Global and single-nucleotide resolution detection of 7-methylguanosine in RNA

Figures & data

Figure 1. Schematic view of chemistry and Bo-sequencing methodology. Schematic overview of the Bo-Seq method. Representation of the sequential steps: i) size-selection of tRnas and de-aminoacylation; ii) chemical reactions leading to the RNA chain cleavage; iii) 5’ phosphorylation, adaptor ligation (RNA adaptors are indicated in blue and green, RNA in red); iv) library preparation and sequencing; v) bioinformatic analysis (reads are indicated in grey).

Figure 2. Cleavage at m⁷G is highest when RNAs are treated with 1 M NaBH₄, aniline and m⁷GMP. (a) schematic representation of prokaryotic 16S rRNA methylated at G527. Treatment with NaBH₄ and aniline lead to the generation of two RNA fragments. (b) optimization of 16S rRNA cleavage reaction at methylated position m⁷G527 with NaBH₄ and aniline. The different reaction conditions applied are indicated in the upper panel of the agarose gel. Ladd: ladder. * indicates the full length 16S rRNA and arrows indicate the derived fragments. (c) schematic representation of human 18S rRNA with m⁷G at position G1639. Cleavage of 18S rRNA using NaBH₄ and aniline lead to the formation of two RNA fragments. Arrow indicates the position of primers in the primer extension reaction. (d) primer extension analysis performed in human 18S rRNA untreated (-) or treated with 1 M NaBH₄ pH 7.5, aniline and m⁷GMP. (e) each cleaved fraction represents the densitometry of the RT-stop bands relative to the full-length product + each RT-stop (m⁷G + N⁺¹ sites) band. The uncleaved fraction represents the densitometry of the full-length product relative to the full-length product + each RT-stop band. (f) schematic representation of tRNA with m⁷G at position G46, and the cleavage that leads to the formation of tRNA fragments. Arrow indicates the position of primers in the primer extension reaction. (g) primer extension analysis performed in human tRnas untreated (-) or treated with NaBH₄ pH 7.5, aniline and m⁷GMP. The arrow indicates the RT-stops. (h) the fraction cleaved represents the relative densitometry of the sum of each RT-stop bands (m⁷G + N⁺¹ sites) relative to the relative densitometry of the sum of each RT-stop bands (m⁷G + N⁺¹ sites) in condition 4.

Figure 3. Heatmap showing normalized tRNA fragments read coverage mapped to individual tRNA isoacceptors derived from NaBH₄-treated RNAs from DU145 cells. Black boxes indicate cleavage site, corresponding to G46 + 1. In red are methylated tRNA isoacceptors.

Figure 4. Detection of m⁷G-methylated and unmethylated tRnas using Bo-Seq. A, B) Representative cleavage score profiles of methylated (a) and unmethylated (b) tRnas from DU145 cells. c, d) read coverage of fragmented tRnas after NaBH₄/Aniline treatment of PC3 METTL1 KO (grey lines) and WT (red lines) RNAs for the indicated m⁷G methylated (c) and non-methylated tRnas (d). Circles represent the tRNA fragment (tRF) start sites. Red arrows indicate cleavage position.

Figure 5. Global detection of m⁷G in tRNA by dot blot. (a) sensitivity of MBL and biovision (BIO) anti-m⁷G antbodies at detecting m⁷G in total RNA. m⁷G-hypermethylated RNA was used as a control. (b) quantitative detection of m⁷G-modified RNAs in increasing levels of total RNA and tRnas from the human cell line PC3 using MBL and biovision (BIO) anti-m⁷G antibodies. One replicate for each condition is shown in the upper panel. Methylene blue staining is used as loading control. (c) bar plots represent densitometry quantification of signal intensity (in arbitrary units) using ImageJ software, normalized to methylene blue. Mean ± SD is represented (n = 3). (d) evaluating the specificity of MBL and biovision anti-m⁷G antibodies by competing with m⁷G-hypermethylated RNA (m⁷G-RNA at 20 μg, 200 μg, and 1000 μg), m⁷GTP (at 1 μg, 2 μM, and 4 μM) and GTP (at 1 μM, 2 μM, and 4 μM). E) METTL1 protein expression levels in single-cell derived clones of PC3 control (WT) and PC3 METTL1 KO cell lines. Tubulin was used as loading control. e, f) dot blot assay of tRnas and large RNAs (>200 nt) from PC3 control (WT) and METTL1 KO cell lines after incubation with MBL (e) and biovision (f) anti-m⁷G antibodies. Two technical replicates from three biological replicates are shown for every condition. Densitometry quantification is shown at the bottom. Mean ± SD are represented (n = 2). Dotted lines represent average densitometry to all METTL1 KO tRNA samples. Statistics: one-tailed Student’s t-test: *p < 0.05; **p < 0.01; ***p < 0.001.

Table 1. Description of available methods for internal m7G mapping in RNA.

