106. A Sleeping Beauty Transposase System in the Context of HD-Ad5/35++ Vectors Achieves Stable In Vivo Transduction of Hematopoietic Stem Cells in Mouse Models

Hematopoietic stem cells (HSCs) are important targets for the gene therapy of infectious diseases, genetic disorders, and cancer. Our aim is to transduce HSCs in vivo after mobilization from the bone marrow and intravenous injection of a gene transfer vector. We employed helper-dependent adenovirus vector containing Ad35 fiber mutants with enhanced affinity to CD46 (HD-Ad5/35++) for gene transfer into HSCs. HD-Ad5/35++ vectors efficiently transduced human HSCs and human CD46-transgenic mouse HSCs in vitro. To trigger chromosomal transgene integration we utilized the hyperactive Sleeping Beauty transposase system (SB100x). The system consists of two vectors; the transposon vector carrying a GFP gene and the SB100x vector that mediates integration of the GFP cassette into a TA dinucleotide of the genomic DNA. As a limitation, this system requires the co-infection of the same target cell with two vectors to achieve stable transduction. This is exemplified in a study with two Ad5/35++ reporter vectors, expressing either GFP or mCherry. Human CD46tg mice were mobilized with GSCF/AMD3100 and intravenously injected with either vector alone or the vector combination. Reporter gene expression in bone marrow localized HSCs was analyzed 3 days later. While 2% of lineage depleted cells expressed either GFP or mCherry only 0.25% expressed both. In spite of these limitations, we tested whether the co-injection of a HD-Ad5/35++ -GFP transposon vector and a HD-Ad5/35++ SB100x vector, supplying a Sleeping Beauty transposase in trans into mobilized mice would lead to stable HSC transduction. Twelve weeks post-injection, we detected stable GFP expression in bone marrow HSCs in the range of 1 to 2%. GFP integration through SB100X was confirmed by genome-wide sequencing of integration sites. Our data indicate that HD-Ad5/35++ vectors preferentially target primitive HSCs and that this increases the chance of co-infection with both vectors. This is supported by a comparison of GFP marking in HSCs and more differentiated marrow and blood cells four weeks after in vivo transduction. While in the HSC-containing LSK fraction of the bone marrow the marking rate was on average 7.7%, marking rates for total mononuclear cells of the bone marrow and peripheral blood were only 2.5% and 0.3%, respectively. The preferential HSC targeting in hCD46tg and humanized mice can be explained by higher CD46 levels on primitive HSCs compared to more differentiated cells in the bone marrow and the dependence of Ad5/35 infection efficiency (at low MOIs) on CD46 density on the cell surface.