Abstract
CD34 has traditionally been used as a marker to enrich human bone marrow stem cells for clinical bone marrow transplants for over five decades. However, several studies support that the most primitive hematopoietic stem cells (HSCs) responsible for long-term hematopoiesis-long-term HSCs (LT-HSCs) -are found within the CD34 negative (CD34 neg) fraction of bone marrow cells. LT-HSCs are challenging to study due to their scarcity and lack of specific markers to identify them. Once these stem cells enter the circulation, they must find their way, through multistep process controlled by a variety of cell surface adhesion molecules, into the bone marrow. In contrast to the CD34 pos stem cells, studies have suggested that LT-HSCs have trouble migrating and engrafting when introduced intravenously. In order to better understand why these deficiencies exist, we set out to fully elucidate the adhesion mechanisms used by various murine hematopoietic stem cell populations to migrate. To this end, we assayed the homing effectors that are either necessary in the first step, i.e., the cell surface adhesion molecule Sialyl Lewis X (sLe x) and its interaction with selectins-mainly E-selectin-on the bone marrow endothelium, or the second step, i.e. the chemokine receptor CXCR4 and its binding to the bone marrow stroma resident chemokine SDF-1α, of cell migration.
Specifically focusing on murine populations of short term HSCs (ST-HSCs; Flk2 negCD34 pos) and LT-HSCs (Flk2 negCD34 neg), flow cytometric and gene expression analysis clearly illustrated distinctive expression patterns of homing effectors with the highest contrast observed in sLe x where Flk2 negCD34 neg HSCs exhibited low expression (< 10%) and Flk2 negCD34 pos HSCs exhibited much higher expression (> 60%) which correlated to higher binding to E-selectin. Of the common E-selectin ligands identified, both Flk2 negCD34 pos and Flk2 negCD34 neg HSCs expressed CD44 and PSGL-1 to equivalent levels while CD43 was expressed at significantly lower levels on Flk2 negCD34 neg HSCs. Moreover, our data indicates that the variation observed in sLe x expression/E-selectin binding correlates with differences in the expression of key glycosyltransferase genes involved in the creation of the sLe x epitope on these E-selectin ligands. mRNA expression analysis of glycosyltransferases revealed that, in general, sialyltransferase (ST3Gals) genes were expressed at higher levels in Flk2 negCD34 neg HSCs while only one fructosyltransferase (FT4) gene was expressed at higher levels in this HSC population. Thus, these data suggest that α2,3 sialylation is intact in Flk2 negCD34 neg HSCs but α1,3 fucosylation is lacking. The expression levels of homing molecules involved in the stages of cell migration after E-selectin binding were also assessed. No differences in the expression of the integrins CD49d, CD49e and CD29 were observed between Flk2 negCD34 neg and Flk2 negCD34 pos HSCs. Interestingly, higher levels of the chemokine receptor CXCR4 were measured on the surface of Flk2 negCD34 pos HSCs compared to Flk2 negCD34 neg HSCs. Likewise, this was reflected in a much lower ability of the Flk2 negCD34 neg HSCs to migrate towards the CXCR4 ligand, SDF-1, in a transwell assay.
Based on these findings that show that both E-selectin binding and CXCR4 mediated migration are compromised in Flk2 negCD34 neg HSCs, we sought to enhance this LT-HSC populations ability to migrate using well-known treatments including recombinant human fucosyltransferase 6 (rhFVI) to create the glycans necessary for selectin binding, as well as the CD26 inhibitor (Diprotin A; Dip A) to enhance the binding of chemokines to their receptors.
Interestingly, although both CD26 inhibition and cell surface fucosylation remarkably enhanced the expression of homing molecules in both HSC populations in vitro, only pretreatment of Flk2 negCD34 neg HSCs with Dip A significantly enhanced engraftment in vivo after transplantation into C57BL/6N mice.
This study represents the first direct analysis of E-selectin ligand expression on murine LT-HSC populations and strongly support that by temporarily enhancing the expression of migration-associated molecules on the surface of these cells, prolonged and efficient HSC engraftment can be achieved. We hope our finding will potentially shed light on methods to optimally utilize these very valuable long-term HSCs in clinical bone marrow and cord blood transplants worldwide.
No relevant conflicts of interest to declare.
