A hybrid kinetic and constraint‐based model of leaf metabolism allows predictions of metabolic fluxes in different environments

Abstract

SUMMARY While flux balance analysis (FBA) provides a framework for predicting steady‐state leaf metabolic network fluxes, it does not readily capture the response to environmental variables without being coupled to other modelling formulations. To address this, we coupled an FBA model of 903 reactions of soybean (Glycine max) leaf metabolism with e‐photosynthesis, a dynamic model that captures the kinetics of 126 reactions of photosynthesis and associated chloroplast carbon metabolism. Successful coupling was achieved in an iterative formulation in which fluxes from e‐photosynthesis were used to constrain the FBA model and then, in turn, fluxes computed from the FBA model used to update parameters in e‐photosynthesis. This process was repeated until common fluxes in the two models converged. Coupling did not hamper the ability of the kinetic module to accurately predict the carbon assimilation rate, photosystem II electron flux, and starch accumulation of field‐grown soybean at two CO2 concentrations. The coupled model also allowed accurate predictions of additional parameters such as nocturnal respiration, as well as analysis of the effect of light intensity and elevated CO2 on leaf metabolism. Predictions included an unexpected decrease in the rate of export of sucrose from the leaf at high light, due to altered starch–sucrose partitioning, and altered daytime flux modes in the tricarboxylic acid cycle at elevated CO2. Mitochondrial fluxes were notably different between growing and mature leaves, with greater anaplerotic, tricarboxylic acid cycle and mitochondrial ATP synthase fluxes predicted in the former, primarily to provide carbon skeletons and energy for protein synthesis.

Rethinking the potential productivity of crassulacean acid metabolism by integrating metabolic dynamics with shoot architecture, using the example of Agave tequilana

“…To extract useful information from large amounts of ’omics data, computational modeling approaches are increasingly important (Shameer et al ., 2018; Chomthong & Griffiths, 2020; Burgos et al ., 2022). With the methods of model integration we developed previously (Kannan et al ., 2019; Shameer et al ., 2022), our CAM model has the potential to interface with transcriptome data, gene regulatory networks and metabolic models based on flux balance analysis (Cheung et al ., 2014; Shameer et al ., 2018; Töpfer et al ., 2020). This will enable the exploration and assessment of the potential of CAM crops of varying forms in a more comprehensive way, identifying potential targets for yield improvement and contributing to CAM biodesign.…”

A systematic overexpression approach reveals native targets to increase squalene production in Synechocystis sp. PCC 6803

“…However, classic FBA does not account for regulatory mechanisms like feedback inhibition or potential metabolite toxicity ( Knoop and Steuer, 2015 ). It could be helpful for future studies to test FSEOF in combination with dynamic extensions of FBA ( Mahadevan et al., 2002 ) and hybrid kinetic and constraint-based models ( Shameer et al., 2022 ). The incorporation of metabolite concentrations, flux rates or kinetic parameters could drastically improve the precision and reliability of the results ( Mahadevan et al., 2002 ; Moulin et al., 2021 ; Shameer et al., 2022 ).…”

“…It could be helpful for future studies to test FSEOF in combination with dynamic extensions of FBA ( Mahadevan et al., 2002 ) and hybrid kinetic and constraint-based models ( Shameer et al., 2022 ). The incorporation of metabolite concentrations, flux rates or kinetic parameters could drastically improve the precision and reliability of the results ( Mahadevan et al., 2002 ; Moulin et al., 2021 ; Shameer et al., 2022 ). For future studies we suggest the combinatorial expression of the identified amplification targets, especially with the native sqs , since it seems to be the rate limiting enzyme in the present study, as well as the combination of amplification and knock-down targets.…”

Multi-scale models of whole cells: progress and challenges

“…Some system-level kinetic models have been developed e.g., (Klipp, 2007;Bordbar et al, 2015;Jamei, 2016), but they usually tend to account for the activity of significantly fewer genes than COBRA models due to a lack of detailed kinetic data for all cellular processes. There have been many methods developed that use Bayesian parameter estimation to predict reasonable thermodynamic and kinetic values to constrain COBRA models e.g., (Liebermeister and Klipp, 2006a;Liebermeister and Klipp, 2006b;Stanford et al, 2013) and subsequently there have been a number of attempts to add kinetic information to FBA models (e.g., (Jamshidi and Palsson, 2008;Adadi et al, 2012;Stanford et al, 2013;Chowdhury et al, 2015;Pozo et al, 2015;Khodayari and Maranas, 2016;Sánchez et al, 2017;Shameer et al, 2022)). Despite this progress, currently the vast majority of FBA models do not contain kinetic information.…”