Loss of the de novo DNA methyltransferases Dnmt3a and Dnmt3b in embryonic stem cells obstructs differentiation; however, the role of these enzymes in somatic stem cells is largely unknown. Using conditional ablation, we show that Dnmt3a loss progressively impairs hematopoietic stem cell (HSC) differentiation over serial transplantation, while simultaneously expanding HSC numbers in the bone marrow. Dnmt3a-null HSCs show both increased and decreased methylation at distinct loci, including substantial CpG island hypermethylation. Dnmt3a-null HSCs upregulate HSC multipotency genes and downregulate differentiation factors, and their progeny exhibit global hypomethylation and incomplete repression of HSC-specific genes. These data establish Dnmt3a as a critical participant in the epigenetic silencing of HSC regulatory genes, thereby enabling efficient differentiation.

Summary The traditional view of hematopoiesis has been that all the cells of the peripheral blood are the progeny of a unitary homogeneous pool of hematopoietic stem cells (HSCs). Recent evidence suggests that the hematopoietic system is actually maintained by a consortium of HSC subtypes with distinct functional characteristics. We show here that myeloid-biased HSCs (My-HSCs) and lymphoid-biased (Ly-HSCs) can be purified according to their capacity for Hoechst dye efflux in combination with canonical HSC markers. These phenotypes are stable under natural (aging) or artificial (serial transplantation) stress and are exacerbated in the presence of competing HSCs. My- and Ly-HSCs respond differently to TGFβ1, presenting a possible mechanism for differential regulation of HSC subtype activation. This study demonstrates definitive isolation of lineage-biased HSC subtypes and contributes to the fundamental change in view that the hematopoietic system is maintained by a continuum of HSC subtypes, rather than a functionally uniform pool.

Clonal haematopoiesis is thought to be a rare condition that increases in frequency with age and predisposes individuals to haematological malignancy. Recent studies, utilizing next-generation sequencing (NGS), observed haematopoietic clones in 10% of 70-year olds and rarely in younger individuals. However, these studies could only detect common haematopoietic clones—>0.02 variant allele fraction (VAF)—due to the error rate of NGS. To identify and characterize clonal mutations below this threshold, here we develop methods for targeted error-corrected sequencing, which enable the accurate detection of clonal mutations as rare as 0.0003 VAF. We apply these methods to study serially banked peripheral blood samples from healthy 50–60-year-old participants in the Nurses' Health Study. We observe clonal haematopoiesis, frequently harbouring mutations in DNMT3A and TET2, in 95% of individuals studied. These clonal mutations are often stable longitudinally and present in multiple haematopoietic compartments, suggesting a long-lived haematopoietic stem and progenitor cell of origin.

Gains and losses in DNA methylation are prominent features of mammalian cell types. To gain insight into mechanisms that could promote shifts in DNA methylation and contribute to cell fate changes, including malignant transformation, we performed genome-wide mapping of 5-methylcytosine and 5-hydroxymethylcytosine in purified murine hematopoietic stem cells. We discovered extended regions of low methylation (Canyons) that span conserved domains frequently containing transcription factors and are distinct from CpG islands and shores. The genes in about half of these methylation Canyons are coated with repressive histone marks while the remainder are covered by activating histone marks and are highly expressed in HSCs. Canyon borders are demarked by 5-hydroxymethylcytosine and become eroded in the absence of DNA methyltransferase 3a (Dnmt3a). Genes dysregulated in human leukemias are enriched for Canyon-associated genes. The novel epigenetic landscape we describe may provide a mechanism for the regulation of hematopoiesis and may contribute to leukemia development.