G-CSF (Filgrastim) has been the standard agent for mobilizing peripheral blood stem cells (PBSCs) for transplantation. An alternative agent, AMD3100 (Plerixafor), has been used to mobilize PBSCs for autologous transplantation and is being tested in allogeneic donors. Since these agents have different mechanisms of action, the stem cells they mobilize may differ. We used a rhesus macaque model to compare CD34+ cells mobilized with these agents. Three macaques were given G-CSF, G-CSF plus SCF, AMD3100, or AMD3100 plus G-CSF. PBSCs were collected by apheresis and CD34+ cells with isolated by immunoselection. For G-CSF mobilization 10 μg/kg/day SQ was given for5 days, for G-CSF plus SCF mobilization 10 μg/kg/day of G-CSF and 200 μg/kg/day SQ of SCF were given for 5 days. For AMD3100 mobilization 1 mg/kg dose was given and for AMD3100 plus G-CSF mobilization 5 days of G-CSF (10 μg/kg/day) and one dose ofAMD3100 (1 mg/kg) were given. A PBSC concentrate was collected 2 hours after the dose of AMD3100 or the last dose of G-CSF. The CD34+ cells were analyzed by global gene and micro RNA (miR) expression analysis. A mean(SD) of 2.0(0.5)×107, 3.2(2) ×107, 2(0.5) ×107, and 12(7) ×107 immunoselected CD34+ were collected for G-CSF alone, G-CSF+SCF, AMD3100, and AMD3100+G-CSF, respectively. Gene expression analysis was performed with a microarray with 17,000 cDNA probes and miR analysis with an array with 827 miRs. Unsupervised hierarchical clustering of the gene expression data separated the samples into 2 clusters. One cluster included all CD34+ cells mobilized with G-CSF and G-CSF+SCF and the other all CD34+ cells mobilized with AMD3100and AMD3100+G-CSF. We found that 1,097 genes were differently expressed among the CD34+ cells mobilized with the 4 protocols (F-test, p<0.005). Hierarchical clustering analysis of these differentially expressed genes separated the samples into three clusters; one with all G-CSF- and G-CSF+SCF-mobilized, one with all the AMD3100-mobilized, and another with all AMD3100+G-CSF-mobilized CD34+ cell samples. When CD34+cells mobilized with AMD3100 alone were compared to those mobilized with G-CSF alone, the AMD3100-mobilized cells were enriched for genes expressed by T cells, B cells and mononuclear phagocytes and red cells (CXCR4, FLT3, CD83, CD3e, CD79a, CD2, hemoglobin b and hemoglobin a2) and the G-CSF-mobilized cells were enriched for genes expressed by neutrophils (proteinase 3, neutrophil cytosolic factor 4, CD16, S100 calcium binding protein, and myeloperoxidase). Genes up-regulated in AMD3100+G-CSF-mobilizedCD34+ cells included many not up-regulated by either agent alone. Hierarchical clustering analysis of the 97 differentially expressed miR (F-test, p<0.05) separated the samples into three clusters; a AMD3100+G-CSF group with all AMD3100+G-CSF-mobilizedsamples, a G-CSF+SCF group with all 3 G-CSF+SCF-mobilized samples, oneAMD3100-mobilized sample and one G-CSF-mobilized samples and a mixed group will two AMD3100-mobilized and two G-CSF-mobilized samples. AMD3100-mobilized cells were enriched for miR-195, let-7c and let-7e and G-CSF-mobilized cells for miR-220, −187and −520. These results demonstrate that the composition of mobilized CD34+ cells is dependent on the mobilization protocol and that CD34+ cells mobilized by a combination of agents are not simply a mixture of cell mobilized by each agent separately. These results suggest that a simple comparison of the number of CD34+ cell mobilized among protocols may not accurately reflect the potency of the mobilized stem cell population.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author