Share

Efficient Strategies for Metabolic Pathway Optimization

Cellular metabolism allows all living organisms — humans and bacteria alike — to convert nutrients into life's building blocks. By combining enzymes from multiple species into a non-natural configuration, new synthetic metabolic pathways may be constructed to manufacture high-value chemicals from low-value biorenewable feedstock. However, identifying the enzyme expression levels that maximize a pathway's productivity remains a major challenge. While enzyme expression is typically altered by either knocking out its gene (zero expression) or greatly overexpressing it on a plasmid, the best enzyme expression levels are usually somewhere in between. We develop systematic methodologies for determinining the optimal enzyme expression levels and their corresponding synthetic DNA sequences, resulting in greater product yield, titer, and productivity. Importantly, these optimization methods do not require biophysical knowledge of the enzymes' interactions or kinetics, which are often unavailable.

Combinatorial Optimization

In engineering, we are often confronted with optimizing a process or product with multiple degrees of freedom — variations in temperature, pressure, flow rate, or material choices — to improve its efficiency. Combinatorial optimization is a procedure that systematically explores a system's degrees of freedom to find the optimal configuration. In metabolic engineering, combinatorial optimization is often employed to vary a metabolic pathway's enzyme expression levels. Random mutations are introduced into the DNA sequence that encodes a pathway and these mutations vary the rate-limiting steps of enzyme production. However, without knowing which specific DNA mutations to perform, this approach results in combinatorial explosion — trillions of mutation combinations yields few beneficial results at extremely high costs.

We are developing a new way to perform combinatorial optimization for metabolic pathways that employs the RBS Library Calculator to systematically vary a metabolic pathway's enzyme expression levels across a 100,000-fold scale while minimizing the number of DNA mutations. The RBS Library Calculator optimizes a degenerate (one-to-many) ribosome binding site sequence that will vary a targeted protein's expression between a minimum and maximum level with a selected number of intermediate points. These RBS libraries contain between 5 and 50 sequences and yield uniform sampling across the selected enzyme expression level range. Our Optimized Search approach eliminates combinatorial explosion and enables the efficient optimization of many-enzyme synthetic metabolic pathways.

enzyme expression levels

The predicted enzyme expression levels of a two-enzyme metabolic pathway are shown, comparing the differences between a random RBS library and an optimized RBS library, generated using the RBS Library Calculator. Each point represents a single pathway variant with varying amounts of enzyme. 1000 pathway variants are shown in both cases. Experimental characterization is employed to identify the optimal pathway variant, which can be located anywhere within the expression level space.

Convergent Optimization

More on this in the future.