miR‑4286 is Involved in Connections Between IGF‑1 and TGF‑β Signaling for the Mesenchymal Transition and Invasion by Glioblastomas
Kuo‑Hao Ho1,2 · Peng‑Hsu Chen1,2 · Chwen‑Ming Shih · Yi‑Ting Lee1,2 · Chia‑Hsiung Cheng1,2 · Ann‑Jeng Liu3 · Chin‑Cheng Lee4 · Ku‑Chung Chen1,2
Abstract
The insulin-like growth factor (IGF)-1 and transforming growth factor (TGF)-β signal pathways are both recognized as important in regulating cancer prognosis, such as the epithelial-to-mesenchymal transition (EMT) and cell invasion. However, cross-talk between these two signal pathways in glioblastoma multiforme (GBM) is still unclear. In the present study, by analyzing data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GSE) 4412, GBM patients with higher IGF-1 levels exhibited poorer survival. Genes positively correlated with IGF-1 were enriched in EMT and TGF-β signal pathways. IGF-1 treatment enhanced mesenchymal marker expressions and GBM cell invasion. A significant positive correlation was observed for IGF-1 with TGF-β1 (TGFB1) or TGF-β receptor 2 (TGFBR2), both of which participate in TGF-β signaling and are risk genes in the GBM process. IGF-1 stimulation promoted both TGFB1 and TGFBR2 expressions. LY2157299, a TGF-β signaling inhibitor, attenuated IGF-1-enhanced GBM cell invasion and mesenchymal transition. By analyzing IGF-1-regulated microRNA (miR) profiles, miR-4286 was found to be significantly downregulated in IGF-1-treated cells and could be targeted to both TGFB1 and TGFBR2. Overexpression of miR-4286 significantly attenuated expressions of the IGF-1-mediated mesenchymal markers, TGFB1 and TGFBR2. Using kinase inhibitors, only U0126 treatment showed an inhibitory effect on IGF-1-reduced miR-4286 and IGF-1-induced TGFB1/TGFBR2 expressions, suggesting that MEK/ ERK signaling is involved in the IGF-1/miR-4286/TGF-β signaling axis. Finally, our results suggested that miR-4286 might act as a tumor suppressive microRNA in inhibiting IGF-1-enhanced GBM cell invasion. In conclusion, IGF-1 is connected to TGF-β signaling in regulating the mesenchymal transition and cell invasion of GBM through inhibition of miR-4286. Our findings provide new directions and mechanisms for exploring GBM progression.
Keywords IGF-1 · TGF-β · TGFBR2 · miR-4286 · Epithelial-to-mesenchymal transition (EMT)
Introduction
Several lines of evidence show that upregulated insulin-like growth factor (IGF) levels may act as putative diagnostic markers in different tumors, especially in malignant brain tumors (Zumkeller and Westphal 2001). Glioblastoma multiforme (GBM), the most common malignant primary brain tumor of the central nervous system, accounts for a 60% prevalence of grade IV gliomas. Due to it being highly mobile and invasive, the prognosis of GBM patients is poor with a median survival of 8–10 months (Urbanska et al. 2014). IGF-1, also called somatomedin C, modulates malignant cellular behavior and plays a crucial role in invasion by GBM (Guvakova 2007; Samani et al. 2007). Clarifying the detailed mechanisms of IGF-1-induced GBM cell invasion would be helpful for researching GBM progression.
The epithelial-to-mesenchymal transition (EMT) plays a critical roles in regulating the migration and invasiveness of neoplastic cells for metastasis (Cevenini et al. 2018). Previous studies reported that IGF-1 induces the EMT through activating surviving of hepatocellular carcinoma (Liu et al. 2018b) and gastric cancer cells (Li et al. 2016). The phosphatidylinositol 3-kinase (PI3K)/Akt-glycogen synthase kinase (GSK)-3β/zinc finger E-box-binding homeobox 2 (ZEB2) signaling pathway is also involved in the IGF-1-induced EMT of gastric cancer cells (Li et al. 2015). Activated NRD1/a disintegrin and metalloproteinase (ADAM) signaling pathway regulates the lipogenesis-mediated EMT in IGF-1-stimulated colon cancer cells (Park and Kim 2018). In GBM, IGF-1 treatment can upregulate mesenchymal marker expressions, leading to generation of a fibroblastic phenotype (Lin et al. 2014). However, the connection between IGF-1 signaling and the EMT in GBM is still unclear.
In addition to IGF-1, several growth factors are also known to participate in GBM progression, such as the transforming growth factor (TGF)-β (TGFB) (Kang and Massague 2004; Bryukhovetskiy and Shevchenko 2016). Higher expression levels of TGF-β around necrotic regions of GBM were significantly correlated with patient survival (Iwadate et al. 2016). TGF-β1 (TGFB1), the dominant isoform in newly diagnosed GBM, but not TGF-β2 (Roy et al. 2018), is vital for regulating GBM cell proliferation and invasion. TGF-β1 induces the EMT of GBM cells through activating Smad-dependent signaling (Song et al. 2019) or the phosphoinositide-dependent kinase 1 (PDK1)/c-Jun pathway (Luo et al. 2018). Since the enhancer of zeste homolog 2 (EZH2)/ AXL/TGF-β axis is relevant to GBM metastasis, it was suggested to be a novel therapeutic target for GBM (Liu et al. 2018a). The mechanisms for regulating TGF-β1 expression in the GBM process remain unexplored.
