Abstract

The kinetics of HPC engraftment correlate to a significant degree with the efficiency of bone marrow (BM) homing. We recently documented that fresh BM-HPC, rendered SDF-1 unresponsive or Gi-signaling refractory, homed normally, due to efficient retention through α4-integrin/VCAM 1. However, when deletion of α4-integrin or VCAM-1 was combined with SDF-1/CXCR4- or Gi-protein-signaling blockade, BM homing was drastically reduced, uncovering non-dominant contributions of SDF-1/CXCR4 and Gi-proteins. When, in contrast, cytokine (SCF) incubated HPC were transplanted, loss of CXCR4- or Gi-signaling alone significantly reduced BM homing, with little compensation by α4/VCAM-1 and endothelial selectins. These studies depicted a flexible hierarchy of cooperating homing pathways, in which the dominance of the individual pathway contribution is shifted in response to the cytokine milieu. Based on these data, we reasoned that different donor transplant sources might differ in the relative contribution of each homing pathway. This concept was tested with G-CSF mobilized (MPB) and fresh steady-state BM (ssBM) HPC. To assess the homing properties of these two cell sources, we controlled the number of transplanted colony-forming cells (CFU-C). We found that murine or human MPB-HPC homed significantly better than the respective ssBM-HPC: 1.16±0.04% per femur vs. 0.80±0.03% for murine cells, and 1.64±0.13% vs. 0.51±0.04% for human cells transplanted to NOD/SCIDβ2microglobulin −/− recipients were recovered 18h after transplantation (n>20, p<0.005). A similar homing advantage was observed 3h and 48h after transplantation. To explore underlying mechanisms, several previously described differences between ssBM and MPB were rigorously addressed and correlated with in vivo homing. Despite down-regulated CXCR4, chemokinesis and SDF-1-directed chemotaxis of MPB-HPC were increased compared to fresh ssBM-HPC. Because of their enhanced motility, blocking of Gi-protein- or CXCR4-signaling reduced MPB-HPC homing by more than 50% (1.19±0.08% vs. 0.45±0.03% for Pertussis toxin treated (@18h), 0.87±0.07% vs. 0.39±0.04% for AMD3100 treated (@ 3h); p<0.005), similarly to cytokine-incubated BM-, but in contrast to fresh ssBM-HPC. We observed virtual loss of expression of CD26, a feedback inhibitor of SDF-1/CXCR4 signaling, from MPB c-kit+ cells, and inhibition of CD26 activity increased BM homing only of ssBM-HPC (from 0.85±0.04% to 1.05±0.05%, p<0.05, as described), but not of MPB-HPC (1.17±0.09% vs. 1.13±0.05%). These novel findings suggest that CD26 down-regulation on MPB cells and their exposure to a “circulatory environment” may contribute to improved migration and homing. Taken together, the data indicate: a) MPB display better homing compared to ssBM, correlating with their modified migratory behavior. b) SDF/CXCR4- and other Gi-signals make dominant contributions to MPB-HPC homing, in contrast to fresh ssBM. c) The recently proposed augmentation of BM homing/engraftment with CD26 inhibitors is ineffective for MPB, due to virtual absence of CD26 expression from MPB-HPC.

Author notes

Corresponding author