|| Checking for direct PDF access through Ovid
N6-methyladenosine (m6A) has been identified as the most abundant modification on eukaryote messenger RNA (mRNA)1. Although the rapid development of high-throughput sequencing technologies has enabled insight into the biological functions of m6A modification2,3,4,5,6,7,8,9,10,11,12,13, the function of m6A during vertebrate embryogenesis remains poorly understood. Here we show that m6A determines cell fate during the endothelial-to-haematopoietic transition (EHT) to specify the earliest haematopoietic stem/progenitor cells (HSPCs) during zebrafish embryogenesis. m6A-specific methylated RNA immunoprecipitation combined with high-throughput sequencing (MeRIP–seq) and m6A individual-nucleotide-resolution cross-linking and immunoprecipitation with sequencing (miCLIP–seq) analyses reveal conserved features on zebrafish m6A methylome and preferential distribution of m6A peaks near the stop codon with a consensus RRACH motif. Inmettl3-deficient embryos, levels of m6A are significantly decreased and emergence of HSPCs is blocked. Mechanistically, we identify that the delayed YTHDF2-mediated mRNA decay of the arterial endothelial genesnotch1aandrhocacontributes to this deleterious effect. The continuous activation of Notch signalling in arterial endothelial cells ofmettl3-deficient embryos blocks EHT, thereby repressing the generation of the earliest HSPCs. Furthermore, knockdown ofMettl3in mice confers a similar phenotype. Collectively, our findings demonstrate the critical function of m6A modification in the fate determination of HSPCs during vertebrate embryogenesis.