Background Chimeric antigen receptor (CAR) T-cells have revolutionized the treatment of hematological malignancies. However, the current autologous approach to CAR T-cell production suffers several drawbacks (difficulty to obtain high quality starting material from patients, limited manufacturing capacity, complex logistics, long vein-to-vein time, high costs), to the point that to date only approximately 10% of eligible patients can receive the treatment. Allogeneic ‘off-the-shelf’ CAR T-cells could help circumvent these issues, but their successful clinical application is hampered by the difficulty to manage graft-versus-host disease (GvHD) and host-versus-graft (HvG). A variety of genetic engineering strategies have been proposed to address these issues, all burdened by the complexity of multiple, potentially risky genome edits. Here we describe a non-gene edited one-step approach for the effective generation of allogeneic CAR T-cells. To this aim, we silenced CD3ζ to target T-cell receptor (TCR) expression, and we downregulated β2M and CIITA to target HLA class I and HLA class II expression, respectively. We aimed at preventing apoptosis induced by CD95L, often expressed by cancer cells, by knocking down CD95.

Methods A microRNA (miRNA)-based short hairpin RNA (shRNA) platform was developed, to allow for the tunable modulation of multiple target genes simultaneously. The platform was equipped with shRNA-derived guide sequences (shGuides) targeting CD3ζ, β2M, CIITA and CD95 simultaneously, and combined with an anti-CD19 CAR. Each shRNA was screened and assessed functionally in vitro.

Results The miRNA-derived shRNA platform targeted against CD3ζ, β2M, CIITA and CD95 in the CAR construct led to a substantial downregulation of TCR, HLA class I, HLA class II and CD95. This, in turn, granted the engineered CAR T-cells protection in in vitro assays mimicking GvHD, HvG and CD95L-induced autophagy, ultimately leading to significantly improved CAR T-cell survival. Notably, thanks to the tunability of gene expression allowed by our platform, we were able to achieve functionally relevant downregulation of HLA class I to avoid recognition by donor T-cells, ensuring at the same time sufficient expression to let CAR T-cells escape NK recognition, without the need of further editing (e.g., HLA E overexpression).

Conclusions These data validate our technology for the effective introduction of multiple functionally relevant edits in CAR T-cells. Moreover, we show in vitro proof-of-concept that our approach can improve allogeneic CAR T-cell viability by allowing evasion from GvHD, HvG and CD95L-induced autophagy. Our technology stands out as an easy, safe, and effective strategy for the generation of allogeneic CAR T-cells.