N6-adenosine methylation in MiRNAs

Abstract

Methylation of N6-adenosine (m6A) has been observed in many different classes of RNA, but its prevalence in microRNAs (miRNAs) has not yet been studied. Here we show that a knockdown of the m6A demethylase FTO affects the steady-state levels of several miRNAs. Moreover, RNA immunoprecipitation with an anti-m6A-antibody followed by RNA-seq revealed that a significant fraction of miRNAs contains m6A. By motif searches we have discovered consensus sequences discriminating between methylated and unmethylated miRNAs. The epigenetic modification of an epigenetic modifier as described here adds a new layer to the complexity of the posttranscriptional regulation of gene expression.

Fig 1. FTO knockdown in HEK293 cell clones FTO1C1, FTO2D4 and FTO3C3.

A) The levels of FTO transcripts were investigated by qRT-PCR. GAPDH was used as a reference gene. Mean ± SD for three pairs of scrambled (scr) and FTO specific siRNAs treated cells are shown. B) Reduced protein levels were revealed by Western blot in all three cell lines transfected with siRNAs targeting FTO mRNA. Exemplary photo is depicted.

Fig 2. Deregulation of miRNAs in FTO knockdown cells.

Mature miRNAs showing increased (A) and decreased (B) steady state levels in FTO knockdown cells. Normalized RNA-seq read numbers of individual miRNAs in FTO knockdown and scrambled siRNA treated cells were compared. Mean ± SD of three independent experiments are depicted. For verification and further studies, qRT-PCR analyses of selected mature miRNAs (C) and primary miRNA transcripts (D) were performed. We did not use other small RNAs as a reference gene for measuring mature miRNAs levels (as suggested by Life Technologies), since depletion of FTO might have an impact on their levels. Therefore, luciferase RNA was used to generate a standard curve and added to the qRT-PCR assays. GAPDH was used as a reference gene for measuring primary miRNAs transcript levels. Mean ± SD from quadruplicates per assay for three independent cell lines (FTO1C1, FTO2D4 and FTO3C3) are depicted. kd, FTO specific siRNA treated cells, scr, scrambled siRNA treated cells.

Fig 3. Knockdown of FTO does not significantly change mRNA levels of genes involved in miRNA biogenesis.

The steady-state mRNA levels of DICER, DROSHA, DGCR8 and ADAR were analyzed by qRT-PCR in cells treated with scrambled (scr) and FTO-specific siRNAs, respectively. GAPDH was used as a reference gene. The observed changes were not significant. Merged values of mean ± SD from triplicates per assay for the three independent cell lines FTO1C1, FTO2D4 and FTO3C3 are depicted. FTO kd, FTO-specific siRNA treated cells, scr siRNA, scrambled siRNA treated cells.

Fig 4. Effect of FTO knockdown on the steady state levels of methylated miRNAs.

X-axis, log2 fold-changes (log2fc) of enrichment after imuunopreciptation with an anti-m6A antibody; y-axis, log2 fold-changes of steady state miRNA levels after FTO knockdown. The values of all 239 methylated miRNAs are shown. The red dotted line is the regression line.