Trispecific targeting of T cells engineered with TCR mimic antibodies to limit antigen escape

Background Antigen loss and tumor heterogeneity present significant challenges for successful immunotherapies. T-cell receptor (TCR)-based therapies rely on the recognition of epitopes derived from intracellular tumor proteins presented by major histocompatibility complex class I molecules on cell surface. Solid tumor cells frequently lack immunoproteasomes, which are crucial for processing and presenting certain immunogenic epitopes. An effective strategy to mitigate the risk of antigen absence and tumor heterogeneity is to simultaneously target multiple tumor antigens, thereby providing critical rescue from disease relapse. Previously, we engineered a TCR mimic monoclonal antibody (TCRm) “ESK2”, specific for Wilm’s tumor 1 (WT1)-derived epitope RMFPNAPYL (RMF) in the context of HLA-A2, into a new chimeric antigen receptor T-cell format, antibody-TCR receptor (AbTCR)-chimeric signaling receptor (CSR). However, the RMF epitope is largely dependent on processing by the immunoproteasomes, which can be lost from leukemia cells and sometimes absent in solid tumor cells. Methods To mitigate antigen loss, tumor heterogeneity and broaden the reach of AbTCR T cells, we combined ESK2 with a new TCRm for an immunoproteosome-independent epitope derived from WT1, VLDFAPPGA (VLD), in the context of HLA-A2 molecules, named ESK3. ESK2 and ESK3 were tandemly engineered into one AbTCR-CSR construct, simultaneously recognizing both the WT1 RMF and VLD epitopes. To add additional specificity and potency, a CSR in these cells was engineered with a single chain variable fragment (scFv) for either CD33 to treat leukemia or mesothelin to treat solid tumors. The specificity and efficacy of the AbTCR-CSRs were evaluated in both in vitro and in vivo. Results In vitro studies demonstrated that the Tri-AbTCR-CSR (CD33 CSR) T cells showed the best killing activity against most acute myeloid leukemia cells. Similar levels of cytotoxicity were exhibited by ESK3 AbTCR-CSR (mesothelin CSR) against most solid tumor cell lines when compared with the Tri-AbTCR or a combination of ESK2 and ESK3 AbTCR-CSR. In animal therapy models, trispecific AbTCR-CSR T cells showed efficacy equivalent to single ESK2-AbTCR or ESK3-AbTCR-CSR T cells, against hematopoietic or solid tumor cells, further supporting the advantage of triple targeting strategy, overcoming epitope loss variants. Conclusions Trispecific T cells targeting immunoproteasome-dependent and independent epitopes of WT1 peptide/HLA-A2 complexes, plus a CSR recognizing a third tumor-associated antigen, present an effective and cost-efficient approach for overcoming tumor immune evasion.