Essential Plasticity and Redundancy of Metabolism Unveiled by Synthetic Lethality Analysis

Abstract

We unravel how functional plasticity and redundancy are essential mechanisms underlying the ability to survive of metabolic networks. We perform an exhaustive computational screening of synthetic lethal reaction pairs in Escherichia coli in a minimal medium and we find that synthetic lethal pairs divide in two different groups depending on whether the synthetic lethal interaction works as a backup or as a parallel use mechanism, the first corresponding to essential plasticity and the second to essential redundancy. In E. coli, the analysis of pathways entanglement through essential redundancy supports the view that synthetic lethality affects preferentially a single function or pathway. In contrast, essential plasticity, the dominant class, tends to be inter-pathway but strongly localized and unveils Cell Envelope Biosynthesis as an essential backup for Membrane Lipid Metabolism. When comparing E. coli and Mycoplasma pneumoniae, we find that the metabolic networks of the two organisms exhibit a large difference in the relative importance of plasticity and redundancy which is consistent with the conjecture that plasticity is a sophisticated mechanism that requires a complex organization. Finally, coessential reaction pairs are explored in different environmental conditions to uncover the interplay between the two mechanisms. We find that synthetic lethal interactions and their classification in plasticity and redundancy are basically insensitive to medium composition, and are highly conserved even when the environment is enriched with nonessential compounds or overconstrained to decrease maximum biomass formation.

The Genome-Wide Interaction Network of Nutrient Stress Genes in Escherichia coli

“…Recently, several groups have attempted to predict synthetic lethal gene pairs in E. coli and other pathogens using metabolic models (45–47), and many of the predicted interactions involve the 82 nutrient stress genes that we have tested experimentally. Our data have confirmed only a small number of the predicted lethal interactions, such as the interaction of panF with the pantothenate biosynthesis genes panBCD , metL with thrA , and the interaction between the NAD biosynthesis genes nadABC and the gene pncB from the NAD salvage pathway (45, 46).…”

Detecting the Significant Flux Backbone of Escherichia coli metabolism

“…This nutritionally rich medium contains the set of compounds defining the minimal medium , i.e., a set of minerals salts and four metabolites representing carbon, nitrogen, phosphorus, and sulfur sources, in addition to the following compounds: amino acids, purines and pyrimidines, biotin, pyridoxine, thiamin, and the nucleotide nicotinamide mononucleotide (see Ref. for specific details).…”

Drug target identification through systems biology

“…A particularly strong point of these models is that they can be used to study synthetic lethality, that is, whether there are two or more proteins that, only when depleted at the same time, will block the flux towards an essential metabolite [10,20]. It is possible to discriminate between synthetic lethal interactions that are due to either a backup process (essential plasticity) or a parallel process (essential redundancy) [21]. Drugging multiple targets at once may prevent easy development of resistance.…”