Integration-specific In Vitro Evaluation of Lentivirally Transduced Rhesus CD34+ Cells Correlates With In Vivo Vector Copy Number

Hematopoietic stem cell (HSC) gene therapy using integrating vectors has a potential leukemogenic risk due to insertional mutagenesis. To reduce this risk, a limitation of <= 2 average vector copy number (VCN) per cell is generally accepted. We developed an assay for VCN among transduced CD34(+) cells that reliably predicts in vivo VCN in 16 rhesus recipients of CD34(+) cells transduced with a green fluorescent protein (GFP) (or yellow fluorescent protein (YFP))-encoding lentiviral vector. Using GFP (or YFP)-specific probe/primers by real-time PCR, VCN among transduced CD34(+) cells had no correlation with VCN among granulocytes or lymphocytes in vivo assayed 6 months post-transplantation. This was a likely result of residual plasmids present in the vector preparation. We then designed self-inactivating long terminal repeat (SIN-LTR)-specific probe/primers, which detect only integrated provirus. Evaluation with SIN-LTR probe/primers resulted in a positive correlation of VCN among transduced CD34(+) cells with granulocytes and lymphocytes in vivo. The transduced CD34(+) cells had higher VCN (25.1 +/- 5.6) as compared with granulocytes (2.8 +/- 1) and lymphocytes (2.4 +/- 0.7). In summary, an integrated provirus-specific real-time PCR system demonstrated nine-to tenfold higher VCN in transduced CD34+ cells in vitro, as compared with VCN in vivo. Therefore, the restriction of <= 2 VCN before infusion might unnecessarily limit gene transfer efficacy.