(482c) Integrated Analysis of the Trade-Off Between Thermodynamics and Yield in Non-Natural Chemical Biosynthesis Pathways

Metabolic pathway design is essential to enabling the production of valuable non-natural chemicals by microorganisms. There are usually multiple pathways capable of producing a chemical of interest. However, there is likely one or a small set of pathways that is optimal from the perspective of yield, titer and productivity.

Computational pathway prediction methods have become increasingly common in metabolic engineering. Given some reactants and a desired final product, these methods enumerate all possible biosynthetic pathways capable of taking the former to the latter. They are combinatorial in nature and result in a multitude of pathways to the product. For example, the Biopathway Predictor developed by Genomatica identified 10,000 routes from a central metabolite to 1,4-butanediol in four to six steps. The brute force approach to pathway design is experimental characterization. This can be done when the number of pathways is small, but it becomes unmanageable when the number of pathways is greater than a few hundred. In addition, there is a significant time and monetary investment associated with several parallel experimental pathway implementations. Hence, there is a requirement for in silico pathway prioritization methods to identify from the multitude of pathways, the optimal one that balances the objectives of yield, titer and productivity.

The most common metric for the in silico comparison of metabolic pathways is maximum theoretical yield. However, cells have evolved to maximize their growth rather than the production of non-natural chemicals, so the yield of implemented pathways often falls short of the maximum. This restricts the use of maximum theoretical yield to the identification of pathways able to reach yield values above a threshold. However, in addition to the maximum theoretical yield, thermodynamic driving force and kinetic bottlenecks can also limit the flux through these non-natural pathways and hence, additional metrics that consider thermodynamics and kinetics are required to further reduce the number of candidate metabolic pathways.

We apply an integrated metric that leverages pathway thermodynamic and kinetic data for the analysis of non-natural pathways, using 1,4-butanediol pathways as a case study. We develop methods for assessing the variation in the enzyme kinetics of these non-natural chemical biosynthesis pathways and quantify the sensitivity of the metric to errors in this estimation. Furthermore, we use the results from the application of this metric to identify infeasible or rate limiting factors which can be used to direct engineering strategies for pathway optimization. We believe that this metric will enable the efficient prioritization of non-natural chemical biosynthesis pathways and accelerate the development of commercially viable microbial strains.