Cellular responses to reactive oxygen species can be predicted on multiple biological scales from molecular mechanisms

Abstract

All aerobically growing microbes must deal with oxidative stress from intrinsically-generated reactive oxygen species (ROS), or from external ROS in the context of infection. To study the systems biology of microbial ROS response, we developed a genome-scale model of proteome damage and maintenance in response to ROS, by extending a genome-scale metabolism and macromolecular expression (ME) model of E. coli. This OxidizeME model recapitulated measured microbial oxidative stress response including metalloenzyme inactivation by ROS and amino acid auxotrophies. OxidizeME also correctly predicted differential expression under ROS stress. We used OxidizeME to investigate how environmental context affects the flexibility of ROS stress response. The context-dependency of microbial stress response has important implications for infectious disease. OxidizeME provides a computational resource for model-driven experiment design in this direction.The human immune system includes phagocytes that use reactive oxygen species (ROS) to combat pathogens. Most microbes are considerably weakened by this oxidative stress, whereas certain pathogens can grow inside the phagosome [1]. How microbes, including pathogens, evolve to tolerate such intense oxidative stress remains a long-standing question. To answer it requires a systemslevel understanding of the many interlinked biochemical and physiological adaptations for responding to ROS damage. Here, we approach this question through in silico experiments subjecting E. coli to varying degrees of oxidative stress in a multitude of environments. To this end, we extended a genome-scale model of E. coli metabolism and protein expression by incorporating ROS damage and response mechanisms.By their chemical nature, ROS can react with a broad array of macromolecules, often with deleterious consequences. In particular, a number of protein targets of ROS have been experimentally verified in detail; however, most of the potential targets of ROS have not been characterized individually. Thus, to better understand ROS stress, we must answer two questions: which biomolecules are damaged to what extent, and what is the cellular response to this damage? Here, we address these questions, focusing on ROS damage to metalloproteins. To predict cellular response to damage, we developed a novel model called OxidizeME. OxidizeME accounts for ROS damage to iron and iron-sulfur cluster cofactors of over 40 enzymes. We also modeled enzyme-catalyzed repair of oxidized iron-sulfur clusters. OxidizeME is able to predict cellular response at the metabolic and gene expression levels. RESULTSA multiscale model of reactive oxygen species damage and response.We developed the OxidizeME model by extending the iLE1678-ME model of E. coli Metabolism and macromolecular Expression (ME model) [2] to account for damage to proteins by ROS, and cellular response mechanisms (Fig. 1) For each protein, OxidizeME computes the the fraction of inactivated enzyme due to cofactor oxidation as a function of intracellular ROS concentrati...

Adaptations of Escherichia coli strains to oxidative stress are reflected in properties of their structural proteomes

“…In addition to the methodology reported above, a number of improvements were implemented for this study and are slated for incorporation into the publicly available pipeline in ssbio. These improvements include 1) the selection of representative experimental structures with the consideration of bound substrates or cofactors [87]; 2) the definition of membrane spanning domains in transmembrane proteins by consolidating information as predicted by TMHMM [88] or OPM [89] and as annotated in UniProt; 3) the selection of quaternary structures of protein complexes by a breadth-first search matching algorithm to determine the structure with the highest quality and coverage of annotated subunits, either from biological assemblies in the PDB or from predicted complexes in the SWISS-MODEL repository.…”

Modeling the multi-scale mechanisms of macromolecular resource allocation

“…Developmental paths of genome-scale resource allocation models. Models or algorithms listed are: FBAwMC [33], FBA ME (Membrane Economics) [36], MOMENT [35], corsoFBA [34], CAFBA [39], GECKO [38], E-matrix [64], E. coli ME [43,44], T. maritima ME [62], iJL1678 (E. coli ME with the protein translocation network) [4], B. subtilis RBA [37], R. solanacearum [58], SectorME [53], FoldME [50], and OxidizeME [65].…”