvironment. Loss of Mir4435-2hg led to a decline in neutrophils and elevation of polymorphonuclear myeloid-derived suppressor cells (PMN-MDSC). In tissue-specific Mir4435-2hg knockout mice, we confirmed that Mir4435-2hg depletion in neutrophils, but not in intestinal epithelial cells, promoted colorectal cancer progression. Mechanistically, Mir4435-2hg depletion enhanced the immunosuppressive ability of PMN-MDSCs by disturbing their fatty acid metabolism. These findings suggest that MIR4435-2HG is a tumor-suppressing lncRNA whose deficiency could increase tumor-infiltrating PMN-MDSCs and enhance the immunosuppressive potential of PMN-MDSCs to promote colorectal cancer development. This provides a theoretical basis for further illustrating the pathogenesis of colorectal cancer and a potential antitumor immunotherapy target.

Figure 1.

Mir4435-2hg deletion promotes colorectal tumorigenesis and progression. A, Meta-analysis for the expression of MIR4435-2HG in colorectal cancer tumor tissues compared to paired adjacent normal tissues for 10 datasets from TCGA and GEO. B, RNAscope assay for detecting the sublocation of MIR4435-2HG in human colorectal cancer tissues. Brown dots represent MIR4435-2HG (data representative of n = 5). Scale bar, 200 μm; zoomed, 50 μm. C, Schematic of three cohorts of AOM/DSS-induced colorectal cancer using wild-type (WT) and Mir4435-2hg^−/− mice. W, week; i.p., intraperitoneal injection. D, Representative H&E staining of colorectum from WT and Mir4435-2hg^−/− mice in the short-term AOM/DSS model. Scale bar, 2.5 mm; zoomed, 250 μm. Quantification of adenomas (E) and lymphoid follicles (F) in the colorectum, counted from three different H&E staining sections of one sample. Each spot represents one mouse, n = 9/group. Gross view (G) and representative H&E staining (H) of colorectum from WT and Mir4435-2hg^−/− mice in the classical term AOM/DSS model, n = 6/group. Scale bar, 2.5 mm; zoomed, 250 μm. Quantification of colon tumor number (I), tumor size (J), and tumor burden (K). For I and K, each spot represents one mouse. For J, each spot represents one tumor. Gross view (L) and representative H&E staining (M) of colorectum from WT and Mir4435-2hg^−/− mice in long-term AOM/DSS model, n = 5 for each group. Scale bar, 1 mm; zoomed, 250 μm. N, Quantification of colon tumor number. Each spot represents one mouse. O, Quantification of tumor size. Each spot represents one tumor. P, Percentage of mice with invasive adenocarcinomas in long-term AOM/DSS model. Q, Survival analysis of WT and Mir4435-2hg^−/− mice with AOM/DSS treatment, n = 10 per group. R, Quantification of tumor number from intestine and colorectum of 4-month Apc^Min/+/WT and Apc^Min/+/ Mir4435-2hg^−/− mice. Each spot represents one mouse, n = 13–14/group. S, Survival analysis of Apc^Min/+/WT and Apc^Min/+/ Mir4435-2hg ^−/− mice, n = 10–11/group. Data are presented as mean±SD; statistical significance was assessed by an unpaired t test (E, F, I, J, K, N, O, R) or log-rank test (Q and S; ns, P > 0.05; *, P < 0.05; **, P < 0.01).

Figure 2.