Previous studies suggested that cross-talk between IGF-1 and TGF-β signals mediates several cellular processes. For example, TGF-β can interact with IGF-enhanced signaling in coordinating cellular dynamics for ear cell morphogenesis and differentiation (Gibaja et al. 2019). TGF-β can also enhance IGF-1 gene promoter activity for osteoblast cell growth (McCarthy and Centrella 2015). Although both IGF-1 and TGF-β are recognized as critical regulators in carcinogenesis, their roles in jointly regulating the EMT of cancers remain unclear, especially in GBM disease. Our study showed that IGF-1 enhanced the EMT and cell invasion through upregulating TGF-β signals in GBM. Furthermore, TGFB1 and TGF-β receptor 2 (TGFBR2) in TGF-β signal pathways were targeted and inhibited by microRNA (miR)-4286, which was downregulated by IGF-1 stimulation in GBM cells. Finally, we concluded that miR-4286 is involved in connections between IGF-1 and TGF-β signaling for the mesenchymal transition and invasion of GBM.
Materials and Methods
Chemicals and Reagents
Human glioblastoma U87-MG and T98G cells were obtained from the Bioresource Collection and Research Center (Hsinchu City, Taiwan). The primary antibodies for detecting N-cadherin, vimentin, TGFB1, TGFBR2, and β-actin were obtained from GeneTex (Hsinchu City, Taiwan). LY2157299 (cat. no. S2230) and temsirolimus (cat. no. CCI-779) were obtained from Selleck Chemicals (Houston, TX, USA). 3-[4, 5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide (MTT) (cat. no. M2128), LY294002 (cat. no. L9908), and U0126 (cat. no. U120) were obtained from Sigma-Aldrich (St. Louis, MO, USA). An enhanced chemiluminescence (ECL) solution (cat. no. WBKLS0500) and polyvinylidene difluoride (PVDF) membranes (cat. no. 539131) were obtained from Millipore (Billerica, MA, USA). Lipofectamine 3000 (cat. no. L3000015), human IGF-1 recombinant protein (cat. no. PHG0071), Trizol® reagent (cat. no. 15596026), and secondary antibodies were obtained from Invitrogen (ThermoFisher Scientific, Waltham, MA, USA). The MultiScribe ™ Reverse Transcriptase Kit (cat. no. N8080234), SYBR® Green Polymerase Chain Reaction (PCR) Master Mix (cat. no. 4309155), TaqMan® miR-4286, and the U6 assay (cat. no. 4427975) were obtained from Applied Biosystems (ThermoFisher Scientific). The dual-luciferase reporter assay system (cat. no. E1910) was obtained from Promega (Madison, WI, USA). All the primer for PCR reaction were produced by Genomics BioSci & Tech (Xizhi, New Taipei City, Taiwan). Unless otherwise specified, all other reagents were of analytical grade.
Cell Culture, Treatments, and Transfection
The Dulbecco’s modified Eagle medium (DMEM, GIBCOBRL; Grand Island, NY, USA) supplemented with 10% fetal bovine serum (Biological Industries, Cromwell, CT, USA), 100 units/ml penicillin, 100 μg/ml streptomycin, 1 mM sodium pyruvate, and 1 mM non-essential amino acids was used to maintain U87-MG and T98G cells All cells were at 37 °C in a 5% CO2 incubator. For drug treatment, overnight-cultured cells were treated with 100, 200 ng/ml, or indicated doses of IGF-1 for the indicated time. For inhibitor treatment, cells were, respectively, treated with 10 μM U0126, 10 μM LY2157299, 10 μM temsirolimus, and 5 μM LY294002 for 1 h, and then 200 ng/ml IGF-1 was added for another 48 h. For the transfection experiments, 105 cells/well were seeded into 12-well plates. After overnight culture, indicated doses of pCDH-miR-4286, 500 ng pmiRGLO-3′ untranslated region (UTR) and mutant plasmids were transfected with lipofectamine 3000 according to the manufacturer’s instructions, respectively. After incubation for 24 h, cells were collected for further experiments.
Cell Viability Assay
The detail method for MTT assays was described in our previous study. After indicated dose of pCDH-miR-4286 and empty pCDH vectors were transfected for 48 h, the MTT assay was conducted.
Immunoblotting Analyses
The detail method for immunoblotting analyses was described in our previous study. The diluted ratio of 1:1000 and 1:5000 with PBS-T buffer were, respectively, used to prepare the primary and secondary antibodies. The antibody-protein complexes were detected by using an ECL non-radioactive detection system.
Pan Cancer, Gene Set Enrichment Analysis (GSEA), and Survival Rate Analysis
The miRNA expression profiling (GSE140297) was obtained as in our previous study (Chen et al. 2019). The miRwalk (Dweep and Gretz 2015) and TargetScan7.2 (Agarwal et al. 2015) were used to predict all target genes of miR-4286. For the pan-cancer analysis, miR-4286 expression levels were obtained from YM522v3 database (Chung et al. 2017). For conducting the GSEA, fgsea R package was used to run the GSEA algorithm with 1000 permutations. Based on their multiples of change in different phenotypes, genes were ranked with hallmark database. Pathways were considered significant as the false detection rate (FDR) of < 0.05. For patient survival analyses, The Cancer Genome Atlas (TCGA) array data and GSE4412 were analyzed using the SurvExpress platform (Aguirre-Gamboa et al. 2013). All probe sets were averaged per sample. According to distinct survival rates, the groups with high and low risk were, respectively, categorized.
