miR-5188 augments glioma growth, migration and invasion through an SP1-modulated FOXO1-PI3K/AKT-c-JUN-positive feedback circuit

Abstract

The biological effect and molecular mechanism of miR-5188 have not been thoroughly investigated. The study aims at elucidating the role of miR-5188 in glioma progression. Human glioma cell lines and tissues were used for functional and expression analysis. Cellular and molecular techniques were performed to explore the functions and mechanisms of miR-5188 in glioma. In our investigation, we demonstrated that miR-5188 promoted cell proliferation, the G1/S transition of the cell cycle, migration and invasion in glioma and reduced the lifespan of glioma-bearing mice. miR-5188 directly targeted FOXO1 and activated PI3K/AKT-c-JUN signalling, which enhanced miR-5188 expression. Moreover, the c-JUN transcription factor functionally bound to the miR-5188 promoter region, forming the positive feedback loop. The feedback loop promoted glioma progression through activating the PI3K/AKT signalling, and this loop is augmented by the interaction between SP1 and c-JUN. Moreover, it was also found that the miR-5188/FOXO1 axis is facilitated by SP1-activated PI3K/AKT/c-JUN signalling. In glioma samples, miR-5188 expression was found to be an unfavourable factor and was positively associated with the mRNA levels of SP1 and c-JUN, whereas negatively associated with the mRNA levels of FOXO1. Our investigation demonstrates that miR-5188 could function as a tumour promoter by directly targeting FOXO1 and participating in SP1-mediated promotion of cell growth and tumorigenesis in glioma.

Keywords: FOXO1; SP1; c-JUN; glioma; miR-5188.

FIGURE 1

miR‐5188 is elevated in glioma tissues and promotes glioma cell proliferation, migration and invasion in vitro by up‐regulating PI3K/AKT activity. A, Increased levels of miR‐5188, as assessed using real‐time quantitative PCR (RT‐qPCR), were positively correlated with the pathology classification status. An unpaired t test was applied for assessing the results of this assay in normal tissues (n = 17), grade II glioma (n = 12), grade III glioma (n = 13) and grade IV glioma (n = 15). An MTT assay (B), colony formation assay (C), cell cycle analysis (D) and EdU incorporation assay (E) were applied to investigate the impact of miR‐5188 on cell proliferation in U87 and U251. Scale bars, 100 μm. F, A transwell assay and Boyden chamber assays were applied to elucidate the impact of miR‐5188 on cell migration and invasion in U87 and U251. Scale bars, 100 μm. G, Western blotting analyses were applied to assess P‐PI3K, PI3K, P‐AKT, AKT, CDK4, CCND1, c‐JUN, N‐cadherin and vimentin protein levels in U87 and U251 cells treated with miR‐5188 inhibitor or mimics. *P < .05; **P < .01; ***P < .001

FIGURE 2

miR‐5188 augments glioma cell growth in vivo. A, Tumorigenicity of U87 cells was increased in miR‐5188‐overexpressing group in comparison with the control group. B, The growth curves of the tumours in nude mice were plotted to explore the impact of miR‐5188 on glioma growth. C, Representative images showing the H&E staining of primary glioma tissues and immunohistochemical analysis of Ki67 and PCNA expression. Scale bars, 125 μm.. D, RT‐qPCR analyses were applied to measure miR‐5188 expression levels in xenografts. E, Survival times calculated through a Kaplan‐Meier analysis revealed that miR‐5188 overexpression (miR‐5188) reduced survival time compared with untreated controls (Mock). **P < .01

FIGURE 3

miR‐5188 down‐regulates FOXO1 through directly targeting FOXO1. A, miR‐5188 and its binding sites in FOXO1 3′UTR. A mutation was constructed in the sequences targeted by miR‐5188 seed region. B, FOXO1 mRNA levels were measured through RT‐qPCR in miR‐5188‐silenced U87 and miR‐5188‐overexpressing U251 cells and normalized to ARF5. C, Western blotting analyses were applied to investigate FOXO1 protein levels in miR‐5188‐silenced U87 and miR‐5188‐overexpressing U251 cells. D, A luciferase reporter assays were applied to elucidate the direct targeting of FOXO1 3′UTR by miR‐5188. E, RNA immunoprecipitation was conducted to confirm the interaction between AGO2‐bound miR‐5188 and FOXO1 mRNA. F, FOXO1 protein levels were measured through immunohistochemistry in xenografts. Scale bars, 125 μm. *P < .05; **P < .01; ***P < .001

