Rap1 Gtpases Regulate the Retention and Engraftment of Hematopoietic Stem and Progenitor Cells

Signaling molecules that control the homing and mobilization of hematopoietic stem and progenitor cells (HSC/Ps) are poorly understood. Rap1, a small-molecular-weight GTP-binding protein belongs to the Ras-like superfamily of GTPases and regulates several signal transduction cascades. Rap1 cycles between a GDP-bound inactive and a GTP-bound active form and exists in two isoforms - Rap1a and Rap1b, which have been implicated in the regulation of actin based functions in non-hematopoietic cells. Although Rap1 has been involved in regulating several hematologic disorders including chronic lymphocytic leukemia, myeloproliferative stem cell disorders, polycythemia vera and sickle cell anemia, its role in the development and function of HSC/Ps has not been investigated. We have generated a mouse model in which both Rap1a and Rap1b isoforms were conditionally deleted in HSC/Ps individually or in combination (double knockout; DKO). Our results demonstrate that deletion of both isoforms of Rap1 results in profound mobilization of primitive hematopoietic stem cells in peripheral blood. In the bone marrow, Rap1ab deficiency shows increased frequency of LSK cells, HPC-1 (LSK CD150-CD48+), HPC-2 (LSK CD150+CD48+) along with an increase in granulocyte-macrophage progenitor cell (GMP) population. Furthermore, spleen size and cellularity were significantly enhanced in DKO mice relative to controls. We hypothesized that Rap1 plays an essential role in regulating the retention of HSC/Ps in the bone marrow (BM) and that loss of Rap1 might inhibit the interaction of HSC/Ps with the BM niche cells, leading to egress of HSC/Ps and thus creating empty space(s) in the marrow for enhanced engraftment of donor derived cells when transplanted under non-myeloablative conditions. To test this, we performed BM transplantation using Rap1ab DKO mice as recipients and WT GFP expressing HSC/Ps as donors in the absence of any myeloablative conditioning. Our long-term engrfatment results showed significantly greater donor derived reconstitution of GFP positive cells in peripheral blood of DKO recipients compared to WT controls (WT: 19.2% vs DKO: 82.18% n=3, *p<0.05), suggesting that loss of Rap1ab creates functional open niche(s) in the BM due to mobilization of endogenous HSC/Ps. To better understand the mechanism behind this observation and to determine whether the GFP donor cells localize closer to the endosteal or vascular niche, we transplanted GFP positive cells into unconditioned (non-myeloablative) WT and Rap1ab DKO mice as described above. We measured the median distance of engrafted GFP cells from the bone surface and vasculature as a measure of proximity utilizing intravital microscopy. DKO recipients, transplanted with WT HSC/Ps preferentially localized to the vascular niche compared to control WT recipients (WT: 8µm vs DKO: 3 µm) and compared to osteoblastic niche, which was comparable in the two recipients, suggesting that GFP+ donor HSC/Ps preferentially localize and engraft near vascular niches providing indirect evidence to suggest that loss of Rap1ab leads to egress of hematopoietic cells from the vascular niche as opposed to osteoblastic niche. We next assessed the potential of Rap1ab deficient cells to engraft in a lethally irradiated host in a competitive repopulation assay. Rap1ab DKO HSC/Ps showed a defect in engraftment as well as multi-lineage reconstitution when transplanted into lethally irradiated hosts compared to WT controls. The defect in engraftment was largely due to impaired homing of DKO HSC/Ps. To assess which specific isoform of Rap1 is essential for mobilization and engraftment/homing of HSC/Ps, we induced deletion in Rap1a and Rap1b separately (single knock out mice) and assessed these mice for peripheral blood cell counts. We found no significant changes in the peripheral WBC counts in single Rap1a KO mice relative to controls; and only a modest increase in single Rap1b KO mice; suggesting that mobilization of HSC/Ps was relatively unperturbed in these mice and requires the loss of both isoforms of Rap1. In contrast, engraftment of HSC/Ps derived from the single KOs of Rap1a and Rap1b was impaired to the same extent as DKO HSC/Ps. These data suggest that loss of single Rap1 isoform contributes similarly to the engraftment of HSC/Ps, whereas the combined loss of both isoforms is required for efficient mobilization of HSC/Ps.