miR-423-5p contributes to a malignant phenotype and temozolomide chemoresistance in glioblastomas

Abstract

Background: Gliomas are based on a genetic abnormality and present with a dismal prognosis. MicroRNAs (miRNAs) are considered to be important mediators of gene expression in glioma tissues.

Methods: Real-time PCR was used to analyze the expression of microRNA-423-5p (miR-423-5p) in human glioma samples and normal brain tissue. Apoptosis, cell cycle, proliferation, immunostaining, transwell, in vitro 2D and 3D migration, and chemosensitivity assays were performed to assess the phenotypic changes in glioma cells overexpressing miRNA-423-5p. Western blotting was used to determine the expression of inhibitor of growth 4 (ING-4)in glioma tissues, and a luciferase reporter assay was conducted to confirm whether ING-4 is a direct target of miR-423-5p. Western blotting was used to identify the potential signaling pathways that are affected in glioma cell growth by miR-423-5p. Xenograft tumors were examined in vivo for the carcinogenic effects of miR-423-5p in glioma tissues.

Results: We first reported that miR-423-5p expression was increased in gliomas and was a potential tumor promoter via targeting ING-4. The overexpression of miR-423-5p resulted in upregulation of important signaling molecules such as p-AKT and p-ERK1/2. In clinical samples, miR-423-5p was dysregulated, and a corresponding alteration in ING-4 expression was observed (P = .0207). Furthermore, the overexpression of miR-423-5p strengthened glioma cell proliferation, angiogenesis, and invasion. Finally, miR-423-5p overexpression also strengthened GBM neurosphere formation and rendered glioma cells resistant to temozolomide (TMZ).

Conclusion: This study establishes that miR-423-5p functions as an oncogene in glioma tissues by suppressing ING-4 and suggests that it has therapeutic potential for glioma.

Keywords: ING-4; chemoresistance; glioma; miR-423-5p.

Fig. 1

miR-423-5p is upregulated in patient glioblastoma (GBM) specimens. (A) Relative expression of miR-423-5p was analyzed by The Cancer Genome Atlas in GBM brain tumor specimens versus healthy brain tissue. (B) Relative miR-423-5p expression levels were validated in 6 healthy brain tissues and 24 glioma tissues. (C) Relative expression levels of miR-423-5p in normal brain tissues and 3 different grades of glioma samples. The 24 glioma tissues were divided into 3 groups (8 glioma tissues in each group): WHO grade II, grade III, and grade IV. The Student t test was used to analyze the significant differences among the 3 groups; * and **indicate significant differences at P < .05 and P < .01, respectively.

Fig. 2

(A) Inhibitor of growth in the ING-4 3′-UTR and a schematic diagram of the reporter construct showed the mutated ING-4 3′-UTR sequence. (B) Direct targeting of the ING-4 3′-UTR by miR-423-5p was validated using a luciferase/3′-UTR reporter assay. U87 and U251 cells were cotransfected with a luciferase/ING-4 wild-type 3′-UTR reporter vector (wt) and 36 nM or 72 nM negative control miR (NC) or miR-423-5p. A reporter vector with a mutated miR-423-5p binding site in the ING-4 3′-UTR (mut) was used as a control. (C) Western blotting analysis showed that the expression levels of ING-4 were decreased in cells with miR-423-5p overexpression but increased in cells with an miR-423-5p inhibitor. (D) Relative expression of ING-4 protein levels was validated by Western blotting analysis in glioma specimens (n = 24) versus normal human brain tissues (n = 6). Fold changes were obtained from the ratio of ING-4-to-GAPDH levels. (E) Spearman correlation analysis was used to determine the correlations between the levels of ING-4 expression and miR-423-5p in human glioma specimens. (Spearman correlation analysis, r = −0.4692, P = .0207). *P < .05, **P < .01 and ***P < .001.

