CD26 Inhibition Can Aid the Homing of Cytokine Activated Mobilized Peripheral Blood (MPB )CD34⁺ Cells to the Bone Marrow (BM) but a Ligand Dependent Attachment Defect Prevents Their Long Term Retention and Subsequent Engraftment

Abstract 141

Short term cytokine exposure reduces the engraftment potential of haematopoietic stem and progenitor cells (HSPC). We have previously shown, in MPB CD34⁺ cells, that this occurs in conjunction with reduced short term homing to the BM of irradiated NOD/SCID animals, and is evident by 4 hours of culture (Ahmed at al, Blood 2004; 103:2079). Homing and engraftment of HSPC is dependent on the SDF-1/CXCR4 axis, that is negatively regulated by CD26, a cell-surface peptidase that cleaves SDF-1. Thus, blockade of CD26 with diprotin A improves engraftment of cord blood HSPC. CD26 levels are low on MPB CD34+ cells, but increase on cytokine exposure, an effect that may account for reduced homing function. We tested the effect of diprotin A treatment on the homing and engraftment of MPB CD34⁺ progenitors in irradiated NOD/SCID mice. Cytokine exposure (SCF, FL, IL3, IL6 for 48–72 hrs) reduced BM homing of progenitors (to 32±7% of uncultured cells, p=0.031). Treatment with Diprotin A significantly improved the homing of cultured progenitors (p=0.0002 cf non-treated cells), to the levels seen in uncultured cells (NS cf homing of uncultured cells with or without Diprotin A). Despite this increase in levels of homing to the BM, long term engraftment of cultured progenitors is not rescued by CD26 blockade, suggesting a defect beyond the initial step of homing to BM.

To assess the BM attachment of transplanted cells, we used the intrabone (IB) assay, injecting cells directly into the tibiae of irradiated recipients thus bypassing the need for homing from the systemic circulation. In comparison to intravenously (IV) injected cells, IB delivery improved engraftment of both cultured and uncultured MPB CD34⁺ cells. At a cell dose of 10⁶, median engraftment of uncultured cells was 13.27% (range 2.43–49.1, n=5) by IB injection compared with 1.5% (0.1–4.6, n=6) by IV delivery (p<0.01). The engraftment defect of cultured cells, however, persisted in IB transplantation. At a cell dose of 10⁶/animal, median engraftment of cultured cells was 0.73% (0.001–10.93, n=12), compared with 13.27% (2.43–49.1, n=5) for uncultured cells (p<0.05). Injecting twice the cell dose for cultured cells (expanded equivalent number) did not rescue this defect. Engraftment of cultured cells in the injected bone (0.055%, range 0.001–9.28, n=6) was significantly lower than that of uncultured cells (14.4%, 0.08–71.2, n=9, p<0.05). These findings suggest that cytokine exposure reduces the ability of MPB CD34+ cells to be retained in the BM.

To study this attachment defect directly, we developed an ex-vivo model where CD34⁺ cells are incubated in long bones of irradiated, 3-week old Sprague Dawley rats after resident HSPC are removed by vigorous flushing. Irradiation, flushing protocols and cell doses were optimized to ensure sensitivity and reproducibility of the assay. Progenitor adherence was assessed by colony assays on infused and recovered attached cells, correcting for animal weight. In this model, cultured progenitors displayed reduced attachment: median number of attached progenitors, 182.1 (23–384, n=26), compared with 385.4 (49–1124, n=30) for uncultured cells (p<0.0001). Attachment of both uncultured and cultured progenitors was reduced by CXCR4 blockade (42%, and 53% reduction respectively, p<0.05 for both), confirming the in vivo relevance of this model. This model provides a novel system to directly study and manipulate the lodgment of HSPC in the BM.

Next, we tested the effect of cytokine exposure on ligand-specific adhesion of MPB progenitors. CD34⁺ cells were incubated on immobilized ligands and progenitor adhesion assessed from the clonogenic output of non-adherent cells. Progenitor adhesion to several putative niche ligands was significantly reduced following cytokine exposure. Specific adhesion of uncultured progenitors to N-cadherin was 31.8±2%, compared to 19.7±3.2% for cultured progenitors (p=0.0058). Cytokine culture also reduced specific adhesion to osteopontin and VCAM-1 (p=0.0025 and p=0.0164 respectively), but not to fibronectin, suggesting that reduced adhesive function of cultured cells is not a global defect.

We conclude that whilst short term homing of cultured MPB CD34+ cells to the BM can be improved by CD26 blockade, the resulting long term engraftment defect remains. This defect is at least partly related to altered adhesive interactions of cultured cells to ligands within the BM.