Metabolomics-informed coarse-grained model enables prediction of cell-free protein expression dynamics

Cell-free expression systems have gained considerable attention for their potential in biomanufacturing, biosensing, and circuit prototyping. All these endeavors make use of the innate metabolic activity of cell lysates. However, our knowledge of the underlying nature of cell-free metabolism remains lacking. In this work, we use non-targeted mass spectrometry to generate time-course measurements of small molecules in E. coli cell lysates (metabolomics) during cell-free protein synthesis (CFPS) to show that the majority of E. coli 's metabolism is active in cell lysate. Furthermore, these data indicate that protein synthesis is a relatively small metabolic burden on cell lysates compared to the background metabolism, and that the build-up of multiple metabolic waste products is fundamentally responsible for stopping cell-free protein expression as opposed to depletion of fuel sources. We use these insights coupled with high-throughput CFPS experiments to develop a foundational coarse-grained mechanistic model of CFPS which we show can be easily recalibrated using Bayesian parameter inference techniques to provide accurate models across hundreds of novel experimental conditions. This work provides new experimental insights into the effects of cell-free metabolism on CFPS and establishes a novel framework for leveraging existing CFPS models in new experimental contexts, opening the avenue to future cell-free applications aimed at building complex systems, from multilayered biological circuits to synthetic biological cells.