5-Azacytidine Depletes Hematopoietic Stem Cells and Synergizes with an Anti-CD117 Antibody to Augment Engraftment of Donor Stem Cells in Immunocompetent Mice

Monoclonal antibody (mAb)-targeting of CD117 (c-Kit) present on hematopoietic stem cells (HSC) is an emerging conditioning strategy for hematopoietic cell transplantation (HCT) to replace standard-of-care alkylator and radiation therapy used to deplete recipient HSC and permit donor HSC engraftment. In preclinical and clinical studies anti-CD117 mAbs, which inhibit stem cell factor (SCF) from binding to CD117, have shown single agent efficacy in permitting donor HSC engraftment in immunocompromised recipients. However, preclinical data suggest that anti-CD117 mAb, ACK2, used as a single agent is insufficient to enable HSC engraftment in immunocompetent mice. Hence, there is a desire to potentiate the effect of naked anti-CD117 mAbs to permit targeted elimination of host HSC for a range of immune competent bone marrow (BM) states, from monogenic disorders, e.g., hemoglobinopathies, to BM clonal malignant diseases. Here, we show that ACK2 synergizes with the widely used hypomethylating agent, 5-Azacytidine (AZA), to deplete HSC and enable robust long-term HSC engraftment in immunocompetent mice.

The cell depleting effect of AZA or ACK2+AZA was first tested on mature and primitive hematopoietic cells of C57BL/6 (B6) mice. Mice received AZA at 5mg/kg i.p. for 5 days either alone or combined with ACK2 at 500 µg i.v. 5 days prior to AZA. BM and spleen were harvested days 6, 10 and 20 after the first AZA dose and analyzed by extended immunophenotype. For HCT studies both congenic [B6 (H-2^b, Thy1.1, CD45.1) into B6 recipients (H-2^b, Thy1.1, CD45.1/CD45.2)] and allogeneic [B6 (H-2^b, Thy1.1, CD45.1) into BALB.B recipients (H-2^b, Thy1.2, CD45.2)] transplantation mouse models were tested.

Single agent AZA induced rapid depletion of CD117+ cells in the BM, including LT-HSC (Lin^-Sca⁺Kit⁺[LSK]CD150⁺CD48^-) and ST-HSC (LSKCD150^-CD48^-). In vivo HSC depletion by AZA has not been previously reported, and was surprising given that most HSC are quiescent, and non-proliferating cells are known to be insensitive to the toxicities of AZA. Examination of HSC populations in the spleen did not suggest that mobilization of LT- or ST-HSC to secondary organs caused the decrease in BM. Rather, AZA-induced myelosuppression was followed by rapid expansion of myeloid-biased MPP and increased HSC proliferation as assessed by Ki67 intracellular protein expression. We hypothesized, that the in vivo HSC-depletive effects of AZA might result from increased sensitivity of proliferating HSC to AZA. Hence, the in vitro effects of different AZA concentrations were tested on proliferating HSC. In both mouse and human HSC cultures, supplemented with SCF and TPO, we observed that AZA reduced the number of cells in a dose dependent manner and decreased cell viability at concentrations of ≥ 0.5 µg/ml.

In vivo, combined ACK2+AZA led to more robust HSC depletion and prolonged time to HSC recovery compared to AZA alone. We then evaluated if blockade of CD117 with its ligand SCF by ACK2 was the cause of this prolonged HSC-depletion. 2B8 mAb, which also binds mouse CD117, but only partially inhibits CD117/SCF signaling, did not result in augmented in vivo HSC depletion (Figure 1A). We further tested if immunocompetent mice treated with ACK2+AZA could be “rescued” with exogenous recombinant SCF. SCF-treatment resulted in faster recovery of LT-HSC. Together these studies suggest that the synergy of ACK2+AZA on the depth and duration of HSC-depletion depends of the disruption of CD117/SCF binding.

Finally, we tested ACK2+AZA conditioning on donor stem cell engraftment in immunocompetent mice. Transplantation of 15-20x10⁶ WBM cells following ACK2+AZA conditioning led to significantly increased stable mean multilineage donor chimerism of >60% in the congenic (Figure 1B) and >20% in the allogeneic setting. Single agent AZA enabled sustained donor engraftment with levels of myeloid donor chimerism of 10-15%, serving as proof of principle that AZA treatment can indeed clear HSC niches and permit engraftment of true HSC. Similar to WBM grafts, 5x10⁴ purified HSC (LSK) engrafted successfully in the ACK2+AZA group, reaching long-term (26 weeks) mean donor myeloid chimerism >25% in both congenic and allogeneic settings.

Our studies reveal a previously unknown effect of AZA on BM HSC and provide a platform for the clinical use of a novel less toxic non-myeloablative conditioning strategy for HCT, which is aimed at the rapid translation into clinical practice.