Matrigel Invasion Assays
The detail method for matrigel invasion assays was described in our previous study (Ho et al. 2018). After, respectively, treating with IGF-1, LY2157299, or transfected with the indicated overexpressing plasmids, the matrigel invasion assays were conducted by using transwell inserts (25-mm polycarbonate membrane, 8-µm pore size; cat. no. PI8P01250; Millipore) coated with Matrigel matrix (cat. no. 356234; Corning, Corning, NY, USA). At least four fields by three researchers with a microscope at 100 × magnification. Experiments were assayed in triplicate.
RNA Extraction and Quantification
The detail methods for RNA extraction and real-time reverse transcription quantitative PCR were described in our previous study (Chen et al. 2019). The primers for detecting TGFB1, TGFBR2, and GAPDH were listed in Supplemental Table S1. The TaqMan microRNA assays (Applied Biosystems) were used to detecting miR-4286 and U6B. The levels of GAPDH and U6B were, respectively, used as internal controls for quantifying mRNA and miR-4286 expression.
Construction of the Plasmids and 3′ UTR Reporter Assays
The detail methods for gene or 3′UTR cloning and 3′UTR reporter assays were described in our previous study (Chen et al. 2019). The DNA fragments containing TGFB1 or TGFBR2 3′UTRs were digested with restriction enzyme XhoI/XbaI, and cloned into downstream of the luciferase gene of pMIRGLO-REPORT luciferase vector (Promega). The PCR products for cloning miR-4286 gene were digested with BamHI/EcoRI enzymes, and the DNA fragments were directly cloned into pCDH vectors (System Biosciences, Palo Alto, CA, USA). PCR primers are listed in Supplemental Table S1. The overlapping PCR was used to create mutagenesis of the miR-4286 target site in the TGFB1 and TGFBR2 3′UTRs.
Statistical Analysis
All data are calculated as the mean ± standard deviation (SD) at least three independent experiments. The unpaired Student’s t-test was used to determine significant differences among groups. Pearson correlation was used to analyze the association between two genes. The log rank test was used to evaluate the correlation between gene expression and patient survival. A value of p < 0.05 was recognized as statistical significance.
Results
Poor Survival‑Associated IGF‑1‑Induced Glioma Cell Invasion and EMT Signaling
To investigate the role of IGF-1 in GBM tumorigenesis, we first analyzed the association between IGF-1 levels and patient survival rates using data in TCGA and GSE4412 databases. We observed a significant association between high levels of IGF-1 and a higher risk of poor survival of patients with a glioma (Fig. 1a, b). Next, to explore IGF1-mediated signaling pathways in GBM tumorigenesis, we analyzed IGF-1-correlated genes using TCGA database. As shown in Fig. 1c, there were 1360 positive genes (r > 0.2 and FDR ≤ 0.05) and 309 negative genes (r < − 0.2 and FDR ≤ 0.05) in TCGA glioma patients. Furthermore, using GSEA analyses, we found that IGF-1 positively correlated genes were significantly correlated with EMT signaling (Fig. 1d, Supplemental Table S2). According to Rohrmann et al. study (Rohrmann et al. 2011), the serum concentration of IGF-1 is variant in patients with brain tumor between 0 and 400 ng/ml. Therefore, 100 and 200 ng/ml of IGF-1 concentrations were used in the present study. To confirm the role of IGF-1 on EMT regulation, we quantified the protein levels of N-cadherin and vimentin after treatment of U87-MG and T98G glioma cells with different doses of IGF1. IGF-1 increased N-cadherin and vimentin protein expressions in both U87-MG and T98G cells (Fig. 1e). Moreover, to explore the effects of IGF-1 on glioma cell invasion with Matrigel invasion assays, we found that IGF-1 significantly increased GBM cell invasiveness (Fig. 1f, g). These data suggested that IGF-1 was associated with a poor prognosis and induced glioma cell invasion and EMT signaling.
TGFB1 and TGFBR2 Genes, Highly Correlated with IGF‑1, are Associated with Poor Prognosis in GBM Patients
Previous studies reported that IGF-1 signaling was mediated by TGF-β signaling (Danielpour and Song 2006). Interestingly, using GSEAs, we found that IGF-1 positively correlated genes were also significantly enriched in TGF-β signaling (Fig. 2a, Supplemental Table S2). Moreover, we analyzed correlations between IGF-1 and TGF-β family gene expressions using TCGA with Pearson correlation, and most gene expressions in the TGF-β family were significantly correlated with IGF-1 (Fig. 2b), especially the highly correlated genes of TGFB1 (r = 0.39, p < 0.01) (Fig. 2c) and TGFBR2 (r = 0.53, p < 0.01) (Fig. 2d). Furthermore, using data in TCGA and GSE4412 databases, we found that GBM patients with higher levels of TGFB1 or TGFBR2 exhibited poor survival (Fig. 2e-h). Taken together, our results suggested that TGF-β signaling is involved in IGF-1-associated poor prognosis of GBM.
