Fig 1. Alternative splicing generates functional METTL3 enzyme isoforms.

(A) N⁶-methyladenosine (m⁶A) in mammalian mRNA is installed by a core complex containing METTL3 and METTL14 [3], as well as additional factors that are not shown. While METTL3 is the catalytic component, METTL14 is required for complex stability and activity. (B) The mouse Mettl3 transcript contains multiple exons, two of which were targeted for mutagenesis to generate genetic knockouts in different studies. While excising exon 4 generated true Mettl3 knockout mouse embryonic stem cells that lack m⁶A in mRNA [5], CRISPR-Cas9-mediated mutagenesis of exon 2 allowed for alternatively spliced Mettl3 transcripts to be generated [4,8]. Some of these new transcript isoforms could be translated in cells, forming functional METTL3 enzyme. The discovery of these METTL3 isoforms finally explains why different cell lines previously thought to completely lack this enzyme show large variation in the amount of residual m⁶A in mRNA.