De novo design and engineering of biological error correction

University Of California, San Francisco, San Francisco CA

Investigators

Abstract

PROJECT SUMMARY / ABSTRACT Biological systems have evolved a remarkable ability of error correction in processes essential for life. Non- equilibrium mechanisms of kinetic proofreading increase the fidelity (correct versus incorrect outcomes) by several orders of magnitude, beyond that expected solely from differences in equilibrium binding to correct versus incorrect binding partners. Yet we are not harnessing this amplification of specificity in applications of biological engineering. The central challenge and goal of this proposal is to pioneer approaches to engineer mechanisms of biological error correction from the ground up that can be integrated generally into engineered systems – for therapeutic purposes or biotechnology – to increase the fidelity of molecular recognition by 2-4 orders of magnitude. Natural error correction systems are complex, incompletely understood, and difficult to manipulate. Here we will instead build generalizable systems de novo with proteins that are computationally engineered to be tunable, controllable, and composable. Based on recent breakthroughs, including from our laboratory, in the precision, accuracy, and complexity of de novo designed protein structures, functions, and dynamics guided by deep learning, we posit that such a transformative advance might now be within reach. We propose conceptual innovations in de novo protein design beyond the state of the field by engineering energy- driven multi-step pathways that implement kinetic proofreading de novo to achieve error correction rivaling that of natural systems. Specifically, we will conceptualize, build, and test generalizable computational approaches to couple mechanisms of protein recognition, modification, conformational change, kinetically controlled assembly, molecular event cascades, and energy consumption. We will build multiple instances of each protein part, molecular function, and system-level behavior, with designed variation in quantitative parameters including rates and affinities. Using in vitro reconstituted multi-protein systems, high-throughput quantitative assays, and modular engineering of cells, we will assess molecular and system-level function using systematic and precise targeted perturbation and parameter sweeps that will be integrated with and inform predictive modeling. Overall, this ambitious plan will lay the foundation for groundbreaking new abilities to correct biological errors and amplify specificity in engineered systems, enabling future advances across discovery science, biomedical applications, and biotechnology.