See Supplemental Experimental Procedures for a more detailed

See Supplemental Experimental Procedures for a more detailed description of methods used in this study.
Author Contributions
Acknowledgments
Introduction Fanconi anemia (FA) is an autosomal recessive disorder associated with birth defects, progressive bone marrow failure, hematopoietic stem cell (HSC) depletion, and cancer predisposition. FA is caused by a disrupted FA-BRCA network and is genetically heterogeneous, with at least 16 complementation groups and respective genes identified so far (Kim and D’Andrea, 2012). Progressive bone marrow failure is the primary cause of morbidity and mortality in FA patients (Kutler et al., 2003). Most patients develop marrow dysfunction within the first decade of life. The symptoms range from mild cytopenia in any lineage to severe aplastic anemia, often initially with thrombocytopenia (Shimamura and Alter, 2010). Red cell macrocytosis is quite common in FA patients and usually precedes the onset of thrombocytopenia. HSC transplantation is the only curative treatment for bone marrow failure in FA. However, androgens have also been widely used to treat cytopenia in FA, especially for patients unable to proceed to transplant or patients with high transplant risk. The most commonly used androgen is oxymetholone (OXM), which is an anabolic-androgenic steroid and a synthetic derivative of testosterone (Shimamura and Alter, 2010). Androgen therapy raises blood counts in ∼50% to 70% of individuals with FA and also works for other forms of aplastic anemia (Dokal, 2003). Despite a long history of androgen use in bone marrow failure syndromes, the mechanism whereby these molecules boost blood counts remains enigmatic (Chute et al., 2010). It has been suggested that androgens stimulate erythropoiesis through an increase in the production of erythropoietin (EPO). However, more recent studies have found no close correlation between androgens and EPO levels (Chute et al., 2010), leading others to speculate that androgens might have a direct effect on bone marrow (T’Sjoen et al., 2005). One recent in vitro study suggested that androgens act by increasing telomerase activity and extending the lifespan of CD34+ stem/progenitor ochratoxin a (Calado et al., 2009). Multiple murine models of FA are available. Among them, Fancd2 mice, Fancp mice, and Fancc-Fancg double knockout mice represent human FA patient phenotypes more closely than the others (Crossan et al., 2011; Houghtaling et al., 2003; Parmar et al., 2010; Pulliam-Leath et al., 2010; Zhang et al., 2010). Fancd2 mice recapitulate the characteristic tumor susceptibility of FA and show an ∼2-fold decrease in hematopoietic stem and progenitor cell (HSPC) populations and a very poor long-term repopulating capacity of bone marrow (Parmar et al., 2010; Zhang et al., 2010). Despite this, the mice have no obvious anemia in their peripheral blood at age 6 months, except for lower platelet counts. Here, however, we found that 18-month-old Fancd2 mice developed spontaneous pancytopenia. We then set out to investigate how OXM benefits FA patients using this aged Fancd2 mouse model.
Results
Discussion Although OXM has been used for patients with aplastic anemia including FA for many years, its mechanism of action remains poorly understood. The studies described herein provide significant information on how this widely used drug affects hematopoiesis and exerts therapeutic benefit. We found that OXM significantly reduced quiescence and promoted proliferation in KSL hematopoietic stem/progenitor cells. This effect was not specific to FA HSPCs, but affected WT mice just as strongly. The percentage of actively cycling cells increased by at least 50% in both genotypes. Interestingly, the effect of OXM on the cell cycle was specific only to HSPC in both mutant mice and controls. Other hematopoietic lineages in the bone marrow did not divide more frequently in response to the drug. This finding indicates that OXM administration has a direct effect on the HSPC compartment and is consistent with the fact that the drug is known to increase the numbers of all blood lineages in humans, including red blood cells, platelets, and neutrophils (Shimamura and Alter, 2010). Although the HSPCs clearly cycle more rapidly on OXM, the number of immunophenotypically defined stem cells did not increase in either mutants or WT mice. This suggests that the newly generated cells rapidly feed into the progenitor compartment rather than generating new stem cells by self-renewal. In fact, our data show that chronic OXM administration eventually results in stem cell exhaustion. The long-term repopulating ability of stem cells from OXM-treated mice was significantly lower than those of placebo-treated controls. This was not unexpected, given the fact that multiple genetic mutations associated with reduced stem cell quiescence have been shown to result in eventual stem cell depletion (Orford and Scadden, 2008). We have previously shown that FA HSPCs already have an accelerated cell cycle and interpreted this finding as a homeostatic mechanism that compensates for the increased loss of stem cells in this disease (Zhang et al., 2010). The administration of OXM can further enhance this compensatory mechanism, at least temporarily, and indeed does improve multiple hematological parameters in FA mice. However, our data predict that OXM will not provide a permanent rescue of hematopoiesis and that a more definitive method of treatment will be eventually needed, even in good early responders. In fact, many FA patients eventually become androgen resistant even if their initial response was positive (Shimamura and Alter, 2010). It is possible that HSCs in these patients have been completely depleted at later stage of the treatment.