Abstract
Even within cell populations enriched for hematopoietic stem cells, there exists phenotypic and functional heterogeneity, with subsets of progenitors differentially contributing to the total character of the transplant. Here, we have used a dynamic in vivo imaging technology to track the differential homing kinetics of such subsets, within a well defined stem cell pool, and identified patterns of short term homing unique to CD31 versus CD34 cells.
Human umbilical cord blood derived (hUCB) hematopoietic progenitor cells (HPC) were isolated on high expression of aldehyde dehydrogenase (ALDH) to yield a population of cells we have previously reported to be enriched for stem cell activity by phenotypic and functional parameters. Cells were plated 12 hours on fibronectin in X-Vivo15 supplemented with SCF, flt-3, and TPO. During incubation, cells were labeled with Bang’s Laboratories 0.9um SPIO nanoparticles containing the Dragon Green fluorescent dye at a ratio of approximately 10:1 beads to cells. Positively loaded cells were immunomagnetically selected and transplanted into sublethally irradiated NOD/SCID/MPS7 mice. The MPS7 mouse model is highly permissive of human cell engraftment, and allows easy identification of donor cells by virtue of a genetic defect in beta-glucuronidase production. Post-transplant, animals underwent MRI every 30 minutes for 12 hours to develop a dynamic, real-time survey of donor migration between organs, and then sacrificed at optimal times for analysis by flow cytometry.
The utility of this approach was validated by quantitative PCR for human beta-globin versus murine rapsyn to correlate loss of MR signal intensity to human cell engraftment. At 12 hours, MR signal intensity had decreased by 29% relative to time 0, with internal standardization against a water control. In contrast, repetitive imaging of a non-transplanted animal resulted in deviation of only +/− 4% from the mean. This loss of MR signal correlated to a spleen engraftment of as few as 10,000 cells (homing efficiency of ~2.5%) in some cohorts. This demonstrates the efficacy and resolution of this paradigm for real time analysis and manipulation of homing in vivo.
We have noted a preferential affinity of CD31+ ALDH high HPC that highly co-express markers for CD117 and CD133 to the splenic vasculature as early as 1 hour post-transplant and peaking at 12 hours post-transplant. Alternatively, we observed initial marrow seeding to contain almost exclusively CD34+ cells with little co-expression of CD31 or even CD133. Additionally, we found that the spleen-homed fractions of cells also express classical markers of activation such as CD71 and CD38, whereas the marrow homed fraction does not. From a functional standpoint, we believe these two subsets comprise two facets of contribution to the total graft, providing short-term radioprotection and cytokine delivery from the spleen-homed compartment, and long term durable engraftment and self-renewal from the marrow-homed compartment.
Author notes
Corresponding author