IGF‑1 Upregulates Glioma Cell Invasion and the EMT Through TGF‑β Signaling
To investigate whether TGF-β signaling participates in IGF-1 signaling, we first tested the effects of IGF-1 on TGFB1 and TGFBR2 expressions. Using a real-time PCR to measure mRNA levels of TGFB1 and TGFBR2, we found that both genes were upregulated in IGF-1-treated glioma cells (Fig. 3a, b). Similarly, IGF-1-increased TGFB1 and TGFBR2 protein expressions were also observed in glioma cells (Fig. 3c). IGF-1 treatment also stimulated TGFB1mediated downstream Smad3 activation. Furthermore, treatment with 10 µM of the TGF-β inhibitor, LY2157299 significantly attenuated IGF-1-induced EMT activation and cell invasion (Fig. 3d–f). These data indicated that TGF-β signaling is involved in IGF-1-induced EMT and cell invasion.
Inhibition of miR‑4286 is Involved in IGF‑1‑Connected TGF‑β Signaling
miRNAs are recognized as crucial mediators for gene regulations (Tsunoda et al. 2014). Furthermore, in our previous study (Chen et al. 2019), we established a miRNA expression profile with IGF-1-treated U87-MG cells as GSE140297. We proposed that miRNAs might participate in IGF-1-mediated TGFB1 and TGFBR2 expressions. By combined analyses with the IGF-1-downregulated miRNA signature and target gene prediction algorithms (Fig. 4a), two miRNAs, miR-4286 and miR-1260a, were selected. Since miR-4286 expression showed the greatest downregulation in IGF-1-treated cells, it was selected as a candidate for the present study (Fig. 4b). To confirm the array data, we measured changes in miR-4286 expression in IGF-1-treated U87-MG and M059K cells (Fig. 4c). We found that IGF-1 stimulation significantly reduced miR-4286 expression in both cell lines. ERK- and PI3K/mTOR-mediated signaling belong to the main downstream pathway regulated by IGF-1 signaling (Weroha and Haluska 2012). Herein, we explored which pathway exhibited a dominant effect in IGF-1 -inhibited miR-4286 expression. After U87-MG cells were, respectively, co-treated with IGF-1 and U0126 (an ERK inhibitor), LY294002 (a PI3K inhibitor), or temsirolimus (an mTOR inhibitor) for 48 h, relative miR-4286 levels were measured. As shown in Fig. 4d, only U0126 showed inhibitory effects on IGF-1-repressed miR-4286 levels, but not LY294002 or temsirolimus. Similarly, the effects of U0126 on IGF-1-repressed miR-4286 expression was also observed in T98G cells (Fig. 4e). In addition, inhibition of ERK signaling attenuated IGF-1-increased TGFB1 and TGFBR2 expressions at both the mRNA and protein levels (Fig. 4f, g). Taken together, our results suggest that IGF-1 repressed miR-4286 expression via an ERK signaling pathway, leading to enhanced TGF-β signaling in GBM.
Identification of TGFB1 and TGFBR2 as Direct Target Genes of miR‑4286
Using the TargetScan 7.2 prediction (Agarwal et al. 2015), TGFB1 and TGFBR2 were predicted to be putative target genes of miR-4286 (Fig. 5a). There are two putative miR4286 binding sites inTGFB1 3′UTR. To further confirm the prediction results, the 3′UTRs of the TGFB1 and TGFBR2 genes containing a miR-4286-binding site were, respectively, cloned into the pmiRGlo reporter plasmids to conduct 3′UTR reporter assays. As shown in Fig. 5b, overexpression of miR-4286 significantly decreased luciferase activities. To further validate that TGFB1 and TGFBR2 expressions were inhibited by miR-4286 via binding to the 3′UTR, five nucleotides located in the critical binding regions of the 3′UTR of the TGFB1 and TGFBR2 genes were mutated by overlapping PCR (Fig. 5a). As shown in Fig. 5b, miR-4286 had no effects on luciferase activity after mutating the miR-4286-targeted site on TGFB1 3U1 and TGFBR2 3UTR, but not TGFB1 3U2. Furthermore, overexpression of miR-4286 significantly reduced mRNA and protein levels of TGFB1 and TGFBR2 (Fig. 5c, d). miR-4286 also decreased activated Smad3 levels. All these findings suggested that IGF-1 enhanced TGFB1 and TGFBR2 through inhibition of miR-4286.
Overexpression of miR‑4286 Attenuates IGF‑1 Functions in GBM
To investigate the expression of miR-4286 during the malignant transformation and development of gliomas, the miRNA expression profile from GSE41033 was used (Fig. 6a). Although miR-4286 levels exhibited no significant difference in normal human neural stem cells (NSCs) and human glioma stem cells (GSCs) due to too small of a sample size, a trend of lower miR-4286 levels in GSCs was observed. We further measured the miR-4286 endogenous levels in normal human astrocyte (NHA) and four GBM cell lines (Fig. 6b). We found that lower miR-4286 levels were observed in GBM cells than that in NHA. Furthermore, to study levels of miR-4286 in various cancers, we found that miR-4286 was downregulated in more than half of the cancer types we examined compared to normal tissues (Fig. 6c). We then analyzed the association between miR-4286 expression and patient survival rates using TCGA data. In head-neck squamous cell carcinoma (HNSC), lung adenocarcinoma (LUAD), pancreatic adenocarcinoma (PAAD), and uterine corpus endometrial carcinoma (UCEC), patients with lower miR-4286 levels exhibited poorer survival rates (Supplemental Fig. 1). To explore the role of miR-4286 in GBM, overexpression of miR-4286 in both U87-MG and T98G cells produced no changes in cell viability (Fig. 6d). However, elevated miR-4286 significantly attenuated IGF-1-induced glioma cell invasion, the mesenchymal transition, and TGFB/ TGFBR2 expressions (Fig. 6e–i). Taken together, our results suggest that miR-4286 may act in a tumor suppressive role in inhibiting and connecting IGF-1 and TGF-β signaling in GBM.
