N6-Adenosine Methylation of miRNA-200b-3p Influences Its Functionality and Is a Theranostic Tool

Abstract

MicroRNAs (miRNAs or miRs) play crucial roles in biological and pathological processes. Some miRNAs also appear as promising biomarkers and therapeutic tools. However, the epitranscriptomic regulation of miRNAs is not yet fully elucidated in all of their fields of application. We report that adenosine methylation of miR-200b-3p inhibits its repressive function toward its mRNA targets such as XIAP by blocking the formation of the miRNA/3' UTR^mRNA duplex. Our data indicate that the adenosine methylation of miR-200b-3p is associated with the survival of glioblastoma patients. Collectively, our data support the idea that the adenosine methylation of miR-200b-3p can be used as a prodrug having a selective cytotoxicity against cancer cells (while being harmless to peripheral blood mononuclear cells [PBMCs], astrocytes, neurons, and hepatocytes).

Keywords: adenosine methylation; biomarker; glioblastoma; miRNA methylation; prodrug.

Graphical abstract

Figure 1

The m6A Methyltransferase METTL3, the m6A Demethylase FTO, and α-Ketoglutarate Regulate the N6-Adenosine Methylation of miR-200b-3p (A) Each bar represents the relative expression level of miR-200b-3p (miR-200b-3p^exp) in 32 samples of GBM. After miRNA extraction from tumors, qRT-PCR was performed to evaluate the relative expression level of miRNA-200b-3p by using non-tumor brain samples as a reference and SNORD6.1 as a housekeeping miRNA. (B) Correlation between the relative expression level of miR-200b-3p and the percentage of N6-adenosine methylation in miR-200b-3p (miR200b-3p^%m6A). This percentage was calculated via the realization of RNA immunoprecipitation (IP) performed with an anti-m6A antibody followed by qPCR analysis (miRIP^m6A-qPCR) (as previously described by Berulava et al.⁶). Each square represents a GBM patient. (C) Absence of correlation between the relative expression level of FTO and the percentage of N6-adenosine methylation in miR-200b-3b. A human FTO ELISA kit (Tebu-Bio, France) was used to estimate the FTO expression. Each square represents a GBM patient. (D) Absence of correlation between the relative presence of α-ketoglutarate (αKG) and the percentage of N6-adenosine methylation in miR-200b-3p. An αKG assay kit (Abcam, France) was used to estimate the relative presence of αKG. Each square represents a GBM patient. (E) High percentage of N6-adenosine methylation in miR-200b-3p is observed in GBM harboring a low level of FTO and αKG. Each square represents a GBM patient. Red squares represent the average ± standard deviation of the two considered subgroups. (F) Impact of the downregulation of FTO on the adenosine methylation percentage of miR-200b-3p%^m6A calculated through the realization of miRIP^m6A-qPCR as previously described. (G) Impact of the αKG treatment on the adenosine methylation percentage of miR-200b-3p (miRNA-200b-3p^%m6A). (H) Impact of the meclofenamic acid (MA) treatment on miRNA-200b-3p^%m6A in U87 cells. After MA treatment (15 μM/24 h, Santa Cruz, France), the miRNA-200b-3p^%m6A was calculated through the realization of miRIP^m6A-qPCR as previously described. (I) Correlation between the relative expression level of METTL3 and the percentage of N6-adenosine methylation in miR-200b-3p. Each square represents a GBM patient. (J) Dot blot illustrating the presence of adenosine methylation in mimetic miR-200b-3p in the presence of METTL3 IP product. Adenosine detection is used as a control. S1/S2/S3 are three independent experiments. Ctrl, synthetic miR-200b-3p adenosine unmethylated; IP-IgG, product of IP performed with anti-IgG (the absence of an adenosine signal indicates that miRNA was not unspecifically immunoprecipitated); IP-METTL3, product of IP performed with anti-IgG. Products of IP were obtained in non-denaturing conditions in order to conserve enzymatic activity according to the manufacturer’s indications. (K) METTL3 knockdown (by siRNA approach) decreases the percentage of adenosine methylation of miR-200b-3p. (L) αFM^score reflects the METTL3, FTO, and αKG expression levels in the 32 GBM samples. Each circle represents a GBM patient. The αFM^score is higher when the percentage of adenosine methylation of miR-200b-3p is higher.

