Background Chimeric antigen receptor therapy – although very efficacious in B cell malignancies – is facing many challenges which limit its success in solid tumors, e.g. on-target off-tumor toxicities, antigen heterogeneity, lack of T cell migration into tumors and an immunosuppressive tumor microenvironment. To better control on-target off-tumor effects and address antigen heterogeneity we developed a modular approach where we equipped T cells with a synthetic agonistic receptor (SAR). The SAR is only activated in the presence of a bispecific antibody (BiAb) cross-linking the receptor with a tumor-associated antigen. While we could show efficacy of the SAR platform in different models, limited infiltration and immune suppression still hamper its function. We could previously demonstrate that T cell infiltration can be enhanced by transduction with carefully chosen chemokine receptors like CXCR6, CCR4 and CCR8. At the same time, gene silencing of checkpoint molecules like PD-1 can make T cells more resistant to immunosuppression, thus we assumed that combining these approaches might generate a desired T cell product.

Materials and Methods All constructs had been generated previously by overlap-extension cloning. The EGFRvIII (E3) SAR consists of extracellular EGFRvIII, transmembrane CD28 and intracellular CD28 and CD3ζ. Human CXCR6-GFP, CCR4-GFP and CCR8-GFP are composed of the chemokine receptors fused to GFP via a 2A sequence. Primary human T cells were retrovirally transduced to stably express the SAR and chemokine receptors. We analyzed migration, cytotoxicity and activation of the single and double (E3 SAR and chemokine receptor) transduced T cells. In addition, PD-1 was knocked out using CRISPR-Cas9 and killing kinetics of target cells and T cell activation were assessed.

Results Co-transduction with chemokine receptors significantly increased migration of E3 SAR T cells to their respective ligand while lysis of target-expressing tumor cell and T cell activation in the presence of BiAb were not affected in vitro. Additionally knocking out PD-1 enhanced killing kinetics and activation of E3 SAR and E3 SAR + CXCR6-GFP transduced T cells compared to corresponding mock electroporated T cells.

Conclusions Using the controllable and modular SAR – BiAb platform SAR T cell activation can be limited by stopping BiAb dosing if adverse events occur. In addition, SAR T cells can be redirected to an alternative tumor-associated antigen by exchanging the BiAb in the case of antigen escape. Here we present add-ons to this approach for increased tumor infiltration and resistance to immunosuppression. Since migration is enhanced upon co-transduction with chemokine receptors and target cell lysis is accelerated upon PD-1 knockout in vitro these two additional modifications seem very promising options to further improve tumor control in vivo.

Disclosure Information M. Schwerdtfeger: None. M. Benmebarek: None. F. Märkl: None. C.H. Karches: A. Employment (full or part-time); Significant; Daiichi Sankyo Deutschland GmbH. A. Öner: None. M. Geiger: A. Employment (full or part-time); Significant; Roche. B. Cadilha: None. S. Endres: None. V. Desiderio: None. C. Klein: A. Employment (full or part-time); Significant; Roche. S. Kobold: None.