MIR4435-2HG regulates colorectal cancer development by remodeling the immune microenvironment. A, Heat map of differentially expressed genes, determined via RNA-seq, in long-term colitis-associated (AOM/DSS) colorectal cancer tumor tissue from WT and Mir4435-2hg^−/− (knockout, KO) mice, n = 3/group. B, Top downregulated biological process terms regulated by Mir4435-2hg depletion identified via gene ontology enrichment. C, Top upregulated biological process terms regulated by Mir4435-2hg depletion. D, Tumor growth curves for the subcutaneous colorectal cancer model, n = 8/group. 1 × 10⁵ MC38 cells were subcutaneously inoculated to WT and Mir4435-2hg^−/− (KO) mice. E, Tumors from WT and Mir4435-2hg^−/− mice in day 22 for the model in D. F, Quantification of tumor weight for the model in D. G, Tumor growth curves for the subcutaneous colorectal cancer model, n = 8/group. Control (mock) or Mir4435-2hg^−/− (KO) MC38 cells were subcutaneously inoculated to WT and Mir4435-2hg^−/− mice. H, Tumors from WT and Mir4435-2hg^−/− mice for the model in G. I, Quantification of tumor weight for the model in G. J, Flow cytometry analysis MC38 cell apoptosis after coculture with splenocytes from WT or Mir4435-2hg^−/− mice at the indicated ratios. K, Quantification of proportion of Annexin V⁺ MC38 cells, n = 3/group. J and K, Data are representative of three independent experiments. Data are presented as mean ± SD; statistical significance was assessed by an unpaired t test (ns, P > 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001).

Figure 3.

Mir4435-2hg depletion increases PMN-MDSCs. A, RNAscope assay showing MIR4435-2HG localization in tumor-infiltrating neutrophils in human colorectal cancer tissues. Representative image of five samples. Neutrophils were identified by their polymorphonuclear morphology. Brown signals represent MIR4435-2HG. Scale bar, 50 μm; zoomed, 10 μm. B, Variations in proportions of neutrophils, monocytes, eosinophils, and lymphocytes in peripheral blood of WT and Mir4435-2hg^−/− (KO) mice at the indicated timepoints for subcutaneous MC38 colorectal cancer tumors, n = 3/group. C, Flow cytometry splenic neutrophil apoptosis from tumor-free WT and Mir4435-2hg^−/− (KO) mice, and subcutaneous MC38 tumor-bearing wild-type (WT_TB) and Mir4435-2hg^−/− (KO_TB) mice, pregated on CD11b⁺Ly6G⁺ cells. D, Quantification of Annexin V⁺ neutrophils for mice in C, n = 3/group. E, Frequency of Annexin V⁺ neutrophils in in vitro cultured bone marrow (BM) cells from WT or KO mice with/without MC38 cultured medium (CM), n = 3/group. F, BIM protein expression assessed by flow cytometry in splenic neutrophils from tumor-free WT and Mir4435-2hg^−/− (KO) mice, and subcutaneous MC38 tumor-bearing wild-type (WT_TB) and Mir4435-2hg^−/− (KO_TB) mice. G, Quantification of BIM mean fluorescence intensity (MFI) from mice in (F), n = 3/group. H, Relative Bcl2l1 mRNA expression in BM-derived MDSCs from WT or KO mice with/without MC38 CM, n = 3/group. I, BCL2L1 protein expression assessed by flow cytometry in splenic neutrophils from tumor-free WT and Mir4435-2hg^−/− (KO) mice, and subcutaneous MC38 tumor-bearing wild-type (WT_TB) and Mir4435-2hg^−/− (KO_TB) mice. J, Quantification of BCL2L1 MFI for mice in (I), n = 4/group. K, Representative flow cytometry of CD14⁺ neutrophils from spleen of tumor-free WT and Mir4435-2hg^−/− (KO) mice, and subcutaneous MC38 tumor-bearing wild-type (WT_TB), Mir4435-2hg^−/− (KO_TB) mice, pregated on CD11b⁺Ly6G⁺ cells. L, Quantification of CD14⁺ neutrophils from K, n = 4 per group. B–L, Data are representative of three independent experiments. Data are presented as mean ± SD; statistical significance was assessed by an unpaired t test (ns, P > 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001).

Figure 4.

Mir4435-2hg depletion leads to increased tumor-infiltrating PMN-MDSCs and decreased T cells. Flow cytometry of tumor-infiltrating total neutrophils (A), CD14⁺ PMN-MDSCs (B), CD3⁺ T cells (C), CD8⁺ T cells (D), monocytes/M-MDSCs (E), total CD45⁺ cells (F), macrophages (G), and M2 macrophages (H) in subcutaneous MC38 tumors, n = 3–8/group. Data are representative of three independent experiments. I, Left, representative IHC of tumor-infiltrating PMN-MDSCs (S100a8) and CD8⁺ T cells (CD8) in the long-term AOM/DSS model. Scale bar, 100 μm. Right, quantification, n = 5/group. Data are presented as mean ± SD; statistical significance was assessed by an unpaired t test (ns, P > 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001).

