Time-course transcriptome data of Actinoplanes sp. SE50/110 during the whole process of acarbose fermentation

ABSTRACT: Metabolic engineering for high-value compounds like acarbose, a type 2 diabetes drug, requires systematic understanding of metabolic regulation. Here, we integrated multi-dimensional systems biology approach in Actinoplanes sp. SE50/110, a non-model bacterium producing acarbose. First, we reconstructed a substantially enhanced genome-scale metabolic model (iASE1267), which features an expanded set of genes, metabolites, and reactions, providing improved metabolic coverage. The model demonstrated high quality with a MEMOTE (metabolic model test suite) evaluation score of 80%, and showed significantly better phenotype simulation under various conditions. By using dual-objective OptRAM in-silico strain design approach on iASE1267 model, we identified two sets of static engineering strategies, overexpression of the 1-epi-valienol-1,7-diphosphate 1-adenylyltransferase AcbR and repression of an adenylosuccinate lyase; overexpression of a dTDP-glucose 4,6-dehydratase and repression of a 4-(cytidine 5'-diphospho)-2-methyl-D-erythritol kinase. Further, the dynamic flux pattern of time-course metabolic models revealed key time-dependent metabolic valves, L-aspartate oxidase (ASPO1), pyruvate carboxylase (PC), and pyruvate kinase (PYK). Additionally, we uncovered a core transcription-metabolism network by utilizing static modification sites as boundary markers, and identified two global negative transcriptional factors (TFs) within this network. Both the two TFs and the four metabolic genes have been experimentally validated with enhanced acarbose titer by 18–23%. This study establishes a framework linking static/dynamic modeling with transcriptional networks to decode metabolic-regulatory mechanisms in non-model bacteria, offering generalizable strategies for optimizing valuable microbial products.

ORGANISM(S): Actinoplanes

PROVIDER: GSE312767 | GEO | 2025/12/10

REPOSITORIES: GEO