Mutant Mgmt Lentivirus Co-Transduction With A Marker Lentivirus Efficiently Enriches For Dual-Vector Expressing Cells In Vivo

Dual gene vectors are limited by insert size constraints, reduced titer, and transcriptional interference. Moreover, high multiplicities of infection (MOI) are needed to compensate for limited expression. We evaluated an alternative strategy; co-transduction with separate lentiviral (LV) vectors. We assessed whether two genes in separate LV vectors efficiently transduce cells to produce dual-positive populations. One LV expressed a mutant form of the DNA repair protein, O6-alkylguanine DNA alkyltransferase (AGT-P140K), and the other LV expressed GFP (both constructs were derived from a 2nd generation SIN vector obtained from D. Kohn). Dual-positive cells were enriched by AGT-P140K mediated resistance to O6-benzylguanine (BG) and BCNU treatment. The two LV vectors were combined at varying ratios and used to transduce K562 cells with the cumulative multiplicity of infection held constant (MOI = 0.5). The level of expression achieved for each gene was proportional to the amount of the specific virus added and the fraction of dual-positive cells was equal to the product of the individual expression percentages (e.g., 9% AGT and 46% GFP yielded 6% dual positive). Equal proportions of the two vectors resulted in the highest level of co-transduction. After BG and BCNU treatment, AGT only and dual positive cells are enriched with equal efficiencies, while GFP only and untransduced cells are lost. Enrichment of co-transduced cells was higher in cultures that were initially transduced with a low AGT:GFP MOI ratio (8.8 fold for 0.05:0.45 vs. 2.5 fold for 0.25:0.25). We next evaluated whether the selective co-transduction strategy could be used to enrich dual positive murine bone marrow progenitors in vivo. 5-FU treated bone marrow cells were co-transduced with LV AGT-P140K and LV GFP (MOI of 10 each) and resulted in 31% AGT and 29% GFP expression ex vivo. These cultures were then transplanted into lethally irradiated recipients. A range of peripheral blood (PB) expression levels (12–33% AGT and 9-31% GFP) was detected and decreased over time in the absence of treatment. Following selection, expression levels increased up to 4 fold for AGT (17% to 75%) and up to 5 fold for GFP (17% to 81%). We then tested whether co-transduction would allow stem cell selection to be coupled to lineage specific GFP expression. A second generation vector (pRRL-GATA-GFP) containing the GATA-1 upstream enhancer (HS2) in place of U3 was previously shown to restrict expression to mature erythroblasts (F. Lotti et al, J Virol. 2002). Donor marrow was co-transduced with LV AGT-P140K and pRRL-GATA-GFP and transplanted into lethally irradiated recipients. After drug treatment, erythroid specific (TER119+) GFP levels were enriched in PB up to 16 fold (3–50%). These data demonstrate that co-transduction with an AGT-P140K and a marker LV can enrich dual expressing cells at limiting MOIs. Co-transduction simplifies the evaluation of in vivo selection by avoiding the constraints of dual gene vector strategies. This method is particularly suited to vector constructs containing complex regulatory sequences or therapeutic applications requiring tissue specific gene expression.