To understanding carbon assimilation in cassava storage roots, a transcriptome-integrated metabolic model has been developed called the rMeCBMx model with significant improvement in carbon flux prediction in the complicated branched pathways: (1) carbon substrates supplied via oxidative pentose phosphate pathway, (2) TCA cycle and glycolysis in respiration pathway, and (3) carbon precursors for alanine biosynthesis. These reactions were consistently weakly predicted by the basic FBA model. This study showed that integration of gene expression data could enhance the prediction power of the previous rMeCBM model, yielding more biologically relevant information on carbon assimilation in developing storage roots of cassava.
Read through in our journal article on the topic of “Transcriptome integrated metabolic modeling of carbon assimilation underlying storage root development in cassava” by Ratchaprapa Kamsen and colleges in Scientific Reports.
The existing genome-scale metabolic model of carbon metabolism in cassava storage roots, rMeCBM, has proven particularly resourceful in exploring the metabolic basis for the phenotypic differences between high and low-yield cassava cultivars. However, experimental validation of predicted metabolic fluxes by carbon labeling is quite challenging. Here, we incorporated gene expression data of developing storage roots into the basic flux-balance model to minimize infeasible metabolic fluxes, denoted as rMeCBMx, thereby improving the plausibility of the simulation and predictive power. Three different conceptual algorithms, GIMME, E-Flux, and HPCOF were evaluated. The rMeCBMx-HPCOF model outperformed others in predicting carbon fluxes in the metabolism of storage roots and, in particular, was highly consistent with transcriptome of high-yield cultivars. The flux prediction was improved through the oxidative pentose phosphate pathway in cytosol, as has been reported in various studies on root metabolism, but hardly captured by simple FBA models. Moreover, the presence of fluxes through cytosolic glycolysis and alanine biosynthesis pathways were predicted with high consistency with gene expression levels. This study sheds light on the importance of prediction power in the modeling of complex plant metabolism. Integration of multi-omics data would further help mitigate the ill-posed problem of constraint-based modeling, allowing more realistic simulation.
Cite this article:
Kamsen, R., Kalapanulak, S., Chiewchankaset, P. et al. Transcriptome integrated metabolic modeling of carbon assimilation underlying storage root development in cassava. Sci Rep 11, 8758 (2021)