HSC Exit From Dormancy Provokes De Novo DNA Damage, Leading To Bone Marrow Failure If Unresolved By The Fanconi Anemia Pathway

Long-term quiescence has been proposed to preserve the genomic stability of hematopoietic stem cells (HSCs) during aging. The current models of HSC aging are limited in their ability to observe both DNA damage in vivo and the consequences of this damage upon hematopoiesis. Fanconi Anemia (FA) is a hereditary multisystem disorder, characterized by defective DNA damage response and progressive bone marrow failure in most patients. However, the existing genetic models of FA do not develop aplastic anemia, suggesting that cell-extrinsic factors may play a causal role. We sought to identify whether physiologic mediators of HSC activation could be used as agonists to provoke DNA damage and HSC attrition in vivo.

Mice were treated with a range of agonists that promote the in vivo exit of HSC from a dormant state into active cycling (polyI:polyC; Interferon-α; G-CSF; TPO; and serial bleeding). Highly purified HSC demonstrated a rapid 3-5-fold induction of DNA damage after treatment with all agonists (p<0.01), as assessed by both enumerating γ-H2AX foci and by alkaline comet assay. Mechanistically, stress-induced exit from quiescence correlated with increased mitochondrial metabolism in HSC, as evaluated by elevated mitochondrial membrane potential (2-fold increased, p<0.01) and superoxide levels (1.5-fold increased, p<0.05). Critically, we could directly implicate these reactive oxygen species in DNA damage as we observed a 1.4-fold increase in 8-Oxo-dG lesions in HSC that had been activated into cycle in vivo(p<0.05). At 48 h post-treatment, γ-H2AX levels began to decrease and this repair was concomitant with an induction of the FA signaling pathway in HSC, as demonstrated by both increased levels of FA gene expression and elevated FANCD2 foci (4-fold induction, p<0.01).

Treatment of Fanca^-/- mice with polyI:polyC led to a HSC proliferative response comparable to wild type (WT) mice but resulted in a 2-fold higher level of activation-induced DNA damage (p<0.05), demonstrating that this repair pathway is involved in resolving activation-induced DNA damage. Four rounds of serial in vivo activation led to a permanent depletion of the most primitive label-retaining Fanca^-/- HSC and this correlated with a 4-fold depletion of functional HSC (p<0.01) as defined by competitive repopulation assays. Subsequent rounds of HSC activation with polyI:polyC resulted in the onset of a severe aplastic anemia (SAA) in 33% of treated Fanca^-/- mice but not in any of the WT controls. SSA was characterized by a dramatic reduction in bone marrow (BM) cellularity, profound thrombocytopenia (21-246x10⁶ platelets/ml), leukocytopenia (0.4-0.5x10⁶ WBC/ml), neutropenia (0.03-0.1x10⁶/ml) and anemia (1.5-2.3 g/dL Hb). Examination of BM HSC/progenitors demonstrated nearly complete loss of HSC, MPP, CMP and CLP (depletion of ≥33x, 8x, 4x and 12x respectively compared to PBS-treated Fanca^-/-controls).

Taken together, these data demonstrates that enforced exit from dormancy in vivo leads to de novo DNA damage in HSC, which is repaired by activation of a FA-dependent DNA damage response. Furthermore, the highly penetrant bone marrow failure observed in Fanconi anemia patients can be recapitulated by the serial application of a physiologic HSC activating signal to Fanca^-/- mice. This suggests that the BM failure in FA may be caused by an aberrant response to HSC activation, most likely during exposure to infection or other physiologic stressors. These data provides a novel link between pro-inflammatory cytokines, DNA damage and HSC dysfunction and may have important clinical implications relevant to both prevention of BM failure in FA and in the study of age-related hematopoietic defects in non-FA patients. Moreover, these data provide the first evidence that FA knockout mouse models accurately recapitulate and provide novel insights into the etiology of BM failure in patients with FA.