Cloning and Functional Analysis of the Rhesus Macaque ABCG2 Gene: Forced Expression Confers an SP Phenotype among Hematopoietic Stem Cell Progeny in Vivo.

Hematopoietic cells can be highly enriched for repopulating ability based upon efflux of the fluorescent Hoechst 33342 dye by sorting for side population (SP) cells, a phenotype attributed to expression of ABCG2, a member of the ABC transporter superfamily. Intriguingly, murine studies suggest that forced ABCG2 expression prevents hematopoietic differentiation. We sought to determine the effects of forced expression of the ABCG2 gene in hematopoietic stem cells in the nonhuman primate model, a model with proven relevance to human hematopoiesis. We cloned the full-length rhesus ABCG2 (rh-ABCG2) cDNA using a series of primers spanning the entire sequence designed using the published human sequence. Sequence homology was greater than 96%. The rh-ABCG2 gene was then introduced into an MFGS based retroviral vector pseudotyped with the RD114 envelope. Mobilized human peripheral blood CD34-positive cells were transduced with either rh-ABCG2 or human GP91-phox vector with no other payload. All transductions were initiated with 4 x10e5 cells using X-VIVO10/1%HSA/4mM/L L-glutamine supplemented with 100ng/ml_FLT3L, 100ng/ml SCF, 100ng/ml TPO and polybrene (5 ug/ml). RD114 vector was concentrated by ultracentrifugation (83,000g 90minutes 4°C). Gene transfer rates to CFU of greater than 80% were achieved using both vectors with similar gene transfer rate estimated by flow cytometry. ABCG2-transduced human peripheral blood progenitor cells (PBPCs) acquired the SP phenotype, but showed significantly reduced growth compared to control (Day 8: cell counts 7.67+/− 2.54 vs. 17.83+/−6.64 x10e5 for ABCG2 and GP91-phox transduced cells, respectively p=0.0024, n=5). We then examined the engraftment of ABCG2-expressing stem and progenitor cells in the rhesus macaque autologous transplant model. GCSF/SCF mobilized PBPCs were collected from 2 animals and the CD34+ cells were divided and transduced with either vector and infused after lethal irradiation. In vivo marking levels post transplant measured in mononuclear cells and granulocytes from peripheral blood and bone marrow ranged initially from 0.5–4% by Realtime PCR, declined equally over time, and were similar between transduced fractions, with no discrepancy between bone marrow and peripheral blood marking. Furthermore, peripheral blood T cells, B cells and granulocytes expressed ABCG2 at levels predicted by vector copy number long term, and the differential of such cells within the SP gate matched that of the non-SP fraction demonstrating no block to differentiation in the large animal. In vitro studies showed selective protection against mitoxantrone among ABCG2-transduced rhesus PBPCs. Our results confirm the existence of rhesus-ABCG2, support its importance in conferring the SP phenotype, suggest no detrimental effect of its overexpression upon hematopoiesis, and imply a potential role for its overexpression as an in vivo selection strategy for gene therapy applications.