Inferring Metabolic States in Uncharacterized Environments Using Gene-Expression Measurements

Department of Bioinformatics (CMBI), Centre for Molecular Life Sciences, Radboud University Nijmegen, The Netherlands
PLoS Computational Biology (Impact Factor: 4.62). 03/2013; 9(3):e1002988. DOI: 10.1371/journal.pcbi.1002988
Source: PubMed


The large size of metabolic networks entails an overwhelming multiplicity in the possible steady-state flux distributions that are compatible with stoichiometric constraints. This space of possibilities is largest in the frequent situation where the nutrients available to the cells are unknown. These two factors: network size and lack of knowledge of nutrient availability, challenge the identification of the actual metabolic state of living cells among the myriad possibilities. Here we address this challenge by developing a method that integrates gene-expression measurements with genome-scale models of metabolism as a means of inferring metabolic states. Our method explores the space of alternative flux distributions that maximize the agreement between gene expression and metabolic fluxes, and thereby identifies reactions that are likely to be active in the culture from which the gene-expression measurements were taken. These active reactions are used to build environment-specific metabolic models and to predict actual metabolic states. We applied our method to model the metabolic states of Saccharomyces cerevisiae growing in rich media supplemented with either glucose or ethanol as the main energy source. The resulting models comprise about 50% of the reactions in the original model, and predict environment-specific essential genes with high sensitivity. By minimizing the sum of fluxes while forcing our predicted active reactions to carry flux, we predicted the metabolic states of these yeast cultures that are in large agreement with what is known about yeast physiology. Most notably, our method predicts the Crabtree effect in yeast cells growing in excess glucose, a long-known phenomenon that could not have been predicted by traditional constraint-based modeling approaches. Our method is of immediate practical relevance for medical and industrial applications, such as the identification of novel drug targets, and the development of biotechnological processes that use complex, largely uncharacterized media, such as biofuel production.

Download full-text


Available from: Sergio Rossell
    • "Model SEED Henry et al., 2010 GLAMM Bates et al., 2011 SuBliMinaL toolbox Swainston et al., 2011 RAVEN toolbox Agren et al., 2013 Path2Models Büchel et al., 2013 EMRA Lee et al., 2014 BNICE Pathway prediction Hatzimanikatis et al., 2005 PathPred Moriya et al., 2010 RetroPath Carbonell et al., 2014a DESHARKY Rodrigo et al., 2008 PathPred Moriya et al., 2010 RELATCH Kim and Reed 2012 XTMS Carbonell et al., 2014b PROM Chandrasekaran and Price 2010 ORACLE Miskovic and Hatzimanikatis 2010 GEM-Path Campodonico et al., 2014 OptKnock Gene Knockout Choon et al., 2014 OptGene Patil et al., 2005 MOMA Segrè et al., 2002 EMILiO Yang et al., 2011 SIMUP Gawand et al., 2013 GDLS Lun et al., 2009 OptForce Pathway amplification Ranganathan et al., 2010 OptORF Kim and Reed, 2010 CosMos Cotten and Reed 2013 FSEOF Nocon et al., 2014 GIMME Transcriptome Becker and Palsson, 2008 E-Flux Colijn et al., 2009 MADE Jensen and Papin 2011 TIGER Jensen et al., 2011 iMAT Bordbar et al., 2012 TEAM Collins et al., 2012 AdaM Töpfer et al., 2012 GX-FBA Navid and Almaas 2012 mCADRE Wang et al., 2012b FCGs Kim et al., 2013 EXAMO Rossell et al., 2013 GIMMEp Proteome Bordbar et al., 2012 GIM 3 E Metabolome Schmidt et al., 2013 MET-IDEA Broeckling et al., 2006 MZmine Katajamaa et al., 2006 Fig. 2. DNA assembly technologies for libraries construction. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Metabolic engineering has expanded from a focus on designs requiring a small number of genetic modifications to increasingly complex designs driven by advances in genome-scale engineering technologies. Metabolic engineering has been generally defined by the use of iterative cycles of rational genome modifications, strain analysis and characterization, and a synthesis step that fuels additional hypothesis generation. This cycle mirrors the Design-Build-Test-Learn cycle followed throughout various engineering fields that has recently become a defining aspect of synthetic biology. This review will attempt to summarize recent genome-scale design, build, test, and learn technologies and relate their use to a range of metabolic engineering applications.
    No preview · Article · Oct 2015 · Metabolic Engineering
    • "Although a draft model can be based on annotated genome sequences, it needs extensive manual curation based on additional biochemical and physiological knowledge. Such generalised genomebased models can be made specific for certain environmental conditions or differentiation stages by restricting them to reactions that are actually active according to transcriptome and proteome data [11] [14] [15]. In multicellular organisms, such data can be used to refine genome-scale models to reflect specific cell types [16]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: To rationalise drug target selection, we should understand the role of putative targets in biological pathways quantitatively. We review here how experimental and computational network-based approaches can aid more rational drug target selection and illustrate this with results obtained for microbes and for cancer. Comparison of the drug response of biochemical networks in target cells and (healthy) host cells can reveal network-selective targets.
    No preview · Article · Aug 2015 · Drug Discovery Today Technologies
  • Source
    • "By integrating transcriptomic data with a global human metabolic model using this method, they predicted tissue-specific metabolic activity in ten different tissues. A method called EXAMO (EXploration of Alternative Metabolic Optima) is an extended version of iMAT that builds a context-specific model [46]. Tailored gene expression using user-defined thresholds may avoid data normalization issues [33]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Several computational methods have been developed that integrate transcriptomic data with genome-scale metabolic reconstructions to infer condition-specific system-wide intracellular metabolic flux distributions. In this mini-review, we describe each of these methods published to date with categorizing them based on four different grouping criteria (requirement for multiple gene expression datasets as input, requirement for a threshold to define a gene's high and low expression, requirement for a priori assumption of an appropriate objective function, and validation of predicted fluxes directly against measured intracellular fluxes). Then, we recommend which group of methods would be more suitable from a practical perspective.
    Full-text · Article · Aug 2014 · Computational and Structural Biotechnology Journal
Show more

Similar Publications