Transition from serum-supplemented monolayer to serum-free suspension lentiviral vector production for generation of chimeric antigen receptor T cells

Background aims Lentiviral vectors (LVs) have been used extensively in gene therapy protocols because of their high biosafety profile and capacity to stably express a gene of interest. Production of these vectors for the generation of chimeric antigen receptor (CAR) T cells in academic and research centers is achieved using serum-supplemented static monolayer cultures. Although efficient for pre-clinical studies, this method has a number of limitations. The main hurdles are related to its incompatibility with robust and controlled large-scale production. For this reason, cell suspension culture in bioreactors is desirable. Here the authors report the transition of LV particle production from serum-supplemented monolayer to serum-free suspension culture with the objective of generating CAR T cells. Methods A self-inactivating LV anti-CD19 CAR was produced by transient transfection using polyethylenimine (PEI) in human embryonic kidney 293 T cells previously adapted to serum-free suspension culture. Results LV production of 8 × 106 transducing units (TUs)/mL was obtained in serum-supplemented monolayer culture. LV production in the serum-free suspension conditions was significantly decreased compared with monolayer production. Therefore, optimization of the transfection protocol was performed using design of experiments. The results indicated that the best condition involved the use of 1 μg of DNA/106 cells, 1 × 106 cells/mL and PEI:DNA ratio of 2.5:1. This condition used less DNA and PEI compared with the standard, thereby reducing production costs. This protocol was further improved with the addition of 5 mM of sodium butyrate and resulted in an increase in production, with an average of 1.5 × 105 TUs/mL. LV particle functionality was also assessed, and the results indicated that in both conditions the LV was capable of inducing CAR expression and anti-tumor response in T cells, which in turn were able to identify and kill CD19+ cells in vitro. Conclusions This study demonstrates that the transition of LV production from small-scale monolayer culture to scalable and controllable bioreactors can be quite challenging and requires extensive work to obtain satisfactory production.