LncRNA MIR17HG promotes colorectal cancer liver metastasis by mediating a glycolysis-associated positive feedback circuit

Abstract

Glycolysis plays a crucial role in reprogramming the metastatic tumor microenvironment. A series of lncRNAs have been identified to function as oncogenic molecules by regulating glycolysis. However, the roles of glycolysis-related lncRNAs in regulating colorectal cancer liver metastasis (CRLM) remain poorly understood. In the present study, the expression of the glycolysis-related lncRNA MIR17HG gradually increased from adjacent normal to CRC to the paired liver metastatic tissues, and high MIR17HG expression predicted poor survival, especially in patients with liver metastasis. Functionally, MIR17HG promoted glycolysis in CRC cells and enhanced their invasion and liver metastasis in vitro and in vivo. Mechanistically, MIR17HG functioned as a ceRNA to regulate HK1 expression by sponging miR-138-5p, resulting in glycolysis in CRC cells and leading to their invasion and liver metastasis. More interestingly, lactate accumulated via glycolysis activated the p38/Elk-1 signaling pathway to promote the transcriptional expression of MIR17HG in CRC cells, forming a positive feedback loop, which eventually resulted in persistent glycolysis and the invasion and liver metastasis of CRC cells. In conclusion, the present study indicates that the lactate-responsive lncRNA MIR17HG, acting as a ceRNA, promotes CRLM through a glycolysis-mediated positive feedback circuit and might be a novel biomarker and therapeutic target for CRLM.

Fig. 1. Comprehensive WGCNA identified the glycolysis-related lncRNA MIR17HG as an oncogene during colorectal cancer liver metastasis.

A, B A total of eight pairs of adjacent normal, CRC and liver metastatic tissues were subjected to lncRNA (A) and mRNA (B) microarray analysis. Heat maps showed differentially expressed genes (eight biological replicates); C Cluster dendrogram of modules identified by WGCNA of lncRNA and mRNA microarray data. D Associations of modules with CRLM; E Four modules (MEdarkgrey, MEpurple, MEsalmon and MEtan) were found to be significantly and gradually increased from adjacent normal to CRC to paired liver metastatic tissues (Student’s t test); F Comprehensive analysis of KEGG pathways involved in glycolysis in MEpurple; G lncRNA MIR17HG expression gradually increased from adjacent normal to CRC to paired liver metastatic tissues (n = 8 for each condition); H The expression levels of MIR17HG in colorectal cancer tissues from patients with and without liver metastasis in the FUSCC dataset; I, J Kaplan–Meier analysis with log-rank test for evaluating MIR17HG expression for predicting the disease-free survival (DFS, I) and overall survival (OS, J) of CRC patients in the FUSCC dataset; K, L Roles of MIR17HG expression in predicting the overall survival of patients without liver metastasis (K) and patients with liver metastasis (L); M Representative 18F-FDG PET/CT images of colorectal cancer patients with low or high MIR17HG expression; N Difference analysis of SUVmax in the MIR17HG-low and MIR17HG-high groups (***P < 0.001).

Fig. 2. MIR17HG promotes glycolysis in colorectal cancer cells.

A Expression levels of MIR17HG in eight colorectal cancer cell lines (Caco2, HT29, SW480, SW620, RKO, HCT116, HCT8, and DLD-1) and one normal colon cell line (FHC) were determined by quantitative RT-PCR. B–C The efficiency of MIR17HG overexpression (B) and knockdown (C) in the indicated cells was validated by quantitative RT-PCR. Controls (ex-NC or sh-NC). D Changes in relative glucose consumption, ATP levels and lactate production in SW480 and HT29 cells after MIR17HG overexpression; E Changes in relative glucose consumption, ATP levels and lactate production in SW620 and RKO cells upon MIR17HG knockdown; F, G Changes in ECAR levels in MIR17HG-overexpressing SW480 (F) and HT29 (G) cells. The ECAR after oligomycin treatment indicates glycolytic capacity; H, I Changes in glycolytic capacity upon MIR17HG knockdown in SW620 (H) and RKO (I) cells. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 3. MIR17HG upregulates HK1 expression to promote glycolysis in colorectal cancer cells.

A mRNA levels of a series of glycolysis-associated genes in MIR17HG-overexpressing HT29 and SW480 cells (upper panel) and in MIR17HG knockdown RKO and SW620 cells (lower panel); B Western blot analysis of GLUT1, HK1, and LDHA protein levels in MIR17HG-overexpressing HT29 and SW480 cells as well as in MIR17HG-knockdown RKO and SW620 cells, in which the mRNA levels differed significantly; C Representative HK1 immunohistochemical (IHC) images in CRC tissue with low MIR17HG expression and in liver metastasis tissue with high MIR17HG expression (left panel); Pearson correlation between MIR17HG and HK1 expression in 105 primary CRC tissues in the FUSCC dataset (right panel); original magnification, ×200; scale bar, 100 μm; D E Relative changes in glucose consumption, ATP levels and lactate production in MIR17HG-overexpressing SW480 cells (D) or MIR17HG-knockdown RKO cells (E) with or without HK1 overexpression; F, G ECAR values in MIR17HG-overexpressing SW480 cells with or without HK1 knockdown (F) and MIR17HG-knockdown RKO cells with or without HK1 overexpression (G). *P < 0.05; **P < 0.01; ***P < 0.001; ns not significant.

Fig. 4. MIR17HG promotes the invasion and liver metastasis of CRC cells in vitro and in vivo by upregulating HK1 expression.

