Role of Osteoclasts in Regulation of Human Hematopoietic Stem Cell Niche and Fate.

Recent studies indicate that osteoblasts play an important role in maintaining hematopoietic stem cells (HSCs) niche in the bone marrow microenvironment. The aim of study was to test the effect of osteoclasts on the fate of HSCs in a long term co-culture assay. To generate osteoclasts, peripheral blood mononuclear cells from mobilized donors were cultured for 6–10 days in αMEM media supplemented with 10% FCS, M-CSF and RANKL. After removal of non-adherent cells, the cultures contained 95% multinucleated osteoclasts and their precursors. These osteoclasts expressed TRAP and formed resorption pits on bone slices (Yaccoby et al., Cancer Res., 2004). CD34+ cells were purified from donor PBSCs and cord blood using immunomagnetic beads separation (>95% purity). Adult and cord blood HSCs were co-cultured with osteoclasts for up to 3 and 10 months, respectively, in media lacking any cytokines. Because osteoclasts do not survive long without M-CSF and RANKL, the HSCs were transferred to fresh osteoclast cultures every 6–10 days. Unlike their tight adherence to stromal cells, HSCs did not adhere to the osteoclasts and were easily recovered from co-cultures by gentle pipetting. Following 1 to 3 weeks of co-culture, committed HSCs rapidly differentiated into various hematopoietic cell lineage, followed by phagocytosis of terminal differentiated hematopoietic cells by the osteoclasts. The remaining HSCs were highly viable (>90% by trypan blue exclusion) and gradually lost their CD34 expression, so that the cultures contained subpopulations of HSCs expressing CD34−/lowCD38+ and CD34−/lowCD38−. Quantitive real time RT-PCR (qRT-PCR) revealed loss of expression of CD34 and reduced expression of CD45 by HSCs co-cultured with osteoclasts longer than 6 weeks. Variable expression of CD34 on HSCs was previously reported in murine but not human HSCs (Tajima et al., Blood, 2001). The co-cultured HSCs showed reduced capacity of generating in vitro hematopoietic colonies, and did not differentiate into osteoclasts upon stimulation with M-CSF and RANKL. We next tested the long term engraftment of these co-cultured HSCs in 2 animal models. In the first model, cord blood and adult HSCs from 2 donors recovered after >6 weeks in co-culture were injected I.V. into irradiated NOD/SCID mice. In the second novel model, co-cultured cord blood and adult HSCs from 2 donors were injected directly into rabbit bones implanted subcutaneously in SCID mice (SCID-rab model), 6–8 weeks after rabbit bone implantation. After 2–4 months, 10%±3% human CD45-expressing cells were identified in the NOD/SCID mice femora and 8%±4% in the SCID-rab mice rabbit bone. Our study suggests that osteoclasts promote rapid differentiation of committed HSCs and induce conversion of CD34+ cells to CD34− stem cells with self renewal potential. Intriguingly, long term co-culture of primary CD138-selected myeloma plasma cells (n=16) with osteoclasts resulted in dedifferentiation of tumor cells from a mature CD45− phenotype to an immature, CD45-expressing cells, suggesting a common mechanism of osteoclast-induced HSC and myeloma cell plasticity. This indicates that osteoclasts are important bone marrow component regulating human HSC niche, plasticity and fate.