Unravelling the effect of proliferative stress and genotoxicity in hematopoietic stem cells in vivo

Hematopoietic Stem and Progenitor Cells (HSPCs) from patients affected by inherited disorders can be corrected with the use of Gene Therapy (GT), providing long-term therapeutic benefit upon reconstitution of the entire hematopoietic system. However, vector insertions may result in oncogene activation or tumour suppressor inactivation, resulting in malignant transformation as recently shown in the lentiviral vector (LV) based clinical trial for adrenoleukodystrophy and/or clonal expansions in different cancer immunotherapy applications. Moreover, the replicative stress accumulated during hematopoietic reconstitution and aging, may induce cellular senescence and trigger a chronic inflammatory response leading to hematopoietic decay. Here, to address the impact of insertional mutagenesis and age on haematopoiesis, we studied the clonal dynamics of hematopoietic reconstitution over time in wild-type (WT) C57 mice transplanted with HSPCs transduced either with a genotoxic LV harbouring the strong retroviral enhancer/promoter Spleen Focus Forming Virus in the Long Terminal Repeats (LTR), or with a safer self-inactivating LTRs and the moderate cellular promoter PGK. Additionally, the same HSPC-GT strategy was applied by transplanting Cdkn2a-/-HSPCs, which lack p16INK4A and p19ARF proteins and thus have no barriers against protooncogene activation. Blood composition and vector integration sites (IS) of B, T, and myeloid cells were monitored overtime (up to 2.5 years). Somatic mutations were identified by analysing the genomic portion of the mouse flanking each IS, and to assess mutation accumulation rates we developed a new Mutation Index (MI) which normalizes the number of mutations by clones and coverage. As expected, vector-driven activation of the Braf protooncogene prompted the acceleration of myeloid tumour onset in the Cdkn2a-/-/genotoxic group compared to the Cdkn2a-/-/nongenotoxic group (p < 0.0001). On the contrary, none of mice receiving WT cells developed tumours. Analysis of the blood composition overtime showed an age dependent skewing towards the myeloid lineage in all mice, which was exacerbated in both groups of mice transplanted with HSPCs transduced with the genotoxic LV. Overall, we retrieved >200,000 IS, corresponding to 9 Gb of sequence genomic information. We found that the MI in both the groups receiving HSPCs transplanted with the genotoxic LV were significantly higher when compared to nongenotoxic groups (p < 0.001) and specifically myeloid clones exhibited a higher mutation frequency compared to B and T cell lineages. Correctly reflecting that the presence of genotoxic insertion resulted in progressive somatic mutation accumulation and insertional mutagenesis. Moreover, the MI of the mice receiving WT HSPCs transduced with the genotoxic LV in the myeloid compartment was significantly higher than mice receiving Cdkn2a-/- cells transduced with the same vector (p < 0.01), reflecting the effect of aging. These results demonstrate for the first time that vector genotoxicity not only increases the risk of oncogenic transformation but has also a negative impact on haematopoiesis by triggering accelerated aging of the hematopoietic system accompanied by accumulation of somatic mutations across hematopoietic lineages in vivo.