Consider Changing the Horse for Your CAR-T?

Main Text

The traditional workhorse for genome-modifying human T cells for therapeutic purposes remains the retroviral vector. However, clinical grade retroviral vectors are costly to produce and can be generated at only a few facilities. The work in this issue of Molecular Theapy by Bishop et al.¹ reveals that non-viral piggyBac transposon-generated CD19-specific chimeric antigen receptor (CAR) T cells have excellent in vivo efficacy in the setting of patient derived CD19+ B-acute lymphoblastic leukemia (ALL) in an animal model. Although piggyBac transposon-generated CAR T cells had previously demonstrated in vitro efficacy,2, 3, 4 sustained activity in vivo had not been reported until now. As previously found for retrovirally-modified CAR T cells, the study also demonstrates that in vivo activity is dependent on CAR structure and not all CD19-CAR constructs mediate potent eradication of malignant cells. Non-viral genome modification methods such as piggyBac transposons have the potential to increase the number of hospitals capable of implementing therapies by decreasing both the vector cost and complexity of the procedure. In this way, this work further represents a democratizing technology that could improve patient access to CAR T cell therapies.

In vitro anti-tumor activity of CAR T cells does not always correlate with in vivo efficacy.⁵ The authors sought to better understand what might improve piggyBac-transposon generated CD19-specific CAR T cell function in vivo. They reasoned that removing the IgG1 Fc region would eliminate interaction with Fc γ receptor-expressing cells and subsequently incorporated different transmembrane and co-stimulatory domains.⁶ Using 4-1BB co-stimulation, as opposed to CD28, led to greater potency and persistence of CD19-directed CAR T cells and provided protection against B-ALL re-challenge even at diminishing doses out to 150 days post-infusion in their xenograft model. Like all pre-clinical CAR T studies in mice, this study involves a xenograft in immune deficient animals. Therefore, a human immune system and malignancy is not fully recapitulated. Nonetheless, the study by Bishop et al.¹ is an important step towards clinical testing and application.

A PubMed search reveals >4,500 articles with the words “chimeric antigen receptor” and adding the word “transposon” to that search reveals 50 articles. One could say that the field of transposon-based genome engineering of therapeutic T cells remains in its infancy. This is despite the emergence and safety of the sleeping beauty transposon system,⁷ which showed activity in human cells before piggyBac,⁸ in CD19-directed CAR T cell therapy.⁹

What has limited more widespread use of transposon systems, like piggyBac, in CAR T investigations? Given the relatively recent discovery of transposon systems that are highly active in human cells, clinical development may be restricted by the patent system. Studies evaluating anti-tumor efficacy or vector enhancements in vitro or in animals may not be limited, but the translation to humans may be limited by those holding intellectual property thereby preventing commercial development. The transfection systems employed and DNA itself can be toxic to human T cells thereby limiting the yield of the therapeutic cells. This may result in a requirement for longer culturing of modified cells ex vivo in order to increase cell numbers to levels need for infusion. Longer ex vivo culture may alter the phenotype of cells over time. Insertional mutagenesis remains a possible concern, although this concern is not mitigated by using integrating viral vectors. Different transposon systems have different integration profiles with piggyBac resembling retroviral vectors,¹⁰ although retroviral vectors remain the workhorse for CAR delivery to T cells due to lack of apparent genotoxicity in mature human T cells. Remobilization of transposons may also be a concern, although transposase expression should be self-limited. Remobilization should be less of a concern with piggyBac, which mediates foot-print free excision compared to sleeping beauty, which leaves behind a foot-print mutation.11, 12

Why are transposon systems such as piggyBac relevant to improving CAR T cell therapy? The piggyBac transposon system offers several advantages with regards to genome engineering human cells (reviewed in ¹³). Transposons provide an efficient non-viral technology for stable genome modification. Transposon systems enable multiplexed transgene integration via transfection of multiple plasmids.¹⁴ Therefore, one could stably modify T cells to express not only a CAR but also one or more other transgenes improving therapeutic potential. Transposons like piggyBac have much larger cargo capacity than viral vectors.¹⁵ Larger transgenes can be inserted or multiple transgenes can be delivered in one vector. The clinical grade vector production cost for plasmids harboring transposons is much less than that for viral vectors. Reduced cost and complexity of vector production alone could have a dramatic impact on the clinical development and widespread availability of CAR T cell therapy for patients. Importantly, mouse models poorly predict the clinical efficacy of CAR T cells. Therefore, the possibility to perform adaptive phase I trials to evaluate different CAR constructs combined with immune response modulators could greatly facilitate the identification of effective therapies. Such studies would be feasible with low cost transposon systems but prohibitively expensive with viral vectors.

Future studies will undoubtedly be directed at overcoming the limitations of using non-viral technologies for CAR T and other cell therapies. Improved transfection methods, use of minicircles,¹⁶ or use of nanoparticles could improve efficiency of transfection. Drugs could be targeted to limit the toxicity of transfected DNA, thereby improving therapeutic cell numbers. The activity of transposon systems will continue to be increased through further engineering of transposon and transposase vectors. The less restricted cargo capacity of transposon vectors could allow further refinement of CAR structure including linker and co-stimulatory domains to further improve CAR T cell efficacy against malignant cells. Long-term safety and genotoxicity studies will need to be performed after using transposons to genome modify human T cells.

Non-viral transposon technologies offer an alternative gene delivery system to viral vectors for CAR T cell production. The authors show that piggyBac remains a promising non-viral system for CAR T cell based therapy. Further CAR optimization may improve efficacy even more in using this approach to attack various malignancies. Ultimately, this study lays the foundation for testing piggyBac transposon-generated CD19-specific CARs in human clinical trials.