Figure 2

The N6-Adenosine Methylation of miR-200b-3p Limits Its Translational Repressor Function toward Anti-apoptotic Players and Confers Poor Prognosis in GBM Patients (A) Absence of correlation between the relative expression level of XIAP and miRNA-200b-3b (miR200b-3p^exp) in all 32 patients included in our study. A human XIAP ELISA kit (Abcam, France) was used to estimate the relative expression level of XIAP. Each square represents a GBM patient. (B) Samples were stratified according to the miR-200b-3p^exp and miR-200b-3p%^m6A parameters in order to distinguish the three indicated groups. Each box represents a sample/patient. For each group, the average of the XIAP expression was analyzed with a human XIAP ELISA kit (Abcam, France) and is represented on the graph. (C) Impact of the in vitro N6-adenosine methylation of miRNA-200b-3p^mimetic on the expression of the XIAP (human XIAP ELISA kit, Abcam, France). (D) Cross-linking immunoprecipitation and qPCR (CLIP-qPCR) were used to investigate the 3′ UTR/XIAP and 3′ UTR/EP300 (internal control) enrichments on GW182, TNRC6B, and IgG (negative control). Experiments were performed using the RiboCluster Profiler kit (CliniScience, France) according to the manufacturer’s instructions. (E) CLIP-qPCR was used to investigate the miR-150-5p (internal control) and miR-200b-3p enrichments on GW182, TNRC6B, and IgG (negative control). Experiments were performed using the RiboCluster Profiler kit (CliniScience, France) according to the manufacturer’s instructions. (F) For each sample, DEVDase activity was estimated as previously described. Each open square represents a sample. A blue square represents the average of samples having miR-200b-3p^m6A >10% or miRNA-200b-3p^exp-low. A gray square represents the average of samples having miR-200b-3p^m6A <10% or miRNA-200b-3p^exp-high. (G) Kaplan-Meier representation of survival curves for GBM patients whose tumors are characterized by a miR-200b-3p^m6A >10% or a miRNA-200b-3p^exp-low (in blue) and by a miR-200b-3p^m6A <10% and a miRNA-200b-3p^exp-high (in gray).

Figure 3

The N6-Adenosine Methylation of miR-200b-3p Selectively Induces Apoptosis in Cancer Cells and Has an Anti-tumor Growth Effect (A) IDH mutation in U87 cells induces a decrease of the αKG rate and an increase of miR-200b-3p^%m6A without FTO or METTL3 expression modification. (B) miR-200b-3p promotes cell death by itself in cancerous and non-cancerous cells (excepted neuron RN33b), while miR-200b-3b induces apoptosis by itself in U87 cells only. A lactate dehydrogenase (LDH)-cytotoxicity assay kit (Abcam, France) was used to estimate the cell death 24 h after the m6A-miR-200b-3b incubation. (C) Representation of our in vivo model of U87-induced GBM. Mice were xenografted with U87 cells. When tumor volume was equivalent to 100 mm³, mice were treated with TMZ and/or miR-200b-3p 5 days per week. Tumor volumes were measured after 3 weeks of treatment. (D) Impact of the adenosine-methylated form of miR-200b-3p on the tumor growth in mice model.

Figure 4

miR-200b-3p Could Also Be Used as a Therapeutic Tool in Other Cancer Types (A) Protein expression of different targets of mir-200b-3p has been studied by ELISA in U87 cells treated with an unspecific oligonucleotide or miR-200b-3p adenosine methylated or not. miR-200b-3p decreases Bcl-2 and PD-L1 expression only when not methylated and has no effect on other proteins studied. (B) In cell lines transfected with m6A-miR-200b-3p, cell death was induced in several cancer cell line types, when cell lines were able to demethylate this miRNA.