Highly Efficient Multiplexed Base Editing with Minimized Off-Targets for the Development of Universal CAR-T Cells to Treat Pediatric T-ALL

Autologous CAR-T therapies have demonstrated remarkable efficacy in treating some hematologic cancers. While these therapies can have substantial clinical benefit for patients, generating bespoke cell therapies creates manufacturing challenges, resulting in inconsistent products and delays in treatment that are often incompatible with effective clinical management of patients. Strategies to create universally-compatible CAR-T therapies, generated from single donors for the treatment of many patients, have been developed as a solution to these challenges, thereby reducing cost of goods, lot-to-lot variability and enabling timely treatment. Mitigating the risks of graft-versus-host-disease (GvHD) and host rejection of CAR-Ts are important components of any strategy to generate these universal therapies. Most first generation approaches utilize DNA double strand break (DSB)-inducing nucleases to ablate the expression of relevant genes in donor T cells to overcome these barriers. However, simultaneous induction of multiple DSBs results in a cell population containing genomic rearrangements, and can lead to significantly reduced cell proliferation. Approaches to develop CAR-T therapies for T cell malignancies, such as T-ALL, encounter additional challenges, including extensive fratricide caused by targeting T cell surface markers such as CD3 and CD7, which are present on both the malignant and the CAR-T cells. Resolving this issue requires additional gene editing, leading to, in the case of nuclease-based strategies, an increased number of DSBs, further genomic rearrangements, and decreased cell expansion.

Base editors (BEs) are a novel class of gene editing reagents that enable programmable, single-base changes in genomic DNA without creating DSBs. Work in the Qasim lab has demonstrated proof of concept for an alternative means of producing universal, fratricide-resistant CAR-T cells for treating T cell malignancies by using cytosine base editor (CBE) technology. Here, we demonstrate a previously-unpublished CBE that produces greatly diminished off-target effects while preserving on-target activity. Using multiplex base editing, we demonstrate simultaneous knockout of four genes (TRAC, CD7, CD52, and PDCD1) with between 80-95% efficiency, producing engineered CAR-T cells with greatly diminished risk of GvHD, graft cell rejection, fratricide, and exhaustion. We show that, in contrast to nuclease editing, concurrent modification of four genetic loci using our reduced off-target CBE produces highly efficient gene knockouts with no detectable genomic rearrangements and no observable change in cell expansion compared to control conditions.

T-ALL is a heterogeneous disease with variable expression of CD3 and CD7 across tumor cells in the same patient. To reduce the risk of antigen escape by tumor cells during the course of treatment with base edited CAR-Ts, we envisioned creating two independent CAR-T populations targeting CD3 (3CAR-Ts) or CD7 (7CAR-Ts). Alone or in combination, base edited 3CAR-Ts and 7CAR-Ts demonstrate robust cytokine release, potent in vitro cytotoxicity, and in vivo tumor control with antigen-positive tumor cells, and display minimal antigen-independent activity. Taken together, our approach addresses existing limitations in CAR-T cell manufacturing and demonstrates that simultaneous base editing using an improved specificity CBE at four target genes is a feasible strategy for generating universal, fratricide-resistant CAR-T cells for the potential treatment of T cell malignancies such as T-ALL. More generally, this program demonstrates the potential for base editing to create highly-engineered cell therapies featuring at least four simultaneous edits, which can confer a wide range of desirable therapeutic attributes.