Mobilization as a Preparative Regimen for Hematopoietic Stem Cell (HSC) Transplantation.

Cytokines release HSC from marrow and thus facilitate their collection from blood. We performed studies in parabiotic mice to determine if microenvironmental niches are vacated when HSC are mobilized. In our initial experiments, ROSA26 (CD45.2+) and Pep3b (CD45.1+) (both C57BL6) mice were joined in parabiosis for 3, 6, 8 or 12 wks. Although their circulations were shared (approximately 50% of granulocytes in the blood of each parabiont had partner phenotype in all studies), few HSC engrafted in partner marrow. Specifically, 1.0–2.5 % of marrow HSC had a partner phenotype, as determined by the transplantation of marrow cells into irradiated secondary recipients (Blood 102,1249,2003). Also, few marrow granulocytes and CFU-GM had a partner phenotype, implying that marrow functions as an intact compartment in which resident HSC give rise to progenitors then mature cells. In contrast, in the spleen, 1.5–3.6% of HSC, yet 38–55% of granulocytes and CFU-GM had a partner phenotype. We then treated each parabiont with one cycle of hG-CSF (25 ug/kg) and hSCF (200 ug/kg) sq qd x 4d on days 17–20 of parabiosis and examined the marrow at 6 weeks (day 42). 10.1±6.2(SD)% of HSC had partner phenotype (p=0.02). When 3 cycles of cytokines were administered beginning days 17, 24 and 31, 13.9±9.0% of marrow HSC had partner phenotype at 6 wks of parabiosis (p=0.01). Experiments were then repeated with AMD3100 (a SDF-1/CXCR4 axis antagonist, gift of AnorMed, Langley, BC, CA; 5mg/kg/mouse sq on day 20; n=3 pairs (6 parabionts)) to demonstrate that HSC mobilization, and not replication, was responsible for these results. At 6 wks of parabiosis, 5.9±2.9% of marrow HSC, 5.7±2.3% of marrow CFU-GM and 5.6±2.8% of marrow granulocytes had partner phenotype (p=0.02), while 7.2±2.7%, 44.0±3.2% and 40.0±6.1% of splenic HSC, CFU-GM and granulocytes had partner phenotype, respectively. These data imply that HSC exited marrow, transited blood, engrafted in open niches in partner marrow, and contributed (normally) to hematopoiesis. Similar percentages of HSC also engrafted in spleen. However, splenic hematopoiesis, as at baseline, derived from CFU-GM, not HSC, engraftment. We next tested a corollary of these findings. If niches are vacated after AMD3100 administration, transplanted HSC might preferentially engraft. Pep3b mice were treated with AMD3100 (5 mg/g sq) and 40 x 10⁶ donor marrow cells (from ROSA26 mice) were transplanted (via tail vein infusion) 6 h later, 2 h later, or both 2 and 6 h later (n=3 mice per condition). Control animals received donor cells but no AMD3100. Donor cell engraftment was assayed at 3m (6 h later group) and 2m (other groups, 3 m data are pending) and was significantly higher in the experimental animals than control. Engraftment was 3.0±1.6, 6.5±4.9, 6.5±3.6 (SD)%, respectively, in the AMD3100-treated mice, and 0.6±0.9, 0.6±0.8 and 2.0±1.9% in the concurrent control studies (all p values = 0.02). Confirmatory experiments in BalbC mice (where higher engraftment rates are anticipated in control studies (Blood 98:1246,2001)) are underway. Our data argue that the number of available niches determine the number of HSC that engraft. As importantly, mobilization with AMD3100 could provide a non-toxic preparative approach to HSC transplantation for genetic (and other nonmalignant) disorders.