N6-methyladenosine (m6A) is the most prevalent covalent modification of mammalian mRNA and is involved in regulating post-transcriptional gene expression. Despite methodological advances, it remains challenging to directly map all m6A sites in the transcriptome. Now, Nature Chemical Biology reports two new methods that enable transcriptome-wide mapping of m6A independently of sequence context.
Existing methods have provided invaluable information about the m6A landscape but each has limitations. Antibody-based m6A-seq (also termed MeRIP-seq) has a low resolution of 100–200 nucleotides. Other methods indirectly infer the positions of modified adenosines from their proximity to RNA–antibody crosslinking sites (in crosslinking and immunoprecipitation (CLIP)-based approaches) or enzymatically modified sites (in deamination adjacent to RNA modification targets (DART)-seq). Approaches based on m6A-sensitive endoribonucleases (such as m6A-REF-seq or MAZTER-seq) detect only the ~16–25% of total m6A sites that fall within their recognition motif.
Detection of m6A is challenging because the N6-methyl group is largely unreactive and therefore difficult to chemically label. Additionally, m6A does not cause nucleotide mis-incorporation or termination during reverse transcription (RT), which precludes the use of mutation- or truncation-based sequencing readouts.
To address these issues, Shu et al. developed m6A-label-seq, a single-nucleotide resolution, transcriptome-wide approach based on metabolic labelling. Normally, m6A methylation enzymes transfer a methyl group from S-adenosyl methionine (SAM) to the N6 position of target adenosines. However, in cells cultured with a SAM analogue, Se-allyl-l-selenohomocysteine (SeAM), m6A sites are instead metabolized to N6-allyladenosine (a6A). Chemically induced cyclization of a6A can lead to mis-incorporations during RT, enabling detection of modified adenosines as A-to-T/C/G mutations in sequencing data.
In proof-of-concept experiments, m6A-label-seq detected a few thousand modified sites in each of three different cell types (human HeLa and HEK293T and mouse H2.35), a subset of which were validated by an orthogonal approach. Comparisons with published data for other single-base resolution techniques (DART-seq, m6A-REF-seq and miCLIP) in HEK293T cells indicated that m6A-label-seq is better at detecting clustered m6A sites. The authors suggest that the superior performance of m6A-label-seq reflects the limitations of indirect and sequence-dependent approaches.
Wang et al. based the second method, m6A-SEAL, on the ability of FTO RNA demethylase to convert m6A into a reactive intermediate, N6-hydroxymethyladenosine (hm6A). Subsequent treatment of hm6A with dithiothreitol (DTT) generated N6-dithiolsitolmethyladenosine (dm6A), which has a free sulfhydryl group that can be functionally tagged with biotin for streptavidin-based affinity purification followed by sequencing.
The authors show that m6A-SEAL specifically detects converted m6A and is highly sensitive — it can be performed on as little as 500 ng of polyA+ RNA. Comparisons of false-positive rates indicate that it is more reliable than MeRIP-seq, DART-seq, m6A-REF-seq and a number of CLIP-based methods. Although the current version of m6A-SEAL has a low resolution, comparable to that of m6A-seq, the authors hope that single-nucleotide resolution can be achieved by adopting readouts based on terminated or disrupted RT. A notable strength of m6A-SEAL is its flexibility. Optimizing conditions should enable the detection of other RNA modifications, such as hm6A or N6,2ʹ-O-dimethyladenosine at the cap +1 position (cap m6Am). Moreover, the sulfhydryl group can be tagged with different functional groups for different applications, such as fluorophores for imaging.
“m6A-label-seq and m6A-SEAL represent complementary, orthogonal techniques”
m6A-label-seq and m6A-SEAL represent complementary, orthogonal techniques that, together with existing approaches, will facilitate a fuller understanding of the distribution and biological roles of m6A RNA modification.
Shu, X. et al. A metabolic labeling method detects m6A transcriptome-wide at single base resolution. Nat. Chem. Biol. https://doi.org/10.1038/s41589-020-0526-9 (2020)
Wang, Y. et al. Antibody-free enzyme-assisted chemical approach for detection of N 6-methyladenosine. Nat. Chem. Biol. https://doi.org/10.1038/s41589-020-0525-x (2020)
About this article
Cite this article
Clyde, D. New tools for transcriptome-wide mapping of m6A. Nat Rev Genet 21, 387 (2020). https://doi.org/10.1038/s41576-020-0248-6