Long-Term, Clonal Tracking Comparing Autologous Transplantation of G-CSF/SCF-Primed Bone Marrow CD34+ Cells with G-CSF/SCF-Mobilized Peripheral Blood CD34+ Cells in Rhesus Macaques

Introduction:

Although autologous transplantation of peripheral blood stem cells (PBSC) - mobilized with a combination of granulocyte colony-stimulating factor (G-CSF) and stem cell factor (SCF) - has been well characterized, the efficacy of G-CSF/SCF-primed bone marrow stem cell (BMSC) transplantation, however, remains unclear and controversial. In our previous lentiviral vector-mediated PBSC and BMSC transplantation study, we reported efficient and long-term hematopoietic reconstitution by PBSC but not by BMSC - the later being associated with the gradual decline of vector markings in two BMSC-transplanted animals, with a loss of marking occurring in most lineages by 26 or 32 weeks after transplant. Follow-up analysis indicated that low-level yet consistent repopulation by BMSC continued in these animals for a longer period. Here we have compared peripheral blood (PB) markings and vector integration sites (VIS) in PBSC- and BMSC-transplanted animals for up to 12 years and 6 years, respectively.

Methods:

Young adult rhesus macaques were treated with G-CSF (10 mg/kg of body weight/day) and SCF (200 mg/kg/day) four days before the cell harvest for transplant. Mobilized PB leukapheresis cell products from five rhesus macaques (95E132, 2RC003, RQ5427, RQ3570, and 96E035) were harvested using a CS3000 Plus blood cell separator. Bone marrow (BM) cells from two animals (95E131 and 96E041) were surgically harvested from their femurs and iliac crests under anesthesia. After harvest, PBSC and BMSC were isolated by Ficoll-Hypaque density centrifugation followed by immunoselection of CD34+ cells, and transduced with HIV-based self-inactivating lentiviral vectors expressing EGFP. Vector-marked cells were then autologously transplanted into the host after total body irradiation (10 Gy). No further priming treatment was provided after transplant. PB from the 5 PBSC- and 2 BMSC-transplanted animals were serially collected over time and cryo-preserved for PCR, flow cytometry, and VIS analyses.

Results:

Both PBSC- and BMSC-transplanted animals showed long-term repopulation for lymphocytes, monocytes, granulocytes, platelets, and red blood cells. PBSC animals showed an average EGFP marking that ranged from 0.32 % to 10.24 %. From these animals, a total of 141 to 4,858 VIS were recovered. We found that the total number of VIS in each animal was proportional to the average EGFP marking in the same animals, and that both of these in turn were linearly correlated with the number of EGFP+ CD34+ cells initially transplanted (1.4 x10⁶ - 28.8 x10⁶ per animal). BMSC-transplanted animals did not show any such patterns. In two BMSC animals, the average EGFP marking levels remained at 0.05 % and 0.10 % until the end point (5 and 6 years) despite the fact that a comparatively large number of EGFP+ CD34+ cells had been transplanted (5.2 x10⁶ and 17.7x10⁶)and a large number of VIS recovered (793 and 680 VIS) in these animals. Temporal VIS analysis of PBSC animals showed that different groups of a large number of PBSC clones repopulated sequentially and reached a point of maximum repopulation at different time points, with some gradually declining after this. BMSC animals also showed a wave-like sequential repopulation similar to the patterns seen in PBSC animals. Unlike PBSC, however, nearly all BMSC clones were detected at a low frequency and at a single time point, except a few larger ones that were detected at multiple time points in a rising and falling pattern. There was no notable difference between the genomic features of VIS in PBSC- and BMSC-repopulating cells.

Conclusions:

Our data suggest that both the BMSC and PBSC consist of highly heterogeneous stem/progenitor cells that can provide long-term polyclonal repopulation through wave-like, sequential repopulation. Unlike PBSC, however, BMSC transplant was inefficient in PB repopulation resulting in only barely detectable markings in PB. The BMSC clonal profiles reflected the clonal patterns seen in PBSC animals, aside from BMSC animals having primarily low-frequency clones. We have previously shown significant differences in immunophenotype and cell cycle status between PBSC and BMSC, where BMSC were significantly lower in Thy-1 expression and had a higher percentage of cells in the S+G2/M phase of the cell cycle than PBSC. These differences may account for the inefficient differentiation and proliferation capabilities of BMSC compared to PBSC shown in this study.