Abstract
We previously demonstrated that engraftment of murine whole bone marrow (WBM) transduced with an oncoretroviral vector using an optimized 5-fluorouracil (5-FU)-based transduction protocol is reduced ~3-fold compared to fresh WBM upon transplantation into sublethally irradiated hosts, although competitive repopulating ability in ablated hosts is not decreased (Goebel et al., Exp. Hematol. 30:1324, 2002). We therefore sought to determine whether marrow cells transduced using a clinically relevant, non-5-FU-containing protocol would engraft more efficiently. Li et al. (Exp. Hematol. 31:1206, 2003) showed that lineage-depleted (lin−) marrow cells from donor mice not treated with 5-FU were effectively transduced, and repopulated myeloablated hosts. We hypothesized that ex vivo culture for gene transfer in the absence of 5-FU would lead to improved donor marrow engraftment in submyeloablated hosts. Lin− cells, isolated from B6.SJL (Boy J; CD45.1+) WBM using a Miltenyi kit and VarioMACS apparatus, were prestimulated in StemSpan serum-free medium with SCF and IL-6 for 48 hours, followed by overnight transduction on RetroNectin-coated plates preloaded with ecotropic SF1-EGFP retroviral supernatant. Cell recovery from the MACS column (1.9 ± 1.4%), bulk transduction efficiency (77.3 ± 12%), and lin− cell purity (58 ± 17%; all from 7–13 experiments) was similar to that previously described. Transplantation of 106 lin− transduced cells into 300 cGy-conditioned congenic C57Bl6/J (B6; CD45.2+) hosts produced only 1.4 ± 0.5% donor chimerism 4–6 months post-transplant, significantly lower than that observed using 106 fresh lin− cells (29 + 18.8%; N = 8–9 hosts each from 2–3 experiments). The percentage of EGFP+ cells in the donor population, nevertheless, was 55.6 ± 18%, indicating that stem cells were marked but engrafted poorly. The repopulating ability of transduced lin− marrow was reduced ~10-fold compared to fresh lin− cells as determined in competitive repopulation assays in ablated hosts. Together, these data suggest that lin cells cultured ex vivo for gene transfer acquired an engraftment defect despite the absence of 5-FU. Increasing the conditioning radiation dose to 550 cGy, a dose used in prior canine and non-human primate gene transfer studies, markedly improved donor chimerism following transplantation of 106 fresh lin− cells (90 ± 1.3% at 4 months, N = 5) or 106 transduced lin− cells (38.5 ± 14% at 2 months, N = 10), suggesting that greater reduction in host stem cell function may be needed for engraftment of cells cultured ex vivo for gene transfer. Ongoing studies to investigate the mechanism responsible for this engraftment defect indicate that expression of adhesion molecules important for homing and engraftment (CD29, 44, 49d, 49e, 62L), CXCR4 expression, and the percentage of cells actively cycling are not significantly altered by the transduction process, although functional studies are underway. The percentage of Sca-1+lin−c-kit+ (SLK) cells in the transduced cell pool is similar to that of freshly isolated lin− cells; thus, transplantation of lin− cells cultured ex vivo for gene transfer results in significantly lower donor chimerism than fresh lin− cells despite the grafts containing similar numbers of SLK cells. Secondary transplants and limiting dilution studies to determine stem cell self-renewal and engraftment capacity before and after ex vivo culture for gene transfer are in progress.
Stem cell integrins: Implications for ex-vivo culture and cellular therapies