Fig. 3

miR-423-5p overexpression promotes cell proliferation, angiogenesis, and invasion. (A) The overexpression of miR-423-5p promoted cell proliferation, which was rescued upon induction of the expression of exogenous ING-4 in both U87 and U251 cells. *indicates significant difference compared with control, #indicates significant difference compared with miR-423-5p/ING-4 treatment at P < .01. (B) Cell-cycle assay results of U87 and U251 glioma cells 3 days after transfection with miR-423-5p or control miR (monitored by flow cytometry). (C) Western blot analysis was performed to indicate the regulation of cell-cycle–regulated proteins, cyclin D1, cyclin E1, and p-RB by miR-423-5p in U87 cells. (D) Apoptosis assay results of U87 and U251 glioma cells 3 days after transfection with the miR-423-5p inhibitor or the control miR were monitored by flow cytometry. (E) Tube formation of human brain microvessel endothelial cells (HBMVECs) transfected with miR-423-5p or control miR was monitored by Matrigel-coated plates with conditioned medium from the indicated cells. Scale bar = 300 μm. (F) miR-423-5p overexpression promoted cell invasion in U87 cells. Cells were transfected with miR-423-5p, which was followed by ING-4 transfection. Scale bar = 100 μm. *P < .05 and ***P < .001.

Fig. 4

Overexpression of miR-423-5p in glioblastoma (GBM) cells strengthens neurosphere formation and migration through multiple effectors. (A) CD133 marker, Nestin, and DAPI staining shown by immunostaining in neurosphere from U87 and N3 cells. Scale bar = 20 μm (B) Neurosphere formation capacity in the presence of miR-423-5p overexpression was determined by a self-renewal assay. miR-423-5p overexpression promoted neurosphere formation of both U87 and N3 cells. Scale bar = 150 μm. (C) The stemness of U87 and N3 cell lines, which were cultured as monolayer (M) or stem cell-like neurospheres (SCs), was determined by Western blot analysis. (D) Cellular signaling and the cellular receptors EGFR and PDGFR were monitored by Western blotting analysis of U87 and N3 cell lines that were cultured as monolayer (M) or stem cell-like neurospheres (SC). Cells were transfected with either negative control miR (NC) or miR-423-5p. (E) The migration of GBM cells was monitored by a spheroid dispersal assay. Representative images and quantification of spheroid migration of U87 GBM cells transfected with either negative control miR (NC) or miR-423-5p. miR-423-5p mediated GBM migration was rescued upon forced coexpression of ING-4. Scale bar = 200 μm. The same results were validated by wound-healing assay. Scale bar = 200 μm. (F) Western blot analysis to demonstrate the regulation of invasion-related proteins, MMP-2 and MMP-9, by miR-423-5p in U87 cells. *P < .05, **P < .01 and ***P < .001.

Fig. 5

miR-423-5p confers resistance to temozolomide (TMZ) by suppressing its target ING-4. (A) Cell proliferation was examined in U87 and U251 cells that were stably expressing miR-NC or miR-423-5p following TMZ treatments at different doses. The CCK8 assay was conducted 48 hours after TMZ treatment. (B) The cell proliferation of U87 and U251 cells, which were treated with 100 μM TMZ, were tested every 24 hours. The overexpression of miR-423-5p renders glioblastoma (GBM) cells resistant to TMZ, while inducing the expression of ING-4 rescued the effects of miR-423-5p. (C) U87 and U251 cells were transfected with miR-423-5p or miR-NC and cultured in 100 μM TMZ. The cells were then determined by apoptosis analysis using flow cytometry 48 hours later. (D) Western blotting. (E) Western blot analysis was performed to assess the regulation of apoptosis-related proteins p53, BCL-2, and Bax by miR-423-5p in U87 cells cultured in 100 μM TMZ. *indicates significant difference compared with control, #indicates significant difference compared with miR-423-5p/ING-4 treatment at P < .01. **P < .01 and ***P < .001.

Fig. 6

miR-423-5p promotes tumorigenicity, angiogenesis, and invasion in vivo. (A) Representative images of mice implanted with intracranial tumors formed by U87 on days 1, 5, 10, 15, 20, 25, and 30. Plot of the Fluc activity by bioluminescence imaging for intracranial tumors. (B) Kaplan–Meier survival curve of animals injected with glioblastoma (GBM) tumor cells U87 transfected with miR-423-5p (N = 10) or miR-NC (N = 10). (C) Cell proliferation and angiogenesis in vivo were determined by Ki-67 and CD31, respectively, using immunohistochemistry. Scale bar = 20 μm. (D) Invasiveness in vivo was determined by hematoxylin & eosin stain. Scale bar = 50 μm. (E) The expression levels of pathway proteins and of EGFR or PDGFR in the miR-423-5p–overexpressing tumor tissues were significantly higher than were those of the miR-NC group according to Western blotting. *P < .05 and **P < .01.