FIGURE 4

Silence of FOXO1 mitigates the inhibitory effects of miR‐5188 knockdown on glioma cell proliferation, invasion and migration. MTT assays (A), EdU incorporation assays (B), flow cytometry assays (C), transwell assays (D) and Boyden assays (E) were performed in U251 or U87 cells treated with control sequence, FOXO1 siRNA or miR‐5188 inhibitor, as indicated. Scale bars, 100 μm. F, Western blotting analyses were applied to measure the protein levels of FOXO1, P‐PI3K, PI3K, P‐AKT, AKT, CCND1, CDK4, c‐JUN, N‐cadherin and vimentin in U251 or U87 cells treated with control sequence, FOXO1 siRNA or miR‐5188 inhibitor, as indicated. G, Western blotting analysis analyses were applied to measure the protein levels of FOXO1, P‐PI3K, PI3K, P‐AKT, AKT, CCND1, CDK4, c‐JUN, N‐cadherin and vimentin protein levels in U87 and U251 cells transfected with control sequence or FOXO1 siRNA. *P < .05; **P < .01; ***P < .001

FIGURE 5

c‐JUN positively enhances miR‐5188 transcription by binding to its promoter region. A, Schematic diagram of the c‐JUN binding sites (site A and site B) inside miR‐5188 promoter regions. B, miR‐5188 levels were measured through RT‐qPCR in c‐JUN‐silenced U87 or U251 cells. C, RT‐qPCR and RT‐PCR analyses were applied to amplify c‐JUN‐binding sites A and B enriched from Chromatin immunoprecipitation assay (ChIP) using anti‐c‐JUN. D, Western blotting analysis analyses were applied to measure the protein levels of c‐JUN, P‐PI3K, P‐AKT, CDK4, CCND1 and FOXO1 expression after c‐JUN suppression. E, ChIP assays were conducted to investigate the impact of FOXO1 and PI3K signalling on c‐JUN‐augmented miR‐5188 transcription in glioma cells. *P < .05; **P < .01; ***P < .001

FIGURE 6

SP1 facilitates miR‐5188 expression in glioma. A, miR‐5188 levels were assessed through RT‐qPCR in U87 or U251 cells transfected with the control vector or SP1 plasmid, as indicated. MTT assays (B) and EdU incorporation assays (C) were performed in U87 and U251 cells treated with SP1 plasmid or both SP1 plasmid and miR‐5188 inhibitor. Scale bars, 100 μm. D, Western blotting analysis analyses were applied to measure the protein levels of P‐PI3K, P‐AKT, CCND1, CDK4, c‐JUN and FOXO1 in U87 and U251 cells treated with SP1 plasmid and miR‐5188 inhibitor. E, Co‐immunoprecipitation (Co‐IP) was applied for detecting the interaction between endogenous SP1 and c‐JUN in U87 and U251 cells. F, Immunofluorescence staining was applied to assess the nuclear colocalization of SP1 and c‐JUN protein in U87 and U251 cells. Scale bars, 25 μm. G, Western blotting analyses were applied to measure the protein levels of c‐JUN in U87 and U251 cells with SP1 knockdown. **P < .01; ***P < .001

FIGURE 7

The correlations between pathoclinical characteristics, miR‐5188 expression and its associated genes. A, miR‐5188 expression was elevated in glioma in comparison with non‐tumour brain tissues. (i) Weak staining of miR‐5188 in non‐tumour brain tissues; (ii) strong staining of miR‐5188 in non‐tumour brain tissues; (iii) negative staining of miR‐5188 in glioma tissues; and (iv) strong staining of miR‐5188 in glioma tissues (insets showing miR‐5188 staining photographed under high magnification). Scale bars, 100 μm. B, Kaplan‐Meier survival analysis was applied to correlate miR‐5188 levels and overall survival in 148 glioma patients. C, The mRNA levels of FOXO1, c‐JUN and SP1 were measured through RT‐qPCR in glioma (n = 40) and non‐tumour (n = 17) brain tissues, normalized to ARF5. P = .0049; P < .0001; P < .0001, respectively. D, Correlations between miR‐5188, FOXO1, c‐JUN, and SP1 levels were explored in glioma tissues (n = 40). P = .0329; P = .0003; P = .0028, respectively