Top of pageAbstract Significant research efforts are directed toward achieving non-viral, DNA-mediated transfer and stable expression of genes in somatic cells for the treatment of diseases. Plasmid DNA has been successfully used to deliver genes into cells, but its usefulness is restricted by the inability to provide sustained expression. One alternative is to include components of an integrating system in the vector design. The Sleeping Beauty (SB) transposase and phiC31 integrase both bring about insertion of DNA sequences into the genomes of human cells, suggesting an alternative to viral transduction. Because each approach requires its own unique genetic elements and catalytic enzyme to mediate gene insertion into cellular chromosomes, we chose to directly compare the effectiveness of integration and stable expression using a two-component system that is internally controlled for the integrating sequence. The integrating vector (pKT2/NAG) consists of (i) a phiC31 integrase recognition site (attB), (ii) eF1-alpha regulated GFP expression unit, and (iii) neomycin resistance gene regulated by the PGK promoter, all flanked by (iv) T2 transposase binding sites (IR/DRs) separated by (v) a colE1 bacterial origin and kanR gene. Overexpression inhibition of the SB transposase suggests a potential limitation with respect to integration efficiency that can be achieved by this vector. To address this issue, we first studied the integrating efficiency of pKT2/NAG by evaluating the effect of the dose of the co-delivered enzyme-encoding plasmid on integration and stable gene expression as determined by G418-resistant colony formation and GFP fluorescence. In each case, pKT2/NAG (500 ng) was co-transfected with 150 ng, 500 ng, or 1500 ng of either CMV-regulated luciferase (pCMV-Luc), transposase (pCMV-SB), or integrase (pCMV-Int) expression plasmids into cultured human HT1080 cells. G418-resistant colony formation exceeded levels achieved by random integration (+ pCMV-Luc) at all tested concentrations of the SB transposase-encoding plasmid, with diminished efficiency at higher concentrations. The inverse relationship was observed for phiC31 integrase, where gene transfer improved as the concentration of pCMV-Int increased relative to pKT2/NAG. Drug-resistant clones have been isolated for studies into: (i) gene silencing by flow cytometric analysis of GFP fluorescence, (ii) insertional preference by cloning and sequencing vector::chromosome junctions, and (iii) insertion frequency by Southern analysis. We have also co-delivered both components into rat multipotent adult progenitor cells and found both systems to effectively mediate G418-resistant colony formation above background. Studies underway in mouse ES cells and in human CD34+ cells using this bifunctional approach will provide a broad assessment of the relative usefulness of these two major non-viral integrating systems for the purpose of ex vivo gene therapy applications in a range of human diseases.