	Method	Type of RNA	Bioinformatics analysis	Advantages	Limitations
m^7G-MeRIP-seq^33	Antibody-based approach using anti-m^7G antibodies	Mammalian tRNA and mRNA	Sequences trimmed with cutadatp2 Alignment done with Tophat2 Reference genome human-hg19 Peak calling with exomePeak and MeTPeak packages Bam-readcount available at https://github.com/genome/bam-readcount RNA-seq data available at GEO GSE112276	Identification of internal m^7G sites in tRNA and mRNA	Required highly specific antibodies lack of sensitivity to detect m^7G in low abundant RNA does not uncover m^7G modification sites at single-nucleotide resolution
m^7G-MaP-seq^35	NaBH4 treatment leads to formation of abasic sites at m^7G positions, which are converted into cDNA mutations during reverse transcription process. Increase mutagenesis rate in treated samples compared with controls identifies the m^7G location	Plant, yeast and human rRNAs and tRNAs	Sequences trimmed with Cutadatp rRNAs from CWR database tRNAs mapped with GtRNAdb Mapping with Bowtie2 Mutation sites identified by mpileup tool Analysis available at https://github.com/jeppevinther/m7g_map_seq Data available at GEO (GSE121927)	It can be applied to different sequencing protocols and allow identification of internal m^7G position in several RNA types. It bypasses the background caused by RNA fragmentation.	NaBH4 treatment increase mutation rate also in ac [4]C, D and m [1]A positions non-stochiometric method high variability between replicates low sensitivity for low abundant RNA and methylated RNA
BoRed-seq^26	Generation of abasic site using NaBH4 at low pH and addition of biotinylated robes to identify methylated RNA	microRNA	Sequenced trimming with Trimmomatic Reference human genome GRCh38/hg38 Mapping with STAR FastQC program available at https://github.com/s-andrews/FastQC Data available at GEO (GSE112180, GSE112181, GSE120454, GSE120455)	Detection of m^7G in low abundant and small RNA, such as miRNA	Complex steps to add biotin-coupled aldehyde reactive probes
m^7G-quant-seq^36	High performance of KBH4-induced abasic sites by using high KBH4 concentrations, HIV reverse transcriptase and 1 mM dNTPs. Mutations and deletions caused by abasic sites during reverse transcription identify the m^7G location with quantitative accuracy	Human rRNA and tRNA	Sequences trimmed with cutadatp2 Alignment done with Tophat2 and bowtie2 Reference genome human-hg38 Analysis of mutation rates and deletions Bam-readcount available at https://github.com/genome/bam-readcount RNA-seq data available at GSE209646	Identification of stoichiometric m^7G modification levels.	complex processing steps and bioinformatic analysis to quantify relative m^7G levels in tRNA could misidentify m^22G and m [1]G as m^7G lack of consistency with other methods identifying only 11 out of the 22 known m^7G-modified tRNA targets.
m^7G-seq^33	Selective biotinylation of NaBH4-induced abasic sites. Mutations caused by misincorporation at biotinylated sites during reverse transcription identify the m^7G location.	Human tRNA and mRNA	Sequences trimmed with cutadatp2 Alignment done with Tophat2 Reference genome human-hg19 Analysis of G→T and G→C mutation rates Bam-readcount available at https://github.com/genome/bam-readcount RNA-seq data available at GEO GSE112276	Identification of low frequency m^7G modification. It bypasses the background caused by RNA fragmentation	Non-stochiometric method high variability between replicates complex step of biotin conjugation of abasic site using hydrazidation
AlkAlaline-seq^32	Generation of abasic site under alkaline condition, 5’ and 3’ dephosphorylation followed by aniline-induced cleavage	Bacterial, yeast, and human cytoplasmic and mitochondrial tRNA and rRNA	Raw reads were trimmed using Trimmomatic v32 software Aligment was performed using Bowtie2 (v2.2.4) Code non available RNA-seq data no available	Reliable technique for precise internal-m^7G profiling in tRNA and rRNA with high specificity	High RNA degradation due to harsh environment limited specificity of the technique due to cross-reactivity with m [3]C and D
TRAC-seq^31	AlkB treatment of size selected RNA to remove major RNA modifications, followed by NaBH4 and aniline treatment	Human and mouse tRNA	Sequences were trimmed using Trim Galore! Aligment was performed using Bowtie tRNAs from GtRNAdb Coverage calculation with Bedtools Analysis available at https://github.com/rnabioinfor/TRAC-Seq RNA-seq data available at GEO (GSE112670)	Specific single-nucleotide resolution profiling of m^7G sites in tRNAs	long procedure (9 days) demethylation step may lead to loss of input material partially impurities in the recombinant protein may remove modifications or induce RNA degradation non-reliable fragmentation: bias on 3’ tRNA fragment amplification before NaBH4 treatment identification of m^7G sites only tested for small and abundant RNA
Bo-seq	Optimised NaBH4-anlinine reaction using adequate concentration of m^7GMP	Human t RNA	sequence trimming fastp tRNA sequences: predicted by tRNAscan-SE combined with gtRNAdb aligment: LAST isotype position assignments: INFERNAL using tRNA HMM profiles Code available at https://github.com/Cancer-Genomics-TH/tRNA_methyl RNA-seq data available at GEO (GSE203255)	Simple, quick, and reliable technique for precise internal-m^7G profiling in human small RNAs	identification of m^7G sites only tested for small and abundant RNA

Torres AG, Batlle E, Ribas de Pouplana L. Role of tRNA modifications in human diseases. Trends Mol Med. 2014;20(6):306–314. doi: 10.1016/j.molmed.2014.01.008

 PubMed Web of Science ®Google Scholar

Shi H, Chai P, Jia R, et al. Novel insight into the regulatory roles of diverse RNA modifications: re-defining the bridge between transcription and translation. Mol Cancer. 2020;19(1):78. doi: 10.1186/s12943-020-01194-6

 PubMed Web of Science ®Google Scholar

Delaunay S, Frye M. RNA modifications regulating cell fate in cancer. Nat Cell Biol. 2019;21(5):552–559. doi: 10.1038/s41556-019-0319-0

 PubMed Web of Science ®Google Scholar