Advancing the Prediction of Unexpected Cellular Behavior Due to Enzyme Promiscuity and Enzyme Solubility

The introduction of non-native synthesis pathways into microbial hosts has been instrumental in transforming cellular organisms into microbial factories that produce commercially useful biomolecules. While experimental efforts have achieved significant success, they are costly and time consuming. The development of computational methods that analyze the feasible design space promise to guide experimentation and expedite novel discoveries. This thesis introduces computational workflows that advance the design of biological systems through the consideration of promiscuous actions of enzymes. While enzymes are recognized to be promiscuous in many biological engineering applications, the effects of promiscuity on cellular physiology is understudied and not well documented. We show that predicting derivative products due to enzyme promiscuity explains metabolomics measurements for Escherichia coli in ECMDB, the E. coli Metabolome Database. Further, predicted products and their catalyzing reactions are suggested to extend iML1515, the most comprehensive genome-scale model of Escherichia coli MG1655. We also show that, after the introduction of heterologous enzymes, cellular hosts have unexpected cellular activity due to two disruption scenarios: the cellular host enzymes are promiscuous towards the synthesis pathway metabolites, and the synthesis pathways enzymes are promiscuous towards host metabolites. The latter may explain unexpected results in the experimental literature. This thesis also introduces a workflow for bridging the gap between synthesis pathway construction in a cellular host and the identification of their genetic implementation. Genes can be sourced from various organisms; however, not all genes are soluble in the host. Our technique, Probabilistic Pathway Assembly with Solubility Confidence Scores (ProPASS), links synthesis pathway construction with the exploration of soluble enzymes within the host. Predicted protein solubility scores are used as a confidence level to quantify the compatibility of each pathway enzyme with the host (E. coli). ProPASS offers experimental guidance to avoid trial-and-error approaches to enzyme selection