NKG2D-CAR memory T cells target pediatric T-cell acute lymphoblastic leukemia in vitro and in vivo but fail to eliminate leukemia initiating cells

Refractory/relapsed pediatric acute leukemia are still clinically challenging and new therapeutic strategies are needed. Interactions between Natural Killer Group 2D (NKG2D) receptor, expressed in cytotoxic immune cells, and its ligands (NKG2DL), that are upregulated in leukemic blasts, are important for anti-leukemia immunosurveillance. Nevertheless, leukemia cells may develop immunoescape strategies as NKG2DL shedding and/or downregulation. In this report, we analyzed the anti-leukemia activity of NKG2D chimeric antigen receptor (CAR) redirected memory (CD45RA-) T cells in vitro and in a murine model of T-cell acute lymphoblastic leukemia (T-ALL). We also explored how soluble NKG2DL affected NKG2D-CAR T cells´ cytotoxicity and the impact of NKG2D-CAR T cells on Jurkat cells. In vitro, we found NKG2D-CAR T cells targeted leukemia cells and showed resistance to the immunosuppressive effects exerted by soluble NKG2DL. In vivo, NKG2D-CAR T cells controlled T cell leukemia burden and increased survival of the treated mice but failed to cure the animals. After CAR T cell treatment, Jurkat cells upregulated genes related to proliferation, survival and stemness, and in vivo, they exhibited functional properties of leukemia initiating cells. The data here presented suggest, that, in combination with other therapeutic approaches, NKG2D-CAR T cells could be a novel treatment for pediatric T-ALL. To explore how CAR T cells affected gene expression, Jurkat GFP-luc cells were co-cultured with NKG2D-CAR T cells for 16h, sorted and used to perform transcriptome analysis by RNAseq. Sorted Jurkat-GFP cells that remained after overnight exposition to NKG2D-CAR T cells were washed in PBS and pelleted in RLT lysis buffer supplemented with 1:1000 β-mercaptoethanol. RNA isolation and sequencing was carried out as our group previously reported (Fernández A et al. Cancers, 2021). Reads from RNA-seq were analyzed to quantify genes through the RSEM (Li B et al. Bioinformatics, 2011) with hg19 as reference for annotation. The differential expression was carried out with edgeR (Robinson MD et al. Bioinformatics, 2009) and the statistical cut-off point was set as FDR < 0.05 and logFC >2. Genes were filtered based on a minimum expression of one Counts Per Million (CPM) in more than one sample. Normalization was performed by the TMM method (trimmed mean of M-values)(Robinson MD et al. Genome Biol, 2010). Functional enrichment of differentially expressed genes was carried out in PFAM (Mistry J. et al, Nucleic Acids Res, 2021), GOTERM_BP_DIRECT (Ashburner M et al. Nat. Genet, 2000) and KEGG REACTOME (Kanehisa M, et al. Protein Sci, 2022)