Multidimensional heuristic process for high-yield production of astaxanthin and fragrance molecules in Escherichia coli

Congqiang Zhang,Vui Yin Seow,Xixian Chen,Heng-Phon Too

doi:10.1038/s41467-018-04211-x

Congqiang Zhang, Vui Yin Seow + Show 2 more

Open Access

https://doi.org/10.1038/s41467-018-04211-x

Copy DOI

Abstract

Optimization of metabolic pathways consisting of large number of genes is challenging. Multivariate modular methods (MMMs) are currently available solutions, in which reduced regulatory complexities are achieved by grouping multiple genes into modules. However, these methods work well for balancing the inter-modules but not intra-modules. In addition, application of MMMs to the 15-step heterologous route of astaxanthin biosynthesis has met with limited success. Here, we expand the solution space of MMMs and develop a multidimensional heuristic process (MHP). MHP can simultaneously balance different modules by varying promoter strength and coordinating intra-module activities by using ribosome binding sites (RBSs) and enzyme variants. Consequently, MHP increases enantiopure 3S,3′S-astaxanthin production to 184 mg l−1 day−1 or 320 mg l−1. Similarly, MHP improves the yields of nerolidol and linalool. MHP may be useful for optimizing other complex biochemical pathways.

Highlights

Optimization of metabolic pathways consisting of large number of genes is challenging
Astaxanthin has been producing at low levels in metabolically engineered Escherichia coli[8,9,10,11,12], Saccharomyces cerevisiae[13,14,15], and Corynebacterium glutamicum[16]
An example is the multivariate modular metabolic engineering (MMME) approach[24], which elegantly segments metabolic pathways into modules regulated by distinct promoters

Summary

Introduction

Optimization of metabolic pathways consisting of large number of genes is challenging. When involving metabolic pathways with large numbers of genes, MMME approaches will require the exploration of a large number of possible combinations of modules To mitigate this challenge, experimental designs (e.g., fractional factorial design) and regression models are used to pinpoint globally optimal phenotypes and identify contributions from the various modules and abiotic conditions[25,26,27]. MHP is a modular pathway optimization process to generate high producer strains by assembling and screening diverse libraries of predefined regulatory elements together with key enzyme variants in a high dimensional combinatorial manner.

Methods

Results

Conclusion