Long-Term Polyclonal and Stable Gene Transfer into Rhesus Repopulating Hematopoietic Stem Cells Using a Simian Immunodeficiency Virus-Based Lentiviral Vector System.

Murine leukemia virus (MLV) vectors have been studied extensively in animal models and utilized for over a decade in clinical trials of gene therapy directed at hematopoietic stem and progenitor cells MLV have a number of limitations, including inefficient transduction of quiescent cells and difficulty in maintaining stable high-level expression. More recently concerns have arisen regarding their safety regarding activation of adjacent proto-oncogenes and resultant leukemogenesis. We have previously reported that lentiviral vectors derived from the simian immunodeficiency virus (SIV) were efficient at transducing rhesus CD34+ cells, resulting in high-level in vivo marking with transduced progeny cells up to one year post-transplantation.(Hanawa et al, 2004) A comparison of vector integration sites in these animals compared to animals receiving MLV-transduced cells revealed different patterns, showing that SIV integrants strongly favored entire transcription units and gene-dense regions of the genome, compared to MLV that favored regions surrounding transcription start sites.(Hematti et al, 2004). Animals receiving MLV-transduced cells had highly non-random engraftment with integrants in or near the the MDS1/EVI1 gene complex. To evaluate long-term safety implications of the SIV vector-mediated CD34+ cell gene transfer, we analyzed the insertional sites in granulocytes, T cell, and B cells from 3 rhesus macaques which were transplanted three years ago with transduced, autologous cytokine-mobilized peripheral blood CD34+ cells. All three animals continued to show significant marking and expression levels in T cells, B cells and granulocytes, with mean GFP + levels of 6.7% (range, 3.3–13.0%), 7.4% (4.2–13.4%) and 5.6% (3.1–10.5%), respectively. Vector insertion site analysis by linear amplification-mediated PCR method at three years continued to show highly polyclonal reconstitution. Subsequent cloning and sequencing data confirmed long-term polyclonality with vector-containing cells and there was no evidence for any worrisome common integration sites, with no integrants detected in the MDS1/EVI1 region, in contrast to results with the MLV vector. These results indicate that the SIV vector system can result in stable and efficient long-term expression in progeny of transduced CD34+ cells, without the worrisome integration profile previously reported in our model with MLV vectors.