Clonal Dynamics of Normal Haematopoiesis with Human Ageing

The haematopoietic system manifests several age-associated phenotypes including anaemia; loss of regenerative capacity, especially in the face of insults such as infection, chemotherapy or blood loss; and increased risk of clonal haematopoiesis and blood cancers. The cellular alterations that underpin these age-related phenotypes, which typically manifest in individuals aged over 70, remain elusive. We aimed to investigate whether changes in HSC population structure with age might underlie any aspects of haematopoietic system ageing.

We sequenced 3579 genomes from single-cell-derived colonies of haematopoietic stem cell/multipotent progenitors (HSC/MPPs) from 10 haematologically normal subjects aged 0-81 years. HSC/MPPs accumulated 17 somatic mutations/year after birth with no increased rate of mutation accumulation in the elderly. HSC/MPP telomere length declined by 30 bp/yr. In cord blood and adults aged <65, a small proportion of HSC/MPPs had unexpectedly long telomeres, as assessed using several criteria for outliers. The proportion of cells with unexpectedly long telomeres reduced in frequency with age. Given that telomeres shorten at cell division, these outlier cells have presumably undergone fewer historic cell divisions, supporting the existence of a rare population of dormant HSCs in humans that declines in frequency with age.

To interrogate changes in HSC population structure with age, we used the pattern of unique and shared mutations between the sampled cells from each individual to reconstruct their phylogenetic relationships. The frequency of branch-points (known as coalescences) in phylogenetic trees in a neutrally evolving, well-mixed population of somatic cells is primarily determined by the product of population size and time between symmetric self-renewal cell divisions (Nt). Smaller populations and more frequent symmetric divisions both increase the density of coalescences. Specific clones can come to dominate either through neutral drift or positive selection.

We found that haematopoiesis in adults aged <65 was polyclonal, with high indices of clonal diversity. The number and pattern of coalescent events in the phylogenies showed that a stable population of 20,000 to 200,000 HSC/MPPs was contributing evenly to blood production in young adult life. In contrast, haematopoiesis in individuals aged >75 showed profoundly decreased clonal diversity. In each elderly subject, 30-60% of haematopoiesis was accounted for by 12-18 independent clones, each contributing 1-34% of blood production. Most clones had begun their expansion before age 40, but only 22% had known driver mutations.

We used the ratio of non-synonymous to synonymous mutations (dN/dS) to identify any excess of non-synonymous (driver) mutations in the dataset. This genome-wide selection analysis estimated that 1/34 to 1/12 non-synonymous mutations were drivers, occurring at a constant rate throughout life, such that the set of 300 - 400 HSC/MPPs sampled from each adult individual harboured around 100 driver mutations, over 10-fold higher than the number of known drivers we could identify. Novel drivers affected a wider pool of genes than identified in blood cancers. The genes DNMT3A, ZNF318 and HIST2H3D were identified as being under significant positive selection in HSC/MPPs, despite ZNF318 and HIST2H3D not being enriched in the setting of myeloid malignancies. Loss of Y chromosome conferred selective benefits on HSC/MPPs in males.

Simulations from a simple model of haematopoiesis, with constant HSC population size and constant acquisition of driver mutations conferring moderate fitness benefits, entirely explained the abrupt change in clonal structure observed in the elderly, which could not be explained by neutral models incorporating drift alone. Our data supports the view that dramatically decreased clonal diversity is a universal feature of haematopoiesis in aged humans, underpinned by pervasive positive selection acting on many more genes than currently known. By old age the majority of HSCs harbour at least one driver mutation. With such ubiquity of driver mutations, selected purely for their competitive advantage within the stem cell compartment, and with the wholesale rewiring of cellular pathways they induce, it is feasible that they may contribute to age-related phenotypes beyond the increased risk of blood cancer.