Discussion
Abnormal cytokine expression in the GBM microenvironment was implicated in the prognosis, therapy, and recurrence (Zhu et al. 2012). EMT and cell invasion, regulated by cytokines, are highly associated with aggressive GBM, and are recognized as therapeutic targets (Liu et al. 2018a; Fedele et al. 2019). However, the connection between different signaling pathways in regulating the EMT and invasiveness of GBM still remains unclear. In this study, we found that IGF-1 treatment enhanced glioma EMT and cell invasion through inducing miR-4286-reduced TGF-β signaling. Using data mining, we identified that IGF-1 expression levels were shown to be positively correlated with TGF-β signaling. Furthermore, inhibition of TGF-β signaling attenuated IGF-1-induced glioma cell invasion and the EMT. In addition, our results indicated that miR4286 was downregulated in IGF-1-treated glioma cells.
TGFB1 and TGFBR2, both critical regulators for TGF-β signaling, were identified as direct target genes of miR4286. Overexpression of miR-4286 significantly repressed TGFB1 and TGFBR2 expressions, resulting in attenuation of IGF-1-induced glioma invasion and EMT signaling. All these findings indicate that IGF-1 enhanced TGF-β signaling-regulated EMT and cell invasion through suppressing miR-4286 expression in GBM.
Several studies reported that TGF-β signaling regulates physiological mechanisms through activating IGF1-mediated pathways. For example, TGF-β-enhanced IGF1-mediated signaling was implicated in ovarian tumor cell proliferation (Alsina-Sanchis et al. 2016). TGF-β regulates osteoblasts in bone growth through increasing IGF-1 promoter activity (McCarthy and Centrella 2015). Activation of IGF-1-mediated signaling is essential to support TGFβ-driven profibrotic myofibroblast functions in idiopathic pulmonary fibrosis (Hernandez et al. 2020). In addition, Walsh et al. identify that IGF-1 treatment induces activation of latent TGF-β1 resulting in increasing invasion and EMT of breast cancer (Walsh and Damjanovski 2011), suggesting that a cross-talk may exist between IGF-1 and TGF-β signaling. In the present study, we found that IGF-1 expression was highly correlated with the EMT and TGF-β signaling in GBM patients. We identified that IGF-1 treatment significantly increased TGFB1 and TGFBR2 expression levels by inhibiting miR-4286, resulting in activation of TGF-β signaling. Using the TGF-β signaling inhibitor, LY2157299, TGF-β signaling was implied to be involved in IGF-1-mediated downstream signaling, suggesting that IGF-1 is connected to TGF-β signaling in the GBM process.
Previous studies show that relationships between TGF-β signaling and miRNA regulation are involved in mediating biological processes and tumorigenesis. For example, TGFβ1 enhanced miR-10b upregulation in regulating GBM cell proliferation, migration, and the EMT (Ma et al. 2017). miR181a-5p directly targets TGFBR1 and influences thymic epithelial cell proliferation via inhibition of TGF-β signaling (Guo et al. 2016). TGF-β1 enhances the miR-424(322)/503 cluster in regulating epithelial tissue remodeling of mammary glands (Llobet-Navas et al. 2014). In the present study, we found that miR-4286 levels were significantly reduced in GBM cells upon IGF-1 treatment via activating ERK signaling. By targeting and inhibiting TGFB1 and TGFBR2 expressions, miR-4286 significantly inhibited IGF-1-mediated EMT signaling and invasion. Consequently, miR-4286 plays a critical role in regulating IGF-1 and TGF-β signaling pathways in GBM. MEK–ERK pathway is recognized as one of the IGF1-mediated downstream signalings (Menu et al. 2004), and involves in mediating GBM progression (Lo 2010).
Through phosphorylation activating, ERK enhances downstream transcription factor (TF)s activation such as activator protein (AP)-1 formed by Fos and Jun, CREB2, and Myc, resulting in regulating gene expression for cell growth, migration, and differentiation (Whitmarsh 2007). In addition to be the transcriptional activators, these TFs also act as repressors in inhibiting gene expression. For examples, AP-1 transcriptionally inhibits estrogen-enhanced progesterone receptor gene expression through binding to AP-1 site (Petz et al. 2004), and blocks transcriptional activity of the nuclear receptor steroidogenic factor 1 (Sirianni et al. 2010). Myc and CREB2 have also been reported to negatively and transcriptionally control gene expression (Grandori et al. 2000; Karpinski et al. 1992). In the present study, by treating with a MEK1/2 Inhibitor U0126, we found that inhibition of MEK–ERK pathway significantly attenuated IGF-1-inhibited miR-4286 expression, suggesting that IGF1-induced MEK–ERK pathway activation is responsible for repressing miR-4286 gene expression. Furthermore, by using bioinformatic tools for TFs prediction, the promoter region (− 3000 bp) of miR-4286 gene contains several AP-1 and Myc binding sites, suggesting that miR-4286 gene expression can be regulated by these TFs. However, which TF involves in IGF-1/ERK-inhibiting miR-4286 expression needs further investigations in the future.
