Oxygen dependency as a critical regulatory axis affecting hematopoietic stem and progenitor cell function

Abstract Hematopoietic stem (HSCs) and progenitor cells (HPCs) are responsible for lifelong replenishment of blood and immune cells and are a source of hematopoietic cell therapies. A comprehensive understanding of extrinsic factors that affect the function and potency of HSCs/HPCs is still desired. HSCs/HPCs are exposed to O2 tensions ranging from <1% to 21% as they function in various anatomical sites or are harvested for clinical applications. Understanding how the various O2 tensions that HSCs/HPCs are exposed to throughout their lifespan affects their function will provide insights into their regulation and represents a potential tool to modify their potency for therapies. We hypothesized that cells may utilize variable local O2tensions to perform distinct functions. We first found that there are significantly different proportions of human HSCs/HPCs in cells isolated from umbilical cord blood, bone marrow, and mobilized peripheral blood, which exist at different oxygenations. This suggests that niche oxygenation may be a factor regulating cell function. To test this, we performed ex vivo expansion assays and colony forming unit (CFU) assays in controlled O2 tensions using CD34+ cells from all three donor sources. Total CD34+ cells and lineage determined progenitors exhibit increased total expansion in high physiologic (14%) and extra physiologic (21%) O2 compared to low physiologic (1-5%). Conversely, HSCs and primitive HPCs showed increased expansion in lower physiologic O2 (5%). Cord blood HPCs showed significantly increased CFU capacity in 5% O2. This suggests that primitive HSCs/HPCs with higher potency are better maintained in low physiologic O2. We tested this with mouse models of expanded cord blood transplantation. CD34+ cells expanded in 1-5% then transplanted to mice showed increased early neutrophil recovery, significantly higher than input, with 1% maintaining the highest proportion of cells driving early neutrophil recovery. All physiologic O2 tensions (1-14%) maintained cells that yielded significantly higher balanced lymphoid/myeloid recovery at intermediate engraftment time points compared to ambient air. Thus, low physiologic O2 tensions maintain functionally potent HSCs/HPCs with early and intermediate engrafting potential. To identify mechanisms underlying this O2-dependent functional shift, we performed single cell RNA-seq. We found that HSCs/HPCs significantly change their transcriptomes during ex vivo culture, but low physiologic O2 partially mitigates this change. Cells at low O2 (1-3%) had lower expression of stress associated gene programs. We validated these stress indicators showing that cells in low O2 exhibited fewer cycling cells, increased quiescent cells, decreased ROS, increased glycolysis to oxidative phosphorylation ratio, and decreased responses to mitochondrial perturbations. We confirmed by western blot that proteins encoded by O2 dependent genes were de-stabilized in high O2, including mTORc pathway proteins like BNIP3, LDHA, and ENO1. Rapamycin treatment to inhibit this mTORc program partially rescued low physiologic maintenance of potent HSC/HPC populations. Finally, we utilized O2-dependent gene signatures integrated with in vivo data to identify genes associated with HSC/HPC functional potency, including DDIT4, PRSS2, and AVP. To validate this, we performed RNA-seq on cord blood units with known mouse repopulating frequencies and found enrichment of a subset of O2-dependent genes in HSCs with high repopulating potential. Finally, siRNA-mediated knockdown of several associated O2-dependent genes led to reduced HPC CFU capacity, confirming their importance in hematopoietic function. Thus, HSC/HPC function is acutely sensitive to local O2 tension. HSCs/HPCs may utilize local low O2tensions to maintain the pool of HSCs in the bone marrow or rapidly initiate differentiation and expansion of HPCs outside the bone marrow in the presence of higher O2. Additionally, our O2-dependent transcriptomic map of HSCs/HPCs, reveals potential targets for altering cell function for clinical use. Our findings also have immediate translational implications in the context of expanded cord blood transplantation and other cell therapies that require ex vivo manipulation, like gene therapy, where O2 may be a critical factor in balancing ex vivo expansion while maintaining enough potent cells for early recovery and long-term hematopoietic durability.