Simultaneous dual CAR expression to prevent relapse in pre-B acute lymphoblastic leukemia

CAR-T cells are a new therapeutic tool to treat pre-B Acute Lymphoblastic Leukemia (ALL): patient's T cells are transduced with a lentivirus to express chimeric antigen receptors (CARs) against CD19 or CD22. However, persisting CAR-T cells generate sustained pressure on leukemic cells that drive mechanisms of persistence such as the loss of the targeted molecule. Indeed, 14% to 33% of relapses following CAR-CD19 therapy are CD19-negative and occur with different CAR-CD19 products. Patients with CD19-negative relapsed leukemia have very poor prognosis. Similarly, a mechanism of leukemic evasion based on diminished CD22 site density has been recently reported after anti-CD22 CAR therapy. To circumvent these limitations, we propose a therapy using CAR targeting more than one antigen. Our hypothesis is that engineering of T cells with a bi-specific CAR-CD19/CD22 construct could prevent the occurrence of immune-evasion and relapse of pre-B ALL. We have graciously obtained the plasmid coding for CAR-CD22 from RJ. Orentas (NIH). To create the CAR-CD19, we have cloned the ScFv sequence from the FMC63 anti-CD19 antibody. Using these ScFv regions, we have generated CAR constructions containing the CD3ζ along with CD28 (2^nd generation). The two CAR molecules (CD19/CD22) are constructed in a bicistronic fashion separated by a T2A self-cleaving peptide. This transgene was cloned in a second-generation lentiviral construct under the control of a SFFV promoter and packaged in VSVg pseudotyped particles. These viral particulates were used to transduce peripheral CD3⁺ T cells, which were stimulated using Dynabeads. To test the specific cytotoxic activity of T cells containing the double CD19-CD22 CAR in an experimental setting mimicking tumour immune evasion, we have generated, by CRISPR, CD19 and/or CD22 knock-out ALL cell lines. The effect of CAR-T cells was tested during a co-culture with the generated ALL knock-out cell lines. We show that T cells bearing the dual CARs kill wild-type ALL cell lines (CD19⁺CD22⁺) or single KO cells lines (either CD19⁺CD22⁻ or CD19^-CD22⁺) but not double KO CD19⁻CD22⁻, confirming that both CAR molecules are expressed at the cell surface and that our bi-specific CAR-T cells can kill clones mimicking cancer immune-evasion at a single target. This strategy to express two CARs could thus be harnessed as a way of limiting immune-escape following CAR therapy.