In the past, few studies investigated the role of miR-4286 in tumor progression. Only one study indicated that miR4286 acted as an onco-miRNA in non-small cell lung cancer (Ling et al. 2019). In contrast, we found an opposite role of miR-4286 in GBM. We suppose that miR-4286 may regulate distinct gene expressions in different cells to modulate the responsiveness of cells to external signals, leading to different cell fate determinations. Additionally, we also found that miR-4286 was downregulated in several cancers, such as COAD, KIRC, PAAD, and READ. Low expression levels of miR-4286 were associated with a high risk of a poor survival rate of patients with HNSC, LUAD, PAAD, and UCEC. These findings indicated that miR-4286 may act as a tumor suppressor in these cancers. More future investigations are required to clarify the detailed molecular mechanisms of miR-4286 in cancers.
In summary, we observed that IGF-1 induced TGF-β and EMT signaling activation, resulting in increasing glioma invasion. We also identified that miR-4286 is involved in cross-talk between the IGF-1 and TGF-β signaling pathways. In addition to TGF-β signaling, IGF-1 is also significantly correlated with TGFB3 and TGFBR3 (Fig. 2b). A study suggested that TGFB3 and TGFBR3 induced downstream SMAD2 and SMAD1/5 phosphorylation for promoting glioma invasion (Seystahl et al. 2017). However, no studies have reported their roles in IGF-1-induced invasion. These should be further investigated in the future. Some limitations exist in this study. In vivo experiments and additional GBM patient tissues are needed for data validation. The detailed mechanisms of IGF-1-reduced miR-4286 expression still need to be further investigated. In conclusion, we suggest that IGF-1-connected TGF-β signaling plays an important role in regulating the GBM microenvironment through influencing miR-4286 expression. Our findings may provide novel mechanisms and directions for investigating GBM diseases.
References
Agarwal V, Bell GW, Nam JW, Bartel DP (2015) Predicting effective microRNA target sites in mammalian mRNAs. eLife. https ://doi. org/10.7554/eLife .05005
Aguirre-Gamboa R, Gomez-Rueda H, Martinez-Ledesma E, Martinez-Torteya A, Chacolla-Huaringa R, Rodriguez-Barrientos A, Tamez-Pena JG, Trevino V (2013) SurvExpress: an online biomarker validation tool and database for cancer gene expression data using survival analysis. PLoS ONE 8(9):e74250. https: //doi. org/10.1371/journ al.pone.00742 50
Alsina-Sanchis E, Figueras A, Lahiguera A, Vidal A, Casanovas O, Graupera M, Villanueva A, Vinals F (2016) The TGFbeta pathway stimulates ovarian cancer cell proliferation by increasing IGF1R levels. Int J Cancer 139(8):1894–1903. https ://doi.org/10.1002/ ijc.30233
Bryukhovetskiy I, Shevchenko V (2016) Molecular mechanisms of the effect of TGF-beta1 on U87 human glioblastoma cells. Oncol Lett 12(2):1581–1590. https ://doi.org/10.3892/ol.2016.4756
Cevenini A, Orru S, Mancini A, Alfieri A, Buono P, Imperlini E (2018) Molecular signatures of the insulin-like growth factor 1-mediated epithelial-mesenchymal transition in breast, lung and gastric cancers. Int J Mol Sci. https ://doi.org/10.3390/ijms1 90824 11
Chen KC, Chen PH, Ho KH, Shih CM, Chou CM, Cheng CH, Lee CC (2019) IGF-1-enhanced miR-513a-5p signaling desensitizes glioma cells to temozolomide by targeting the NEDD4L-inhibited Wnt/β-catenin pathway. PLoS ONE 14(12):e0225913. https: //doi. org/10.1371/journ al.pone.02259 13
Chung IF, Chang SJ, Chen CY, Liu SH, Li CY, Chan CH, Shih CC, Cheng WC (2017) YM500v3: a database for small RNA sequencing in human cancer research. Nucleic Acids Res 45(D1):D925– d931. https ://doi.org/10.1093/nar/gkw10 84
Danielpour D, Song K (2006) Cross-talk between IGF-I and TGF-beta signaling pathways. Cytokine Growth Factor Rev 17(1–2):59–74. https ://doi.org/10.1016/j.cytog fr.2005.09.007
Dweep H, Gretz N (2015) miRWalk2.0: a comprehensive atlas of microRNA-target interactions. Nat Methods 12(8):697. https :// doi.org/10.1038/nmeth .3485
Fedele M, Cerchia L, Pegoraro S, Sgarra R, Manfioletti G (2019) Proneural-mesenchymal transition: phenotypic plasticity to acquire multitherapy resistance in glioblastoma. Int J Mol Sci. https ://doi.org/10.3390/ijms2 01127 46
Gibaja A, Aburto MR, Pulido S, Collado M, Hurle JM, Varela-Nieto I, Magarinos M (2019) TGFbeta2-induced senescence during early inner ear development. Sci Rep 9(1):5912. https ://doi. org/10.1038/s4159 8-019-42040 -0