Figure 5.

Loss of Mir4435-2hg enhances the immunosuppressive potential of PMN-MDSCs. A, Heat map of differentially expressed genes, determined via RNA-seq, in BM neutrophils from naïve 8-week WT and Mir4435-2hg^−/− (KO) mice, n = 3/group. B, Top upregulated biological process terms regulated by Mir4435-2hg depletion in BM neutrophils. GSEA for cholesterol homeostasis (C) and (D) fatty acid metabolism. Flow cytometry (E) and quantification of lipid accumulation (BODIPY staining; F) of splenic neutrophil from tumor-free WT and Mir4435-2hg^−/− (KO) mice, and PMN-MDSCs from subcutaneous MC38 tumor-bearing wildtype (WT_TB) and Mir4435-2hg^−/− (KO_TB) mice 20 days after MC38 cells injected, n = 3/group, pregated on CD11b⁺Ly6G⁺ cells. MFI, mean fluorescence intensity. G, Lipid accumulation (BODIPY staining) of WT and Mir4435-2hg^−/− BM-derived PMN-MDSCs with or without the stimulation of CT26 or MC38 cultured medium, n = 3/group. Relative Arg1 (H), Nos2 (I), and Cox2 mRNA expression (J) in BM-derived MDSCs from cultures in G, assessed by qRT-PCR, n = 3/group. K, Flow cytometry of ROS (DCFH-DA probe) in BM-derived PMN-MDSCs from cultures in G, pregated on CD11b⁺Ly6G⁺ cells. L, CFSE flow cytometry analysis of T-cell and CD8⁺ T-cell proliferation after coculture with WT or Mir4435-2hg^−/− MDSCs for 72 hours, n = 3/group. E–L, Data are representative of three independent experiments. Data are presented as mean ± SD; statistical significance was assessed by an unpaired t-test. ns, P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 6.

Neutrophil-specific deletion of Mir4435-2hg promotes colorectal cancer progression. A, Relative Mir4435-2hg expression in neutrophils and intestinal epithelial cells from naïve 8-week Mir4435-2hg^flox/flox, Mir4435-2hg^flox/flox Villin-Cre, and Mir4435-2hg^flox/flox S100a8-Cre mice, n = 2–3/group. B, Tumor growth curves for the subcutaneous MC38 colorectal cancer model, n = 8/group, tumors were harvested at day 22. C, Tumors from Mir4435-2hg^flox/flox and Mir4435-2hg^flox/flox S100a8-Cre mice from B. D, Quantification of tumor weight from mice in B. E, Flow cytometry of lipid accumulation of splenic neutrophil from tumor-free or subcutaneous MC38 tumor-bearing Mir4435-2hg^flox/flox and Mir4435-2hg^flox/flox S100a8-Cre mice, n = 5/group. F, Lipid accumulation in the bone marrow-derived PMN-MDSCs from the indicated mice with/without MC38 culture medium (CM), n = 3/group. G, Gross view of colorectum from Mir4435-2hg^flox/flox, Mir4435-2hg^flox/flox Villin-Cre, and Mir4435-2hg^flox/flox S100a8-Cre mice in AOM/DSS model, n = 6/group. H, Quantification of tumor number from mice in (G); each spot represents one mouse. I, Quantification of tumor size from mice in G; each spot represents one tumor. J, Quantification of tumor burden from mice in G; each spot represents one mouse. K, Left, representative IHC of tumor-infiltrating PMN-MDSCs (S100a8) and CD8⁺ T cells (CD8). Scale bar, 100 μm. Right, quantification, n = 5/group. (A, E, and F) Data are representative of three independent experiments. Data are presented as mean ± SD; statistical significance was assessed by an unpaired t test (ns, P > 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001).