Improving 3-hydroxypropionic acid production in E. coli by in silico prediction of new metabolic targets

Abstract

3-Hydroxypropionic acid (3-HP) production from renewable feedstocks is of great interest in efforts to develop greener processes for obtaining this chemical platform. Here we report an engineered E. coli strain for 3-HP production through the β-alanine pathway. To obtain a new strain capable of producing 3-HP, the pathway was established by overexpressing the enzymes pyruvate aminotransferase, 3-hydroxyacid dehydrogenase, and L-aspartate-1-decarboxylase. Further increase of the 3-HP titer was achieved using evolutionary optimizations of a genome-scale metabolic model of E. coli containing the adopted pathway. From these optimizations, three non-intuitive targets for in vivo assessment were identified: L-alanine aminotransferase and alanine racemase overexpression, and L-valine transaminase knock-out. The implementation of these targets in the production strain resulted in a 40% increase in 3-HP titer. The strain was further engineered to overexpress phosphoenolpyruvate carboxylase, reaching 0.79 ± 0.02 g/L of 3-HP when grown using glucose. Surprisingly, this strain produced 63% more of the desired product when grown using a mixture of glucose and xylose (1:1, C-mol), and gene expression analysis showed that the cellular adjustment to consume xylose had a positive impact on 3-HP accumulation. Fed-batch culture with xylose feeding led to a final titer of 29.1 g/L. These results reinforce the value of computational methods in strain engineering, enabling the design of more efficient strategies to be assessed. Moreover, higher production of 3-HP under a sugar mixture condition points towards the development of bioprocesses based on renewable resources, such as hemicellulose hydrolysates.