A, B Effects of knockdown or overexpression of HK1 on the migration ability of MIR17HG-overexpressing SW480 cells (A) and MIR17HG-knockout RKO cells (B) were detected by a wound healing assay; C, D Effects of HK1 knockdown or overexpression on the invasive ability of MIR17HG-overexpressing SW480 cells (C) and MIR17HG-knockdown RKO cells (D) were determined by a transwell assay; E Representative HE images of liver tissues obtained from nude mice with (lower panel) or without (upper panel) liver metastasis, original magnification, ×200; scale bar, 50 μm; F, G The number of liver metastatic foci in five randomly selected mice of each group was counted under a microscope; H, I Overall survival of each group of mice injected with engineered SW480 (H) or RKO (I) cells (n = 10 for each group). *P < 0.05; **P < 0.01; ***P < 0.001; ns not significant.

Fig. 5. MIR17HG sponges miR-138-5p to upregulate HK1 expression in colorectal cancer cells.

A The cytoplasmic and nuclear distribution of MIR17HG in colorectal cancer cells was measured by quantitative RT-PCR; B The binding sites in miR-138-5p, MIR17HG and the HK1 3′UTR were predicted by RNAhybrid; C The sequences of wild-type (WT) miR-138-5p and the designed mutant (MUT); D Levels of MIR17HG and β-Actin mRNA after streptavidin capture were measured in colorectal cancer cells transfected with wild-type biotinylated miR-138-5p or its mutant; E, F SW480 (E) and RKO (F) cells were transfected with wild-type miR-138-5p or the mutant. After AGO2 immunoprecipitation and validation of the efficiency by western blot analysis, the levels of MIR17HG and β-Actin mRNA were quantified by quantitative RT-PCR, and immunoprecipitate (IP)/input ratios were compared; G The sequences of reporter genes containing wild-type MIR17HG or its mutant without the miR-138-5p binding site (BS); H Effects of overexpression of wild-type miR-138-5p or its mutant on luciferase activity of wild-type or mutated MIR17HG in SW480 and RKO cells; I The sequences of reporter genes containing the wild-type HK1 3′ UTR or its mutant without the miR-138-5p binding site (BS); J, K Effects of miR-138-5p or the negative control miRNA on the luciferase activity of the wild-type or mutated HK1 3′ UTR in SW480 (J) and RKO (K) cells. **P < 0.01; ***P < 0.001; ns not significant.

Fig. 6. MIR17HG sponges miR-138-5p to promote glycolysis in CRC cells.

A, B Relative changes in glucose consumption, ATP levels and lactate production induced by transfection of miR-138-5p mimics into SW480 cells with or without MIR17HG overexpression (A) and transfection of miR-138-5p inhibitors into RKO cells with or without MIR17HG knockdown (B); C, D Effects of transfecting miR-138-5p mimics on ECAR levels in SW480 cells with or without MIR17HG overexpression (C) and effects of transfecting miR-138-5p inhibitors into RKO cells with or without MIR17HG knockdown (D). *P < 0.05; **P < 0.01; ***P < 0.001; ns not significant.

Fig. 7. MIR17HG promotes the invasion and liver metastasis of colorectal cancer cells by sponging miR-138-5p in vitro and in vivo.

A, B A wound healing assay was used to evaluate the effects of miR-138-5p mimics or inhibitors on the migration ability of SW480 (A) and RKO (B) cells upon MIR17HG overexpression and knockdown and in the corresponding control cells; C, D A transwell assay was applied to evaluate the effects of miR-138-5p mimics or inhibitors on the invasive ability of SW480 cells (C) or RKO cells with MIR17HG overexpression or knockdown and their corresponding control cells (D); E Representative HE images of liver tissues obtained from nude mice with (lower panel) or without (upper panel) liver metastasis, original magnification, ×200; scale bar, 50 μm; F,G The number of liver metastatic foci in mice of each group (n = 5 for each group) was counted under a microscope; H, I Overall survival of each group of mice injected with engineered SW480 (H) or RKO (I) cells (n = 10 for each group). *P < 0.05; **P < 0.01; ***P < 0.001; ns not significant.

Fig. 8. Lactate accumulation upregulates MIR17HG transcriptional expression via the p38/Elk-1 pathway.

A Effects of the monocarboxylate transporter inhibitor CHC (5 mM) on MIR17HG levels in SW480 and RKO cells stimulated with lactate (20 mM) for 12 h; B Different concentrations of L-lactate were used to stimulate SW480 and RKO cells for 12 h prior to quantitative determination of MIR17HG expression; C Nuclear translocation of phosphorylated p38 in lactate-stimulated cells was verified by western blot analysis; D Effects of the p38 chemical inhibitor SB203580 (10 μM for 12 h) on MIR17HG transcription in SW480 and RKO cells stimulated with lactate; E, F MIR17HG levels were detected after transfection of sh-Elk-1 plasmids into SW480 (E) and RKO (F) cells with or without lactate stimulation; G Effects of lactate and SB203580 on activation of p38 and Elk-1 in SW480 and RKO cells; H Schematic illustration of the binding site of Elk-1 in the MIR17HG promoter (upper panel); a ChIP assay was performed to investigate the roles of the two Elk-1 binding sites on the transcriptional activity of the MIR17HG promoter (lower panel); I Schematic illustration of reporter genes with the wild-type promoter of MIR17HG and its three mutants (upper panel); four kinds of reporter gene vectors were transfected into SW480 or RKO cells stimulated with lactate to confirm the binding site of activated Elk-1 in the MIR17HG promoter (lower panel). *P < 0.05; **P < 0.01; ***P < 0.001; ^###P < 0.001; ns and ns not significant. *, **, *** and ns indicate the statistical difference between the corresponding group and the group with no administration in Fig. 8A, D, E and F. ^### and ns indicate the statistical difference between the corresponding group and the group stimulated by lactate only.