Skip to main content
Advertisement

< Back to Article

A genome-scale metabolic model for the denitrifying bacterium Thauera sp. MZ1T accurately predicts degradation of pollutants and production of polymers

Fig 1

Workflow to build the metabolic model of Thauera sp. MZ1T using a semiautomatic approach.

An initial draft M-model was reconstructed using three sets of BLASTp parameters (e-value, query length, and identity percentage) from three template models present in BiGG (Escherichia coli K-12 substr. MG1655, Klebsiella pneumoniae subsp. pneumoniae MGH 78578, and Pseudomonas putida KT2440). NCBI reference sequence annotation (GenBank) was employed in GPR associations. The RAVEN and COBRA toolboxes for MATLAB were employed in the reconstruction, refinement, and validation of the model. The resulting optimized draft model and constituents of the BOF were manually curated. Protein, RNA, and DNA components of the BOF were estimated based on the total coding sequences. Disconnected metabolites were linked to the metabolic pathways using bioinformatics databases and experimental evidence. Four detailed metabolic modules were carefully added to the M-model to show specific metabolic capabilities of Thauera sp. MZ1T: 1) aromatic compound degradation under aerobic and anaerobic conditions, 2) N metabolism including denitrification (with nitric and nitrous oxide partial denitrification), oxidative phosphorylation with nitrate as electron acceptor, and DNRA, 3) PHAs and PHB production, and 4) EPS precursor production. The resulting model was validated using experimental data retrieved from the literature. The iterative model refinement process included manual curation, gap-filling, and curation under heterotrophic conditions with different oxygen concentrations depending on the experimental environments. PHB and EPS production was simulated using a set of 36 C sources under aerobic and anaerobic conditions to estimate the compounds with higher yields. The final model (iThauera861), containing 1,744 metabolites, 2,384 reactions, and 861 genes, predicted growth up to 95% of accuracy for 60 C and N substrates.

Fig 1

doi: https://doi.org/10.1371/journal.pcbi.1012736.g001