Grandori C, Cowley SM, James LP, Eisenman RN (2000) The Myc/ Max/Mad network and the transcriptional control of cell behavior. Annu Rev Cell Dev Biol 16:653–699. https: //doi.org/10.1146/ annur ev.cellb io.16.1.653
Guo D, Ye Y, Qi J, Zhang L, Xu L, Tan X, Yu X, Liu Q, Liu J, Zhang Y, Ma Y, Li Y (2016) MicroRNA-181a-5p enhances cell proliferation in medullary thymic epithelial cells via regulating TGF-beta signaling. Acta Biochim Biophys Sin 48(9):840–849. https ://doi. org/10.1093/abbs/gmw06 8
Guvakova MA (2007) Insulin-like growth factors control cell migration in health and disease. Int J Biochem Cell Biol 39(5):890–909. https ://doi.org/10.1016/j.bioce l.2006.10.013
Hernandez DM, Kang JH, Choudhury M, Andrianifahanana M, Yin X, Limper AH, Leof EB (2020) IPF pathogenesis is dependent upon TGFbeta induction of IGF-1. FASEB J 34(4):5363–5388. https: // doi.org/10.1096/fj.20190 1719R R
Ho KH, Chang CK, Chen PH, Wang YJ, Chang WC, Chen KC (2018) miR-4725-3p targeting stromal interacting molecule 1 signaling is involved in xanthohumol inhibition of glioma cell invasion. J Neurochem 146(3):269–288. https ://doi.org/10.1111/jnc.14459
Iwadate Y, Matsutani T, Hirono S, Shinozaki N, Saeki N (2016) Transforming growth factor-beta and stem cell markers are highly expressed around necrotic areas in glioblastoma. J Neuro-Oncol 129(1):101–107. https ://doi.org/10.1007/s1106 0-016-2145-6
Kang Y, Massague J (2004) Epithelial-mesenchymal transitions: twist in development and metastasis. Cell 118(3):277–279. https ://doi. org/10.1016/j.cell.2004.07.011
Karpinski BA, Morle GD, Huggenvik J, Uhler MD, Leiden JM (1992) Molecular cloning of human CREB-2: an ATF/CREB transcription factor that can negatively regulate transcription from the cAMP response element. Proc Natl Acad Sci USA 89(11):4820– 4824. https ://doi.org/10.1073/pnas.89.11.4820
Li C, Li J, Wu D, Han G (2016) The involvement of survivin in insulinlike growth factor 1-induced epithelial-mesenchymal transition in gastric cancer. Tumour Biol 37(1):1091–1096. https ://doi. org/10.1007/s1327 7-015-3909-x
Li H, Xu L, Zhao L, Ma Y, Zhu Z, Liu Y, Qu X (2015) Insulin-like growth factor-I induces epithelial to mesenchymal transition via GSK-3beta and ZEB2 in the BGC-823 gastric cancer cell line. Oncol Lett 9(1):143–148. https ://doi.org/10.3892/ol.2014.2687
Lin YC, Lin JC, Hung CM, Chen Y, Liu LC, Chang TC, Kao JY, Ho CT, Way TD (2014) Osthole inhibits insulin-like growth factor1-induced epithelial to mesenchymal transition via the inhibition of PI3K/Akt signaling pathway in human brain cancer cells. J Agric Food Chem 62(22):5061–5071. https ://doi.org/10.1021/ jf501 047g
Ling C, Wang X, Zhu J, Tang H, Du W, Zeng Y, Sun L, Huang JA, Liu Z (2019) MicroRNA-4286 promotes cell proliferation, migration, and invasion via PTEN regulation of the PI3K/Akt pathway in non-small cell lung cancer. Cancer Med 8(7):3520–3531. https: // doi.org/10.1002/cam4.2220
Liu CA, Chang CY, Hsueh KW, Su HL, Chiou TW, Lin SZ, Harn HJ (2018a) Migration/invasion of malignant gliomas and implications for therapeutic treatment. Int J Mol Sci. https ://doi.org/10.3390/ ijms1 90411 15
Liu F, Sun Y, Liu B, Lu J, Li H, Zhu H, Gao H, Zhou X, Chang H (2018b) Insulin-like growth factor-1 induces epithelial-mesenchymal transition in hepatocellular carcinoma by activating survivin. Oncol Rep 40(2):952–958. https ://doi.org/10.3892/or.2018.6516
Llobet-Navas D, Rodriguez-Barrueco R, Castro V, Ugalde AP, Sumazin P, Jacob-Sendler D, Demircan B, Castillo-Martin M, Putcha P, Marshall N, Villagrasa P, Chan J, Sanchez-Garcia F, Pe’er D, Rabadan R, Iavarone A, Cordon-Cardo C, Califano A, Lopez-Otin C, Ezhkova E, Silva JM (2014) The miR-424(322)/503 cluster orchestrates remodeling of the epithelium in the involuting mammary gland. Genes Dev 28(7):765–782. https ://doi.org/10.1101/ gad.23740 4.114
Lo HW (2010) Targeting Ras-RAF-ERK and its interactive pathways as a novel therapy for malignant gliomas. Curr Cancer Drug Targets 10(8):840–848. https ://doi.org/10.2174/15680 09107 93357 970
Luo D, Xu X, Li J, Chen C, Chen W, Wang F, Xie Y, Li F (2018) The PDK1/cJun pathway activated by TGFbeta induces EMT and promotes proliferation and invasion in human glioblastoma. Int J Oncol 53(5):2067–2080. https ://doi.org/10.3892/ijo.2018.4525
Ma C, Wei F, Xia H, Liu H, Dong X, Zhang Y, Luo Q, Liu Y, Li Y (2017) MicroRNA-10b mediates TGF-beta1-regulated glioblastoma proliferation, migration and epithelial-mesenchymal transition. Int J Oncol 50(5):1739–1748. https ://doi.org/10.3892/ ijo.2017.3947
