HLA and TCR Knockout by Zinc Finger Nucleases: Toward “off-the-Shelf” Allogeneic T-Cell Therapy for CD19⁺ Malignancies.

Abstract 3766

Cell therapy by infusion of T cells can reconstitute immunity to combat pathogens and malignancies. However, the time required to manufacture T cells with the desired properties and in sufficient numbers ex vivo is often incompatible with the treatment window for patients. Furthermore, autologous T cells from patients with advanced disease may have compromised function and be tolerant to desired antigens. A potential solution would be an approach to infuse allogeneic T cells that avoids immune-mediated rejection caused by host T cells recognizing disparate major or minor histocompatibility antigens on the infused cells. To broaden the application of T cell therapy, we investigated whether HLA gene expression can be disrupted by designer zinc-finger nucleases (ZFNs). ZFNs comprise a zinc finger DNA binding domain designed to bind a specific DNA sequence fused to the cleavage domain of Fok I endonuclease. Since FokI dimerization is required to introduce a double strand break (DSB), we generated ZFN pairs that flank the intended DNA target sequences in the required spatial conformation. Cellular repair of the DSB by error-prone non-homologous end joining allows disruption of HLA gene expression. As an initial proof of concept experiment, transfection of ZFN pairs designed to target exon 3 of the HLA-A locus into the human kidney cell line HEK293 resulted in 10% genetic modification of the HLA-A loci. We generated clones of HEK293 cells that showed deletion or insertion mutations within the ZFN binding site of one or both HLA-A alleles leading to early termination of translation. These HLA-A^null HEK293 clones evaded HLA-A-restricted lysis by T cell clones, even after interferon-γ and TNF-α treatment was used to upregulate HLA expression. Since only transient expression of ZFNs is needed to disrupt a target gene, we tested the ability to disrupt HLA-A gene expression by electro-transfer of in vitro-transcribed ZFN mRNA into primary T cells. We show that a single administration of the mRNA encoding the ZFNs targeting HLA-A could render over 40% of primary T cells HLA-A negative. We enriched the HLA-A^null population by paramagnetic bead separation to obtain a pool of T cells >90% of which lack HLA-A expression. An attractive potential clinical application for HLA^null allogeneic T cells is to redirect their specificity independent of HLA via expression of a chimeric antigen receptor (CAR) targeting CD19. Thus, we eliminated HLA-A expression from CD19-specific CAR⁺ T cells and demonstrated that they (i) evade HLA-A-restricted lysis by T cell clones, and (ii) specifically lysed CD19⁺ tumor targets. Finally, to further improve this T cell product and eliminate potential deleterious immune mediated recognition by the endogenous T cell receptor (TCR) on allogeneic CAR⁺ T cells, we used ZFN pairs targeting the TCR α or the TCR β locus. Transient expression of these ZFNs resulted in permanent disruption of endogenous TCR expression and a highly enriched αβ TCR^null cell population could be generated by paramagnetic bead selection. These data support our plans to develop allogeneic T cells as “off-the-shelf” biologics that can be infused on demand as “drugs”.