(19g) Invited Talk: Engineering Specificity in Next-Generation Immunotherapies

Engineered proteins have had profound impacts in the clinical setting. This class of molecules is vast, including a wide variety of functionalities, such as 1) synthetic receptors that turn engineered inputs into augmented natural outputs, 2) inhibitory proteins that block cell-cell and cell-ligand interactions from affecting cell behavior, and 3) molecular imaging agents capable of detecting disease with drastically improved sensitivity and specificity compared to conventional means. Although these advances have been great, many challenges remain in the pipeline of engineering new proteins to meet clinical demands. Chief among these challenges is the engineering of new proteins with specific binding function, such as selectivity for individual protein family members or the ability to inhibit protein-protein interactions. In this talk, I will discuss efforts to overcome each of these challenges and applications of the described technologies. First, I will describe the engineering of new chimeric antigen receptors (CARs) that use mutated IL13 variants to selectively target cancer cells overexpressing the cancer-testis antigen IL13Rα2 while sparing tissue expressing the ubiquitous receptor IL13Rα1. Through stringent in vitro and in vivo experiments, we studied the levels of binding affinity required to achieve this selectivity, and validated these results using in vivo CAR T cell trafficking studies to understand how CAR binding affinity influences CAR T cell localization. Second, I will discuss my lab’s recent efforts in developing a novel yeast surface display platform that uses inhibition rather than target binding as a selective pressure for ligand selection. We have demonstrated the ability to display a natural receptor extracellular domain or natural ligand simultaneously with a candidate inhibitory protein on the yeast surface. Application of an EGFR-Erbitux model system leads to robust detection of the exclusion of a titrated soluble competitor. Recent progress in further optimizing this system for affinity ranges and construct architecture will be discussed.