[Paper Review] A compact model of Escherichia coli core and biosynthetic metabolism

Marco Corrao, Hai He+3 more|arXiv (Cornell University)|Jun 24, 2024

Microbial Metabolic Engineering and Bioproduction4 citations

TL;DR

This paper presents iCH360, a manually curated, medium-scale metabolic model of *Escherichia coli* K-12 MG1655 focusing on core and biosynthetic metabolism. By extracting a subnetwork from the genome-scale iML1515 model and enriching it with updated annotations, thermodynamic data, and quantitative constraints, the model enables more accurate and interpretable analyses such as enzyme-constrained FBA, elementary flux mode analysis, and thermodynamic screening, offering a robust reference for systems biology and metabolic engineering applications.

ABSTRACT

Metabolic models condense biochemical knowledge about organisms in a structured and standardised way. As large-scale network reconstructions are readily available for many organisms, genome-scale models are being widely used among modellers and engineers. However, these large models can be difficult to analyse and visualise and occasionally generate predictions that are hard to interpret or even biologically unrealistic. Of the thousands of enzymatic reactions in a typical bacterial metabolism, only a few hundred form the metabolic pathways essential to produce energy carriers and biosynthetic precursors. These pathways carry relatively high flux, are central to maintaining and reproducing the cell, and provide precursors and energy to engineered metabolic pathways. Focusing on these central metabolic subsystems, we present iCH360, a manually curated medium-scale model of energy and biosynthesis metabolism for the well-studied bacterium Escherichia coli K-12 MG1655. The model is a sub-network of the most recent genome-scale reconstruction, iML1515, and comes with an updated layer of database annotations and with a range of metabolic maps for visualisation. We enriched the stoichiometric network with extensive biological information and quantitative data, enhancing the scope and applicability of the model. In addition, we assess the properties of this model in comparison to its genome-scale parent and demonstrate the use of the network and supporting data in various scenarios, including enzyme-constrained flux balance analysis, elementary flux mode analysis, and thermodynamic analysis. Overall, we believe this model holds the potential to become a reference medium-scale metabolic model for E. coli.

Motivation & Objective

• To address the limitations of large genome-scale models (GEMs), such as computational complexity and biologically unrealistic predictions, by creating a medium-scale model focused on essential metabolic subsystems.
• To overcome the scope limitations of existing core models like ECC and ECC2, which exclude key biosynthetic pathways relevant to metabolic engineering.
• To develop a model that integrates stoichiometric, thermodynamic, and functional annotation data to improve predictive accuracy and biological interpretability.
• To enable advanced analyses such as enzyme-constrained FBA, elementary flux mode (EFM) analysis, and thermodynamic screening by providing a manageable yet comprehensive network.

Proposed method

• The model, named iCH360, is derived as a subnetwork of the iML1515 genome-scale model, focusing on central and biosynthetic pathways essential for energy and precursor production.
• Manual curation was performed to ensure consistency in reaction directionality, gene-reaction associations, and functional annotations, with integration of updated database mappings.
• The stoichiometric network was enriched with thermodynamic data, including standard Gibbs free energy changes ($\Delta G'^\circ$), and uncertainty bounds via a probabilistic framework.
• A quadratically constrained program (QCP) was formulated to compute the Max-Min Driving Force (MDF), incorporating uncertainty in thermodynamic estimates and physiological metabolite concentration bounds.
• Enzyme-constrained flux balance analysis (ecFBA) was applied using the model to assess flux distributions under enzyme capacity constraints.
• Probabilistic thermodynamic analysis (PTA) was performed using the PTA package to compute flux-force efficacy ($\eta$) and evaluate thermodynamic states under physiological conditions.

Experimental results

Research questions

• RQ1 Can a medium-scale metabolic model of *E. coli* be constructed that retains key biosynthetic and core metabolic pathways while improving interpretability and accuracy over large GEMs?
• RQ2 How does the integration of thermodynamic constraints and uncertainty improve the biological realism of flux predictions in a curated model?
• RQ3 To what extent can the iCH360 model support advanced analyses such as elementary flux mode (EFM) analysis and enzyme-constrained FBA compared to genome-scale models?
• RQ4 How well do predicted flux-force efficacies correlate with experimental enzyme abundance data in a physiologically relevant context?
• RQ5 Can the model serve as a reliable reference for metabolic engineering and systems biology applications due to its balance of detail, accuracy, and tractability?

Key findings

• The iCH360 model comprises 360 reactions and 334 metabolites, forming a compact yet comprehensive representation of *E. coli* core and biosynthetic metabolism, derived from the iML1515 GEM.
• The model includes updated gene-reaction mappings and extensive functional annotations, enhancing its utility for systems biology and model interpretation.
• Thermodynamic analysis revealed that 72 reactions with flux-force efficacy $\eta > 0.5$ operate far from equilibrium, indicating strong driving forces consistent with high flux and irreversible operation.
• The Max-Min Driving Force (MDF) computation, using a 90% confidence interval and physiological metabolite concentration bounds, yielded a robust estimate of thermodynamic feasibility.
• Flux-force efficacy analysis showed a strong correlation between high $\eta$ values and high enzyme abundance, validating the model’s predictive power for physiological flux states.
• The model enables efficient application of advanced analysis techniques such as EFM and ecFBA, which are computationally infeasible for full-scale GEMs, thus offering a practical alternative for detailed metabolic investigation.