Development of Anti-CD3 Chimeric Antigen Receptor (CAR)-T Cells for Allogeneic Cell Therapy of Peripheral T-Cell Lymphoma (PTCL)

Infusion of CAR-T cells has produced dramatic clinical responses in patients with B-cell leukemias and lymphomas as well as multiple myeloma, but cell-based and potentially curative therapies for PTCL are not available. We sought to develop an effective CAR T-cell therapy for PTCL by targeting CD3.

PTCL develops from mature T-cells and tumors from most subtypes retain high and uniform CD3 expression. CD3 is not expressed in non-hematopoietic tissues but is highly expressed in T lymphocytes. Therefore, expression of an anti-CD3 CAR in T lymphocytes is likely to result in T-cell self-killing, precluding consistent manufacturing of sufficient CAR-T cell numbers. Indeed, when we transduced peripheral blood T-cells with a lentiviral vector delivering a second-generation CAR targeting CD3 (anti-CD3–41BB-CD3zeta), massive fratricide rapidly ensued. To solve this problem, we downregulated surface CD3 expression using a CD3 protein expression blocker (PEBL). CD3 PEBL is composed of an anti-CD3 single-chain variable fragment and an intracellular retention domain that anchors the cognate antigen in the endoplasmic reticulum and Golgi apparatus before degradation. Expression of CD3 PEBL in T-cells completely abolished cell surface CD3 expression, without affecting cell proliferation and cytotoxicity (Kamiya et al. Blood Adv 2018). When T-cells were transduced with lentiviral vectors containing anti-CD3 CAR and CD3 PEBL, fratricide was limited, resulting in high numbers of viable anti-CD3 PEBL-CAR T cells. With peripheral blood cells from 6 healthy donors, the median percentage of CAR+ cells obtained was 45.5% (range 30%-62%) and the median expansion achieved 9-10 days post T-cell activation and transduction was 46-fold (range 11-95). Anti-CD3 PEBL-CAR T cells had potent and specific cytotoxicity against CD3+ target cells in short- and long-term assays and proliferated vigorously in the presence of CD3+ target cells. We further assessed the anti-tumor efficacy of anti-CD3 PEBL-CAR T cells in immunodeficient mice engrafted with CD3+ Jurkat cells expressing firefly luciferase and eGFP. Anti-CD3 PEBL-CAR T cells produced a distinct anti-leukemic effect, with treated mice showing a marked reduction in leukemic cell burden (Fig. 1A).

Importantly, CD3 PEBL also prevented T-cell receptor (TCR)/CD3 display on the cell surface and abolished TCR-mediated signalling. Minimal signs of graft-versus-host disease (GvHD) were detected in immunodeficient mice treated with anti-CD3 PEBL-CAR T cells, whereas mice treated with non-transduced T-cells developed GvHD which caused substantial weight loss (Fig. 1B).

In conclusion, PEBL-mediated blockade of surface TCR/CD3 complex expression prevents fratricide mediated by anti-CD3 CAR and abrogates TCR-based antigen recognition by T-cells. The latter property is critical for the manufacturing of allogenic CAR-T cells without GvHD potential. The technology developed in this study fits well in current cGMP cell manufacturing protocols and it does not pose the risk of off-target gene editing. Taken together, these features warrant testing of this novel anti-CD3 CAR-T therapy in the treatment of relapsed/refractory PTCL.