McCarthy TL, Centrella M (2015) Androgen receptor activation integrates complex transcriptional effects in osteoblasts, involving the growth factors TGF-beta and IGF-I, and transcription factor C/EBPdelta. Gene 573(1):129–140. https: //doi.org/10.1016/j. gene.2015.07.037
Menu E, Kooijman R, Van Valckenborgh E, Asosingh K, Bakkus M, Van Camp B, Vanderkerken K (2004) Specific roles for the PI3K and the MEK-ERK pathway in IGF-1-stimulated chemotaxis, VEGF secretion and proliferation of multiple myeloma cells: study in the 5T33MM model. Br J Cancer 90(5):1076–1083. https ://doi.org/10.1038/sj.bjc.66016 13
Park GB, Kim D (2018) Insulin-like growth factor-1 activates different catalytic subunits p110 of PI3K in a cell-type-dependent manner to induce lipogenesis-dependent epithelial-mesenchymal transition through the regulation of ADAM10 and ADAM17. Mol Cell Biochem 439(1–2):199–211. https ://doi.org/10.1007/s1101 0-017-3148-0
Petz LN, Ziegler YS, Schultz JR, Nardulli AM (2004) Fos and Jun inhibit estrogen-induced transcription of the human progesterone receptor gene through an activator protein-1 site. Mol Endocrinol 18(3):521–532. https ://doi.org/10.1210/me.2003-0105
Rohrmann S, Linseisen J, Becker S, Allen NE, Schlehofer B, Overvad KK, Olsen A, Tjonneland AAM, Melin B, Lund E, Vineis P, Grioni S, Tumino R, Palli D, Matiello A, Bonet C, Chirlaque M-D, Sanchez M-J, Rodriguez L, Dorronsoro M, Ardanaz E, Lagiou P, Trichopoulou A, Trichopoulos D, Dossus L, Grote VA, Boeing H, Aleksandrova K, Bueno-de-Mesquita HB, van Duijnhoven F, Peeters PHM, Khaw K-T, Wareham NJ, Key TJ, Rinaldi S, Romieux I, Gallo V, Michaud DS, Riboli E, Kaaks R (2011) Concentrations of IGF-I and IGFBP3 and brain cancer risk in the European Prospective Investigation into Cancer and Nutrition (EPIC). Cancer Epidemiol Biomark Prev. https ://doi. org/10.1158/1055-9965.EPI-11-0179
Roy LO, Poirier MB, Fortin D (2018) Differential expression and clinical significance of transforming growth factor-beta isoforms in GBM tumors. Int J Mol Sci. https: //doi.org/10.3390/ijms1904111 3
Samani AA, Yakar S, LeRoith D, Brodt P (2007) The role of the IGF system in cancer growth and metastasis: overview and recent insights. Endocr Rev 28(1):20–47. https ://doi.org/10.1210/ er.2006-0001
Seystahl K, Papachristodoulou A, Burghardt I, Schneider H, Hasenbach K, Janicot M, Roth P, Weller M (2017) Biological role and therapeutic targeting of TGF-beta3 in glioblastoma. Mol Cancer Ther 16(6):1177–1186. https ://doi.org/10.1158/1535-7163. MCT-16-0465
Sirianni R, Nogueira E, Bassett MH, Carr BR, Suzuki T, Pezzi V, Andò S, Rainey WE (2010) The AP-1 family member FOS blocks transcriptional activity of the nuclear receptor steroidogenic factor 1. J Cell Sci 123(Pt 22):3956–3965. https ://doi.org/10.1242/ jcs.05580 6
Song Y, Chen Y, Li Y, Lyu X, Cui J, Cheng Y, Zheng T, Zhao L, Zhao G (2019) Resveratrol suppresses epithelial-mesenchymal transition in GBM by regulating Smad-dependent signaling. BioMed Res Int 2019:1321973. https ://doi.org/10.1155/2019/13219 73
Tsunoda S, Avezov E, Zyryanova A, Konno T, Mendes-Silva L, Pinho Melo E, Harding HP, Ron D (2014) Intact protein folding in the glutathione-depleted endoplasmic reticulum implicates alternative protein thiol reductants. eLife 3:e03421. https ://doi.org/10.7554/ eLife .03421
Urbanska K, Sokolowska J, Szmidt M, Sysa P (2014) Glioblastoma multiforme – an overview. Contemp Oncol (Pozn) 18(5):307–312. https ://doi.org/10.5114/wo.2014.40559
Walsh LA, Damjanovski S (2011) IGF-1 increases invasive potential of MCF 7 breast cancer cells and induces activation of latent TGF-β1 resulting in epithelial to mesenchymal transition. Cell Commun Signal 9(1):10. https ://doi.org/10.1186/1478-811x-9-10
Weroha SJ, Haluska P (2012) The insulin-like growth factor system in cancer. Endocrinol Metab Clin N Am 41(2):335–350. https: //doi. org/10.1016/j.ecl.2012.04.014
Whitmarsh AJ (2007) Regulation of gene transcription by mitogenactivated protein kinase signaling pathways. Biochem Biophys Acta 1773(8):1285–1298. https ://doi.org/10.1016/j.bbamc r.2006.11.011
Zhu VF, Yang J, Lebrun DG, Li M (2012) Understanding the role of cytokines in Glioblastoma Multiforme pathogenesis. Cancer Lett 316(2):139–150. https ://doi.org/10.1016/j.canle t.2011.11.001
Zumkeller W, Westphal M (2001) The IGF/IGFBP system in CNS malignancy. Mol Pathol 54(4):227–229. https ://doi.org/10.1136/ mp.54.4.227