Phenotype-specific estimation of metabolic fluxes using gene expression data

Area of Science:

• Systems Biology
• Metabolic Engineering
• Computational Biology

Background:

• Cellular metabolism is influenced by gene expression, but post-transcriptional regulation creates discrepancies between transcriptomes and metabolic fluxomes.
• Accurately inferring metabolic flux from transcriptomic data requires accounting for unobserved regulatory mechanisms.

Purpose of the Study:

• To develop a method for systematically accounting for unobserved mechanisms in fluxome inference given a transcriptome.
• To present Pheflux, a constraint-based model designed to infer the most likely and least biased fluxome.

Main Methods:

• Developed Pheflux, a constraint-based model that maximizes Shannon's entropy of fluxes per mRNA.
• Validated Pheflux using 13C metabolic flux analysis (MFA) data from yeast and bacteria.
• Applied Pheflux to analyze transcriptomes from The Cancer Genome Atlas (TCGA) for normal and tumor cells.

Main Results:

• Pheflux accurately estimates the carbon core metabolism when benchmarked against experimental 13C fluxes.
• Analysis of TCGA data revealed significantly higher glucose yields on lactate in breast, kidney, and bronchus-lung tumor cells compared to normal counterparts.
• The observed metabolic alterations in tumor cells are consistent with the Warburg effect.

Conclusions:

• Pheflux provides an efficient method for inferring cellular metabolic states from transcriptomic data.
• The model can be effectively utilized to study eukaryotic cell metabolism, particularly in the context of cancer.
• Findings support the role of altered glucose metabolism, characteristic of the Warburg effect, in various cancer types.