iPSC-Derived Dual Antigen Receptor T Cells Targeting GD2 and LMP2 Antigens for Extranodal NK/T-Cell Lymphoma, Nasal Type

Epstein-Barr virus (EBV)-associated lymphomas generally exhibit poor prognosis. Given that EBV antigens, LMP1 and LMP2, are often expressed on EBV-associated lymphomas, these lymphomas should be a good target of antigen-specific cytotoxic T lymphocytes (CTL) therapy. Among these, advanced-stage ENKL is a highly aggressive disease. We previously reported that functionally rejuvenated LMP2-specific CTLs (LMP2-rejTs), generated from induced pluripotent stem cells (iPSCs), robustly suppress ENKL in vivo with remarkable persistence for more than seven months. To mitigate the risk of tumor antigen escape, we next generated dual-antigen receptor rejuvenated T cells from iPSCs by introduction of LMP1-directed chimeric antigen receptor (CAR) into LMP2-rejTs which have already demonstrated in vivo persistence. We reported that these cells recognize dual antigens via CAR and native T cell receptor, resulting in cooperative antitumor effects in vivo.

We have recently confirmed that ENKL cells often express disialoganglioside (GD2). GD2 has been identified as a promising target for CAR T-cell therapy aimed at GD2-expressing tumors. A number of Phase 1 clinical trials, utilizing GD2-directed CAR (GD2-CAR) T-cell therapy in neuroblastoma patients, have demonstrated not only safety but also efficient in vivo expansion, all while avoiding significant toxicities.

For clinical implementation, we have developed iPSC-derived dual antigen receptor T cells, where we incorporated GD2-CAR into LMP2-rejTs, resulting in GD2-CARrejTs. We reprogrammed an LMP2-CTL clone derived from a healthy donor into iPSCs using a Sendai virus vector, followed by transduction with the lentiviral GD2-CAR vector. Utilizing flow cytometry, we evaluated CAR transgene expression in the GD2-CARrejTs and found a substantial expression rate of 91.8%. Additionally, we determined a nearly perfect LMP2 antigen specificity of 99.9% for the GD2-CARrejTs, also measured by flow cytometry.

To investigate whether GD2-CARrejTs exhibited cytotoxicity against ENKL than GD2-CARTs, we performed ⁵¹Cr release assays. The cytotoxicity of GD2-CARrejTs was significantly stronger against an ENKL cell line (NK-YS), than that of GD2-CARTs (79.8% vs 24.8%; p < 0.0001) at an effector to target ratio of 40:1. Next, to observe the in vivo antitumor effect of GD2-CARrejTs against NK-YS, NK-YS cells labeled with Firefly luciferase / GFP were intraperitoneally injected into NOG mice. Four days after tumor injection, mice were divided into untreated (control) and treated (GD2-CARrejTs) groups. Tumor signal was significantly suppressed in mice treated with GD2-CARrejTs compared to untreated mice on Day 35 (p = 0.0318).

To delineate the disparities in cytotoxicity, we undertook a single-cell RNA sequencing analysis and confirmed that GD2-CARrejTs exhibited significantly lower TIGIT expression in comparison to GD2-CAR T cells. In fact, we noted an amplification of the cytotoxic impact of GD2-CAR T cells when paired with TIGIT antibodies against GD2-expressing tumor cells. However, the enhanced cytotoxicity still did not exceed that of GD2-CARrejTs.

From these results, we conclude that GD2-CARrejTs may constitute a promising therapeutic approach against refractory ENKL. The most significant advantage of iPSC-derived T cell therapy lies in its clonality and potential for an unlimited supply of therapeutic T cells. Additional gene-editing of iPSC-derived CARrejT cells could potentially lay the groundwork for allogeneic, "off-the-shelf" GD2-CARrejT therapy, thereby introducing a promising new therapeutic avenue for patients with ENKL in the foreseeable future.