Validation of RetroPath, a computer‐aided design tool for metabolic pathway engineering

Abstract

Metabolic engineering has succeeded in biosynthesis of numerous commodity or high value compounds. However, the choice of pathways and enzymes used for production was many times made ad hoc, or required expert knowledge of the specific biochemical reactions. In order to rationalize the process of engineering producer strains, we developed the computer-aided design (CAD) tool RetroPath that explores and enumerates metabolic pathways connecting the endogenous metabolites of a chassis cell to the target compound. To experimentally validate our tool, we constructed 12 top-ranked enzyme combinations producing the flavonoid pinocembrin, four of which displayed significant yields. Namely, our tool queried the enzymes found in metabolic databases based on their annotated and predicted activities. Next, it ranked pathways based on the predicted efficiency of the available enzymes, the toxicity of the intermediate metabolites and the calculated maximum product flux. To implement the top-ranking pathway, our procedure narrowed down a list of nine million possible enzyme combinations to 12, a number easily assembled and tested. One round of metabolic network optimization based on RetroPath output further increased pinocembrin titers 17-fold. In total, 12 out of the 13 enzymes tested in this work displayed a relative performance that was in accordance with its predicted score. These results validate the ranking function of our CAD tool, and open the way to its utilization in the biosynthesis of novel compounds.

An automated Design-Build-Test-Learn pipeline for enhanced microbial production of fine chemicals

“…Flavonoids are among the most structurally diverse classes of natural products 19 , and pinocembrin serves as a key precursor to this diversity 20 . Automated enzyme selection for the pinocembrin pathway had already been validated using our RetroPath 21 software and served as a compatible start point for the rest of the pipeline. The four selected enzymes (phenylalanine ammonia-lyase (PAL), chalcone synthase (CHS) and chalcone isomerase (CHI) (all from Arabidopsis thaliana ) and 4-coumarate:CoA ligase (4CL) (from Streptomyces coelicolor , strain ATCC BAA-471/A3(2)/M145)) convert l -phenylalanine to (2 S )-pinocembrin with the requirement for malonyl-CoA (Fig.…”

“…2Combinatorial optimization of the (2S)-pinocembrin pathway through the Design/Build/Test/Learn cycle. a A biosynthetic pathway composed of four enzymes (PAL, 4CL, CHS, and CHI; see Supplementary Table 2) was initially selected 21 . In the first DBTL cycle, a combinatorial library totaling 2592 pathway configurations was designed by varying the order of pathway genes, promoter parts (P trc and P lacUV5 ), and plasmid copy numbers (pSC101 and p15a).…”

“…2a, Build) and sequenced-verified. The library was screened for production of pinocembrin and the pathway intermediate cinnamic acid in E. coli DH5α using an HTP 96-Deepwell plate-based growth pipeline with media and culture conditions approximated to that of Fehér et al 21 This initial library was found to produce pinocembrin titers ranging from 0.002 to 0.14 mg L −1 (Fig. 2a, Test and Supplementary Data 3).…”

“…Other studies have employed additional strategies for boosting titers in the flavonoids pathway, including chassis engineering to increase malonyl-CoA 21,23,24 or phenylalanine 24 availability, resulting in titers of up to 40 mg L −1 21,23,24 ; and regulation of growth through substrate feeding (glucose and phenylalanine) and pH regulation in fermenters, resulting in the highest reported titer of 68 mg L −1 23 . In order to assess the production capabilities of the prototype selected by the pipeline, we performed additional screening for optimal chassis, growth media, and strain.…”

Exploring biochemical pathways for mono-ethylene glycol (MEG) synthesis from synthesis gas

“…The design/identification of biochemical pathways to convert gases to MEG is non-trivial (Medema et al, 2012) and can be tackled with the help of state-of-the-art cheminformatics and metabolic engineering tools, such as BNICE.ch (Hadadi et al, 2016;Hatzimanikatis et al, 2005), DESHARKY (Rodrigo et al, 2008), GEM-path (Campodonico et al, 2014), RetroPath (Fehér et al, 2014), and XTMS . Almost all of these cheminformatics tools are based on the retrobiosynthesis principle (Carbonell et al, 2013;Hadadi and Hatzimanikatis, 2015;Medema et al, 2012), in which a set of defined biotransformation rules are iteratively applied to connect a target compound "retrosynthetically" to the metabolites of either a host organism's metabolic network, or a biochemical database such as KEGG (Kanehisa et al, 2016), MetaCyc (Caspi et al, 2014), and PubChem (Kim et al, 2016).…”