BPS2026 – Development of a multiscale framework for engineering protein condensates as microreactors

Living cells organize the biochemical reactions that drive metabolism in space and time into organelles. Membraneless organelles (also known as biomolecular condensates), formed by phase separation, can yield enhanced reaction rates for native processes like purine synthesis and so present a unique opportunity for control of reaction dynamics of synthetic pathways for metabolic engineering purposes. In this work scaffold proteins were designed to recruit client enzymes and small molecules of interest to biomolecular condensates using a ML-driven bioinformatics algorithm and molecular dynamics simulations. Because of the locally increased concentration of protein and substrate, due to mass-action the rate of enzymatic activity within the dense phase of the condensate is elevated over homogeneous solution. Here this toolkit was implemented for two sustainable manufacturing applications: to increase the in vitro enzymatic synthesis of acetoin, a flavoring compound and precursor to biofuels and α-hydroxyketone pharmaceuticals; and to increase the in vivo utilization of xylose, one of the most abundantly available types of renewable raw materials. This unified theoretical, computational, and experimental platform can be applied to design of synthetic condensates as reactors for green industrial chemical production.