| Metabolic engineering promises to reduce our reliance on non-renewable chemical synthesis methods by harnessing microbial metabolisms to convert simple renewable resources, such as sugars, into useful chemicals, such as biofuels. Most metabolic engineering efforts require expressing heterologous enzymes in a microbial host, such as E. coli. Both the process of producing enzymes and the chemical metabolites that are produced can have deleterious impacts on host fitness, however. Here we detail efforts to use experimental and computational methods to better understand how to mitigate the deleterious impact of a metabolic pathway on its host. Firstly, we sought to computationally predict how metabolic pathway titers could be improved, and product toxicity reduced, by implementing genetic feedback networks that can sense pathway metabolite concentrations in a cell and respond by up- or down-regulating enzyme levels. To this end, we developed a computational methodology for modeling and simulating large sets of genetic feedback networks acting on a metabolic pathway that produces 4-aminocinnamic acid, or 4-ACA. Our analysis revealed genetic feedback network architectures and implementations that promise to improve pathway titers. Secondly, we experimentally analyzed the burden imposed by expressing a Phenylalanine Ammonia Lyase enzyme, PAL2 from Arabidopsis thaliana, which converts 4-aminophenylalanine (4-AF) to 4-ACA in Escherichia coli cells. We uncovered how the timing and level of induction of a PAL2-superfolder GFP fusion (PAL2-sfGFP) influences cell growth, expression of PAL2-sfGFP, and in vivo cellular catalysis of 4-AF to 4-ACA conversion. We then identified stationary phase promoters that can be used to autonomously express high levels of PAL2-sfGFP, increase conversion, and reduce deleterious impacts. Lastly, we produced 4-AF from glucose in E. coli using a three-enzyme operon, PapABC, from Pseudomonas fluorescens, achieving titers of around 100 mg/L, forming the basis for a pathway to produce 4-ACA. Together, this work serves to improve our understanding of how to mitigate the deleterious impacts of a metabolic pathway on its host by using computational model analysis and by controlling in vivo enzyme expression dynamics. |
