Foxo3 transcription factor is a critical regulator of hematopoietic stem cell (HSC) quiescence and hematopoietic and leukemic stem cell maintenance. In particular, loss of Foxo3 has been shown in models of both acute and chronic myeloid leukemia to prevent the maintenance of leukemogenesis, indicating that devising strategies to inhibit Foxo3 will be of critical therapeutic value. Nonetheless mechanisms that control Foxo3 activity in HSC remain unknown. FoxO are regulated by post-translational modifications including phosphorylation, acetylation and redox modulation that together determine FoxO subcellular localization and activity. In response to growth factors stimulation FoxO are phosphorylated by AKT that promote their cytosolic localization and inhibition of transcriptional activity. Interestingly, in normal bone marrow HSC and in leukemic stem cells Foxo3 is constitutively nuclear despite readily detectable phospho-AKT, strongly suggesting that negative phosphorylation may not be the sole Foxo3 regulatory mechanism in these stem cells. While acetylation of Foxo3 is linked to oxidative stress, and oxidative stress is known to activate Foxo3, whether acetylation activates or inhibits Foxo3 remains unclear and it is proposed to be context-dependent. Therefore, we sought out to determine how acetylation of Foxo3 impacts its function in HSC. To address this question we generated Flag-tagged-Foxo3 mutants mimicking (lysine to glutamine) or abrogating (lysine to arginine) acetylation of all five putative acetylation sites using a PCR-based site-directed mutagenesis strategy and then cloned these mutants into the retroviral MSCV-IRES-GFP (MIG) vector for generation of retroviral supernatant and efficient transduction of bone marrow mononuclear cells. We first evaluated subcellular distribution of Foxo3 mutants in human embryonic kidney (HEK)-293 cells. Interestingly Foxo3 is mainly cytoplasmic in HEK-293 cells but hydrogen peroxide (H2O2) treatment induces Foxo3 nuclear translocation. Ectopic expression of Foxo3 mutants mimicking acetylation where five putative acetyl-lysine residues (5KQ) are mutated in HEK-293 cells and subsequent H2O2 treatment impairs Foxo3 nuclear localization (>50% reduction of nuclear Foxo3), although to a lesser extent than some of the single mutants, indicating that distinct acetylated lysines may impact differently Foxo3 activity. Moreover, protein expression of Foxo3 acetylation-mimic mutants is increased as compared to wild type Foxo3 in HEK-293 cells suggesting that acetylation affects Foxo3 protein stability. In order to analyze the impact of Foxo3 acetylation in vivo, we transduced bone marrow mononuclear cells (BMMC) freshly isolated from 5-fluoracil (5-FU) treated mice with acetyl-lysine mimic mutant Foxo3. GFP-positive transduced BMMC were FACS-sorted, cytospun onto glass slides and analyzed by immunostaining with anti-Flag antibody, in order to discriminate exogenously expressed Foxo3-WT versus Foxo3-mutants. In agreement with HEK-293 results, overexpression of Foxo3 mutant in which five acetylation sites were mimicked (5kQ) showed a significant decrease in nuclear localization, indicating that acetylation may lead to cytoplasmic Foxo3 and ultimately abrogate its activity in BMMC. Importantly, BMMC overexpressing Foxo3-5kQ injected into lethally irradiated mice (Colony Forming Unit-Spleen – CFU-S) produced significantly less spleen colonies at day 12, indicating that the Foxo3 acetyl-lysine mimic mutant may function as a dominant-negative Foxo3. We lastly investigated Foxo3 activity in a SIRT1 conditionally knockout mouse model since SIRT1 is the main Foxo3 deacetylase. In agreement with results described here we found that in SIRT1−/−LSK cells, Foxo3 is mainly cytoplasmic. Importantly and consistent with data generated from ectopic expression of Foxo3 acetyl-lysine mutants, Foxo3 protein expression is significantly (two-fold) increased in SIRT1−/− Lin−Sca1−cKit+ myeloid progenitors. These findings strongly support the notion that acetylation of Foxo3 abrogates its function in hematopoietic stem and progenitor cells. They also suggest that acetylation of Foxo3 impacts its protein stability. Altogether these results are important for understanding the mechanism of regulation of HSC activity and are likely to have significant therapeutic value.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.