Human cord blood-derived primitive CD34-negative hematopoietic stem cells (HSCs) are myeloid-biased long-term repopulating HSCs

Hematopoietic stem cells (HSCs) possess both self-renewal and multi-lineage differentiation abilities and maintain lifelong hematopoiesis. Recent studies have revealed that the murine HSC compartment consists of heterogeneous HSC subpopulations in terms of their lineage-biased differentiation potentials. Dykstra et al. have recently categorized murine HSCs as α-, β-, γand δ-cells, according to the contribution ratio of myeloid or lymphoid cells in the repopulation assays. In addition, other groups have reported that lineage-biased HSCs could be prospectively isolated by their surface immunophenotypes. For example, CD41 murine HSCs have shown to possess a long-term (LT) repopulating capacity and showed a marked myeloid-biased reconstituting capacity. In addition, the murine platelet-primed von Willebrand factor (vWF)-positive HSCs have LT-myeloid-biased lineage repopulation potentials and can self-renew. Furthermore, Morita et al. reported that the murine HSC compartment could be segregated according to the expression levels of CD150 antigen, and CD150 HSCs had the most potent self-renewal activities, as well as the the highest myeloid-biased lineage differentiation potentials. All of these reports demonstrated that myeloid-biased murine HSCs have a LT-repopulating capacity and can produce lymphoid-biased HSCs. Therefore, myeloid-biased HSCs were thought to be most primitive HSCs in the murine HSC hierarchy. Conversely, it has not yet been elucidated whether the human HSC compartment consists of homogeneous or heterogeneous HSC subsets in terms of the lineage-biased differentiation potentials. It has long been believed that human HSCs and hematopoietic progenitor cells (HPCs) exist only in the CD34-positive (CD34) fraction. However, we recently identified CD34-negative (CD34) SCID-repopulating cells (SRCs) in the human cord blood (CB) using an intra-bone marrow injection (IBMI) technique. In addition, we developed high-resolution purification methods for these CB-derived CD34 SRCs using 18Lineage (18Lin)-specific antibodies. These highly purified lineage-depleted CB-derived CD34 cells possessed not only SRC abilities but also colony forming cell (CFC) abilities in the methylcellulose semi-solid culture. These observations clearly demonstrated that human CB-derived HSCs/HPCs (HSPCs) exist not only in the CD34 but also in the CD34 fraction. However, the functional differences between these human CD34 HSPCs are not fully elucidated. Therefore, in this study, we precisely analyzed the differences in the differentiation potentials between CD34 HSPCs in vivo and in vitro. In order to compare the in vivo differentiation potentials of CD34 SRCs, we first performed an SRC assay. The CB-derived 18Lin-negative (18Lin) CD34 cells, both of which contain highly purified CD34 SRCs, were transplanted into NOD/Shiscid/IL-2 Rγc null (NOG) mice using the IBMI technique. Then, the percentages of CD19, CD33 and other types of cells (defined as human CD45, CD19, CD33 cells) in the human CD45 cells produced from CD34 SRCs in the mouse BMs were serially analyzed (schematically presented in Supplementary Figure S1). The human CD45 cell repopulation capacities of both CD34 SRCs were not significantly different (Figure 1a and Supplementary Figure S2). These data were consistent with our previously reported data that both CD34 SRCs possessed comparable human CD45 cell repopulation capacities. However, the differentiation potentials of these CD34 SRCs with regard to the CD33 myeloid cells were clearly different. These CD34 SRCs showed significantly higher rates of CD33 myeloid cell repopulation (Figures 1e and 2). At 5–6, 12 and 18–24 weeks after transplantation, CD34 SRCs showed significantly higher percentages of CD33 cells (74.4, 30.4 and 29.8%, respectively; Po0.01) compared with CD34 SRCs (22.8, 13.1 and 17.7%, respectively) in the mouse BMs (Figures 2g–i). Surprisingly, a number of the mice that received CD34 SRCs showed exclusively human CD33 myeloid cell repopulation at 5 weeks after transplantation (Figure 2d and Supplementary Figure S3B). However, these CD34 SRCs were not myeloid-committed progenitors. Because all of the mice received CD34 SRCs showed multi-lineage human hematopoietic cell reconstitution at 18–24 weeks after transplantation (Figures 1 and 2, Supplementary Figures S2 and S3). We have previously reported that these CD34 SRCs possessed secondary and tertiary (41 year) multi-lineage reconstituting abilities as did CD34 SRCs. The percentages of CD33 cells in the mice that received CD34 SRCs were gradually decreased from the early-to-late weeks after transplantations (Figures 2g–i), and concomitantly the percentages of CD19 B-lymphoid cells increased (Figures 2j–l). In contrast, CD34 SRCs produced significantly higher percentages of CD19 cells compared with CD34 SRCs until 12 weeks after transplantation (Figures 2a, b, j and k). At 5–6, 12 and 18–24 weeks after transplantation, the mean percentages of CD19 cells in the mouse receiving both CD34 SRCs were 66.7 and 21.0% (Po0.01), 80.8 and 64.7% (Po0.05), and 59.7 and 65.6% (P= 0.496), respectively (Figures 2j–l). Therefore, CD34 SRCs predominantly produced CD19 cells in the mouse BM at each time point (Figure 2 and Supplementary Figure S3A). These results were consistent with recently reported data. We next further analyzed the multi-lineage differentiation potentials of CD34 − SRCs. At 18–24 weeks after transplantation, mice were killed and the human hematopoietic multilineage reconstitutions in the mouse left tibia (injection site) were analyzed by FACS. Both CD34 − SRCs could produce comparable levels of CD34 progenitor cells, CD19 B lymphocytes, CD14 monocytes, CD41 megakaryocytes and CD3 T lymphocytes in the murine BM (Figures 1b–d, f and g and Supplementary Figure S2), as we reported previously. However, CD34 SRCs produced higher percentages of CD33 cells compared with those of CD34 SRCs (Figures 1e and 2g–i), as above-mentioned. On the contrary, CD34 SRCs produced a significantly higher percentage of CD235a cells compared with CD34 SRCs (Figure 1h). Collectively, these results demonstrated, for the first time, that human CB-derived CD34 SRCs are myeloid-biased SRCs. Citation: Blood Cancer Journal (2015) 5, e290; doi:10.1038/bcj.2015.22