Metabolic network model guided engineering ethylmalonyl-CoA pathway to improve ascomycin production in Streptomyces hygroscopicus var. ascomyceticus

Abstract

BackgroundAscomycin is a 23-membered polyketide macrolide with high immunosuppressant and antifungal activity. As the lower production in bio-fermentation, global metabolic analysis is required to further explore its biosynthetic network and determine the key limiting steps for rationally engineering. To achieve this goal, an engineering approach guided by a metabolic network model was implemented to better understand ascomycin biosynthesis and improve its production.ResultsThe metabolic conservation of Streptomyces species was first investigated by comparing the metabolic enzymes of Streptomyces coelicolor A3(2) with those of 31 Streptomyces strains, the results showed that more than 72% of the examined proteins had high sequence similarity with counterparts in every surveyed strain. And it was found that metabolic reactions are more highly conserved than the enzymes themselves because of its lower diversity of metabolic functions than that of genes. The main source of the observed metabolic differences was from the diversity of secondary metabolism. According to the high conservation of primary metabolic reactions in Streptomyces species, the metabolic network model of Streptomyces hygroscopicus var. ascomyceticus was constructed based on the latest reported metabolic model of S. coelicolor A3(2) and validated experimentally. By coupling with flux balance analysis and using minimization of metabolic adjustment algorithm, potential targets for ascomycin overproduction were predicted. Since several of the preferred targets were highly associated with ethylmalonyl-CoA biosynthesis, two target genes hcd (encoding 3-hydroxybutyryl-CoA dehydrogenase) and ccr (encoding crotonyl-CoA carboxylase/reductase) were selected for overexpression in S. hygroscopicus var. ascomyceticus FS35. Both the mutants HA-Hcd and HA-Ccr showed higher ascomycin titer, which was consistent with the model predictions. Furthermore, the combined effects of the two genes were evaluated and the strain HA-Hcd-Ccr with hcd and ccr overexpression exhibited the highest ascomycin production (up to 438.95 mg/L), 1.43-folds improvement than that of the parent strain FS35 (305.56 mg/L).ConclusionsThe successful constructing and experimental validation of the metabolic model of S. hygroscopicus var. ascomyceticus showed that the general metabolic network model of Streptomyces species could be used to analyze the intracellular metabolism and predict the potential key limiting steps for target metabolites overproduction. The corresponding overexpression strains of the two identified genes (hcd and ccr) using the constructed model all displayed higher ascomycin titer. The strategy for yield improvement developed here could also be extended to the improvement of other secondary metabolites in Streptomyces species.Electronic supplementary materialThe online version of this article (doi:10.1186/s12934-017-0787-5) contains supplementary material, which is available to authorized users.

Genome‐Scale Metabolic Reconstruction of Actinomycetes for Antibiotics Production

“…In the case of fungi, also producing secondary metabolites, 24 Penicillium GEMs were recently reconstructed using MetaCyc database and a Penicillium rubens GEM as references . Analysis of the resulting Penicillium GEMs revealed that metabolic diversity across the species was observed in secondary metabolism, not primary metabolism, which was consistent with the metabolic conservation analysis of 31 Streptomyces species from another study . GEM‐based pan‐genome analysis will allow better understanding of not just how secondary metabolism works across species, but also evolutionary processes of individual species, including habitats and specific growth conditions.…”

Generation of Streptomyces hygroscopicus cell factories with enhanced ascomycin production by combined elicitation and pathway‐engineering strategies

“…ascomyceticus H16. Escherichia coli JM109 was used to propagate all plasmids (Wang, Wang et al, ). E. coli ET12567/pUZ8002 was used as a nonmethylating plasmid donor strain for intergeneric conjugation with S. hygroscopicus var.…”

“…E. coli ET12567/pUZ8002 was used as a nonmethylating plasmid donor strain for intergeneric conjugation with S. hygroscopicus var. ascomyceticus (Wang, Wang et al, ). The integrative E. coli ‐ Streptomyces shuttle vector pIB139 containing the ermE * promoter ( P ermE* ) was used for gene overexpression in S. hygroscopicus var.…”

“…The construction of a genome‐scale metabolic network model of S. hygroscopicus var. ascomyceticus allowed the identification of two target genes ( hcd and ccr ) that resulted in a 1.43‐fold improvement (reached 438.95 mg/L) when simultaneously overexpressed (J. Wang, Wang, Song, & Wen, ). Most of the target genes engineered in these reports are involved in primary metabolic pathways and ascomycin biosynthesis precursor pathways.…”

The ModelSEED Database for the integration of metabolic annotations and the reconstruction, comparison, and analysis of metabolic models for plants, fungi, and microbes

“…Beyond the capacity of models to give structure and chemical meaning to functional annotations in biology, these models are also valuable for their predictive capacity. Today, models can be used to predict a wide range of biological phenotypes, including: (i) respiration, photosynthesis, and fermentation types (Cheung et al 2014;de la Torre et al 2015;Marcellin et al 2016;Edirisinghe et al 2016;Meadows et al 2016;Mendoza et al 2017;Marshall et al 2017;Chen et al 2018;Shameer et al 2018;Lieven et al 2018); (ii) feasible growth conditions and Biolog phenotype array profiles (Plata et al 2015;Bosi et al 2016;diCenzo et al 2016;Hartleb et al 2016); (iii) essential genes and reactions (Ding et al 2016;Khodayari and Maranas 2016;Zhang et al 2018;Xavier et al 2018;Guzmán et al 2018); (iv) potential existing or engineerable by-product biosynthesis pathways (Alper et al 2005;Milne et al 2011;Park et al 2013;Chen and Henson 2016;Harder et al 2016); and (v) the yields and even titre available for those pathways (Zuñiga et al 2016;Wang et al 2017;Li et al 2018;Niu et al 2019).…”