Preview

PDF, English Download (43MB) | Terms of use

Abstract

The acquisition of mutations within the genome of hematopoietic stem cells (HSCs) is of particular importance as this is a likely driver of malignant transformation for many leukemias, as well as a hallmark of ageing. Importantly, the study of normal physiologic mediators of HSC mutation acquisition is a largely underdeveloped area. The advent of next generation sequencing (NGS) technologies provides a route to interrogate this phenomenon but is complicated by the fact that a genome-wide analysis at the clonal level is necessary to determine the mutational signature of individual HSCs.

We have developed an in vitro model that allows for the sensitive assessment of clonal mutations occurring within single HSCs using whole genome sequencing. In order to optimize our clonal mutation analysis, we performed a benchmarking exercise where we deeply sequenced an individual HSC colony to ~90X coverage and performed somatic nucleotide variant (SNV) analysis at various down-sampled coverages. Importantly, we established that the number of mutations called increases in an almost linear fashion with increasing coverage, until a plateau is reached at around 30X coverage. We additionally developed optimal filtering parameters, which demonstrated a much-improved capacity for discerning true positive and false positive mutations at low coverage, compared to previously published methodologies.

Using this optimized sequencing pipeline, we collected and sequenced HSC clones from young and old mice, as well as those exposed to stress agonists known to induce HSC cycling. We additionally employed a genetic label-retention system to segregate dormant and actively cycling HSCs in order to assess whether mutations are predominantly acquired during replication. Genomic coverage of the majority of these HSC colonies ranged from 30-40X. As seen in humans, we found a progressive increase in mutation burden with age within the murine HSC compartment, corresponding to a rate of ~40 SNVs per year. Furthermore, these HSCs had mutational signatures corresponding to that observed in aged human tissues. In contrast to previous reports, data from the label-retention model demonstrated that this age-associated increase in mutation burden correlated with HSC replication, as dormant aged HSCs had similar mutation burdens to young HSCs. Seemingly contradictory to this finding, we observed no difference in mutation burden upon stress-induced cycling of HSCs. However, it appears we have unintentionally introduced a large selection bias with regards to the agonist-treated clones and future work will focus on rectifying this caveat.

In summary, we have developed a sensitive and specific analysis to accurately detect mutations within individual HSCs. From our results we have clearly demonstrated that mutation acquisition within HSCs accumulates with age and that this increase correlates with an increase in replication history. We envisage that these findings will be an important step towards interrogating whether replication stress is a biologically relevant driver of genome instability in HSCs.