Region-specific RNA m6A methylation represents a new layer of control in the gene regulatory network in the mouse brain

Abstract

N⁶-methyladenosine (m⁶A) is the most abundant epitranscriptomic mark found on mRNA and has important roles in various physiological processes. Despite the relatively high m⁶A levels in the brain, its potential functions in the brain remain largely unexplored. We performed a transcriptome-wide methylation analysis using the mouse brain to depict its region-specific methylation profile. RNA methylation levels in mouse cerebellum are generally higher than those in the cerebral cortex. Heterogeneity of RNA methylation exists across different brain regions and different types of neural cells including the mRNAs to be methylated, their methylation levels and methylation site selection. Common and region-specific methylation have different preferences for methylation site selection and thereby different impacts on their biological functions. In addition, high methylation levels of fragile X mental retardation protein (FMRP) target mRNAs suggest that m⁶A methylation is likely to be used for selective recognition of target mRNAs by FMRP in the synapse. Overall, we provide a region-specific map of RNA m⁶A methylation and characterize the distinct features of specific and common methylation in mouse cerebellum and cerebral cortex. Our results imply that RNA m⁶A methylation is a newly identified element in the region-specific gene regulatory network in the mouse brain.

Keywords: N6-methyladenosine; RNA methylation; epitranscriptomic mark; mouse cerebellum; mouse cerebral cortex.

Figure 1.

Protein expression of m⁶A writer and eraser genes in adult mice cerebellum and cerebral cortex. (a) Western blot analysis of METTL3, METTL14, WTAP, ALKBH5 and FTO in adult mouse cerebellum (B1, B2) and cerebral cortex (C1, C2). Beta-actin was used here as an internal control. Experiments were performed in biological triplicates using six mice in total. Representative results are shown here. (b,c) Representative images of immunostaining of METTL3, METTL14, WTAP, ALKBH5 and FTO using paraffin sections of adult mice cerebellum (b) and cerebral cortex (c). ML, molecular layer; IGL, inner granule cell layer; PCL, Purkinje cell layer. Scale bars represent 200 µm. The laminar structure of the cortex was labelled from layer I to layer VI. Experiments were performed in biological triplicates using three male and three female mice in total and representative data are given here. Enlarged images in the square area are shown in the lower panels and scale bars represent 10 µm in (b) and 50 µm in (c).

Figure 2.

Region-specific m⁶A methylation in the mouse cerebellum and cerebral cortex. (a, b) Venn diagrams showing the numbers of overlapping m⁶A transcripts in the two biological replicates of m⁶A-IP in the cerebellum (a) and the cerebral cortex (b). (c) Venn diagram showing the numbers of genes commonly or specifically expressed in the cerebellum or the cerebral cortex. (d) Venn diagram showing the numbers of CMRs and SMRs. Numbers of specifically expressed genes among SMRs are shown in parentheses. (e) Column chart showing the numbers of common and specific m⁶A peaks in mouse cerebellum and cerebral cortex. The blue bars indicate common peaks, while the orange bars indicate specific peaks. (f, g) Box plots showing the methylation levels of cerebellar RNAs and cortical RNAs by comparing the median fold enrichment at peak levels (f) and gene levels (g). (h, i) Box plots showing the methylation levels of CMRs and SMRs by comparing the fold enrichment at the peak level (h) or gene level (i). Wilcoxon test was performed for statistical analysis. ***p value < 2.2 × 10⁻¹⁶.

Figure 3.

Distribution patterns of m⁶A peaks in mouse cerebellar and cerebral cortical mRNAs. (a) Sequence logo representing the deduced consensus motif through clustering of all enriched m⁶A peaks in the cerebellum and the cerebral cortex. (b–d) Enrichment of all m⁶A peaks (b), common m⁶A peaks (c) and specific m⁶A peaks (d) along the whole mRNA transcripts. (e) Statistics of numbers of m⁶A peaks enriched in different regions along the mRNA transcripts. Total m⁶A peaks in the cerebellum (total-cer) and cortex (total-cor); specific m⁶A peaks in the cerebellum (Spec-cer) and cortex (Spec-cor); common m⁶A peaks in the cerebellum (Com-cer) and cortex (Com-cor) were included and compared.

Figure 4.

Universal RNA methylation in different types of neural cells. (a) Column charts showing the methylation status of different cell type-enriched genes in the cerebellum and the cortex. GC, granule neuronal cell; PC, Purkinje cell; BG, Bergmann glia cell; NC, neuronal cell; ODC, oligodendrocyte; ASC, astrocyte. The numbers of cell type-enriched genes are shown above each column; the numbers of methylated genes among those cell type-enriched genes are shown in the column. (b,c) Cumulative distribution function of log2-fold enrichment of m⁶A peaks in three kinds of cell type-enriched genes in mouse cerebellum (b) or in mouse cerebral cortex (c). (d) IGV plots showing the methylation status of representative cell type-enriched genes in the cerebellum and the cortex. The grey reads are from non-IP control (input) libraries; red and purple reads are from m⁶A-IP libraries of mouse cerebellum and cerebral cortex, respectively. Arrows show the direction of transcription. Y-axis represents the normalized numbers of reads count. Positions of m⁶A peaks are highlighted in the blue box.

Figure 5.

GO analysis of commonly and specifically methylated genes in mouse cerebellum and cerebral cortex. (a,b) GO functional analysis of the commonly methylated genes containing the top 3000 m⁶A peaks in mouse cerebellum (a) and cerebral cortex (b). (c,d) GO functional analysis of cerebellar (c) and cortical (d) specifically methylated genes.

Figure 6.

Methylation status of mouse synaptic mRNAs and FMRP target mRNAs in the mouse brain. (a,b) Venn diagrams showing the overlap between mouse cortical methylated genes and the genes encoding postsynaptic (a) or presynaptic proteins (b). (c) Column charts showing the methylation status of FMRP target mRNAs in comparison with all methylated mRNAs in mouse cerebellum and cerebral cortex. Orange bars indicate the unmethylated FMRP target mRNAs. The numbers on blue bars indicate the amount of methylated RNAs in each sample. (d) Pie charts showing the distribution of m⁶A peaks in methylated FMRP target mRNAs in mouse cerebellum and cerebral cortex. (e) Box plots showing the median methylation levels of FMRP target mRNAs in comparison with all methylated genes in mouse cerebellum and cerebral cortex, respectively. (f) Proposed model of the role of m⁶A RNA methylation in FMRP-induced translational repression. The target mRNAs are maintained at an appropriate RNA methylation level to be recognized by FMRP and enter a translational repression state via ribosome stalling. Upon relevant synaptic stimulus, mRNA methylation is altered, resulting from changed expression of one or more kinds of m⁶A writer, eraser or reader genes. Either increase or decrease in mRNA methylation may cause the dissociation of FMRP from target RNAs and then enable the ribosome elongation to reactivate translation. The one, two or three asterisks represent decreased, appropriate and increased methylation levels.