Functional De Novo Proteins as Custom Molecular Tools for Bioengineering: from Targeted Cancer Immunotherapy to Modular Biosensors

Abstract

Traditional protein engineering methods use naturally existing proteins as a starting point and are intrinsically limited to small perturbations of a protein’s original structure and function. Computational de novo protein design is free of such limitations and enables the creation of designer proteins with custom functions and ideal biochemical properties, which make excellent tools to address diverse bioengineering challenges. In this work, I focus on two applications. First, I describe the development of conditionally active cytokine mimics for targeted cancer immunotherapy. We take advantage of the robust folding of a de novo designed Interleukin-2 mimetic protein to design a split version consisting of two components that do not elicit signaling individually but regain activity when co-localized. With this approach, we aim to overcome the severe side-effects of systemic cytokine therapies. Second, I describe the design of modular and tunable biosensors from broadly applicable design principles by using a robust de novo designed protein switch. We demonstrate the modularity of this biosensor platform by designing functional biosensors for multiple analytes of interest.