73. Enhanced Engineering of Chimeric Antigen Receptor (CAR)-Modified T Cells Using Non-Viral Sleeping Beauty Transposition from Minicircle Vectors

Abstract

Adoptive therapy with T cells that were modified with gamma-retroviral and lentiviral (LV) vectors to express a CD19-specific chimeric antigen receptor (CAR) has shown remarkable efficacy in pilot clinical trials. However, there are persistent concerns with the use of viral vectors with regard to safety, as well as the cost and scale of vector production required for making CAR T-cell therapy available to patients on a global level. In this study, we have refined non-viral Sleeping Beauty (SB) transposition to provide an effective and broadly applicable gene-transfer strategy with superior safety profile. We demonstrate that SB transposition of CAR transgenes from minicircle (MC) DNA vectors enables dramatically improved transposition rates compared to conventional plasmids. MC-derived CD19-CAR transposons display a highly favorable integration profile and confer stable CAR expression that permits potent anti-tumor functions and rapid manufacture of CAR T-cell products.We prepared MC transposon donor vectors encoding a CD19-CAR or eGFP from corresponding parental pT2 plasmids through site specific recombination, and a MC encoding SB100X transposase. Each MC contained only the promotor and gene of interest, and was devoid of antibiotic resistance gene and bacterial origin of replication. We then performed transfections into CD8+ and CD4+ T cells of healthy donors (n=7) and analyzed transposition rate and stability that could be accomplished when CD19-CAR transposon and SB100X transposase were encoded by MCs vs. equimolar amounts of corresponding conventional plasmids. In each donor, we achieved a significantly higher transposition rate with MCs (mean 49.8% on day 14 post transfection) compared to plasmids (mean 12.8%; 4.4-fold difference; p<0,001). Expression of the CD19-CAR was stable over multiple rounds of expansion and several weeks in culture. Importantly, MC transfection was not only more effective, but also substantially less toxic compared to plasmids and on average, a 6-fold higher yield of CAR T cells could be obtained within 14 days of culture without the need for feeder cell expansion. In functional experiments, CD19-CAR T cells modified with our enhanced MC-based SB transposition strategy were equally effective as LV transduced CD19-CAR T cells, and exerted high levels of specific cytolytic activity, cytokine production including IFN-g and IL-2, and productive proliferation after stimulation with CD19+ lymphoma. Moreover, a single administration of SB-modified CD19-CAR T cells leads to complete eradication of systemic lymphoma in a murine xenograft model (NSG/Raji-ffLuc), and was at least equivalently potent as LV transduced CD19-CAR T cells prepared from the same donor. To address safety, we determined the gene copy number (n=5, range 3-8 in CD8+ T cells) and performed a comprehensive genomic insertion site analysis. The data show a close-to-random integration profile of MC-derived CD19-CAR transposons, without preference for highly expressed or cancer related genes. Intriguingly, a significantly higher proportion of SB integrations had occurred in genomic safe harbors compared to LV integrations (7-fold difference; p<0,001), close to the perfect score expected for random integration. In conclusion, we demonstrate the potential to manufacture CAR T cells using virus-free SB-mediated transposition from MC DNA vectors. The superior safety profile, high level stable transposition rate and ease-of-handling MC vectors position our novel approach to become a preferred gene-transfer strategy in advanced cellular and gene-therapy.