Mechanism-informed design of allosteric protein sensors

Columbia University Health Sciences, New York NY

Investigators

Abstract

PROJECT SUMMARY The overall vision of our research program is to deeply understand how proteins occupy an ensemble of conformations to enable biological functions, towards computationally designing new, functional protein ensembles from scratch. Designing protein ensembles that sense and respond to environmental perturbations represents a major challenge in biophysics. However, natural proteins have evolved many mechanisms for sense/response behavior, often without requiring dramatic changes to the protein’s average structure, and in the context of protein folds that are strongly conserved across kingdoms. We do not generally understand how natural protein ensembles are conserved and repurposed for diverse functions while maintaining relatively few protein folds because we have few strategies to determine the energetic contribution of every amino acid in a protein to a global conformational change. Understanding the energetics of protein ensembles of all sizes in solution at high resolution would enable a deeper familiarity of how natural sense/response behaviors evolved. Quantitative descriptions of how proteins shift conformations to respond to perturbations would then inform design constraints to guide efforts in reengineering natural protein sensors for functional responses to environmental signals. In the next five years, we plan to develop a new way to understand how natural proteins behave as small molecule sensors and reengineer them to respond to user-defined ligands. First, we will establish and benchmark a method to determine the contribution of each amino acid in a protein to its folding and binding Gibbs free energies from medium-resolution hydrogen-deuterium exchange/mass spectrometry (HDX/MS) data. HDX/MS is a powerful technique to assess protein ensembles in solution. Building on our recent work to solve the mechanism of ligand-driven allostery in a dimeric bacterial transcription factor, we will then apply our method to quantitatively understand how other members of this ancient protein family respond to small molecule ligands, as well as monomeric proteins with the same fold. This work will expand our understanding of the evolution of allosteric conformational changes and oligomerization in this protein fold and provide a detailed roadmap for computationally redesigning these as protein sensors. We then propose to redesign the transcription factors as small molecule sensors in accordance with the energetic requirements for functional response in the family as determined in by HDX/MS using physics-based and machine learning computational methods for two applications: (1) transcriptional regulators in bacteria for biological upcycling of poly(ethylene-terephthalate) plastic waste, and (2) human cell membrane receptors for metabolite sensing in the solid tumor microenvironment.