Decoupling Environment-Dependent and Independent Genetic Robustness across Bacterial Species

Abstract

The evolutionary origins of genetic robustness are still under debate: it may arise as a consequence of requirements imposed by varying environmental conditions, due to intrinsic factors such as metabolic requirements, or directly due to an adaptive selection in favor of genes that allow a species to endure genetic perturbations. Stratifying the individual effects of each origin requires one to study the pertaining evolutionary forces across many species under diverse conditions. Here we conduct the first large-scale computational study charting the level of robustness of metabolic networks of hundreds of bacterial species across many simulated growth environments. We provide evidence that variations among species in their level of robustness reflect ecological adaptations. We decouple metabolic robustness into two components and quantify the extents of each: the first, environmental-dependent, is responsible for at least 20% of the non-essential reactions and its extent is associated with the species' lifestyle (specialized/generalist); the second, environmental-independent, is associated (correlation = approximately 0.6) with the intrinsic metabolic capacities of a species-higher robustness is observed in fast growers or in organisms with an extensive production of secondary metabolites. Finally, we identify reactions that are uniquely susceptible to perturbations in human pathogens, potentially serving as novel drug-targets.

Figure 1. Illustration of a simple metabolic network producing an essential constituent of biomass on different growth media.

Figure 2. The distribution of non-essential, essential and conditionally-essential/non-essential reactions versus environmental diversity across the 487 organisms studied.

Lines represent the linear regression calculated for each group.

Figure 3. The metabolic networks of species with similar network size and different topological properties.

(A) Clostridium botulinum: Network size, 189; connectivity, 5.2; centrality (mean shortest path), 3.7; robustness (NGR), 0.85. (B) Helicobacter acinonychis: Network size, 191; connectivity, 4.1; centrality (mean shortest path), 5.4; robustness (NGR), 0.56. Red circles - essential reactions; green circles - non-essential reactions.

Figure 4. Observed versus predicted NGR.

Predicted values are derived from a generalized linear predicting NGR from the growth rate and fraction of secondary metabolites of each species. Growth rate data was available for 109 species including 17 anaerobic (red), 37 aerobic (blue), 40 facultative (green), 4 microaerophilic, and 11 unknown.

Figure 5. Distribution of pathways for the synthesis of 10-formyltetrahydrofolate in human pathogens and human commensal organisms.

Maroon squares: metabolites; blue squares: reactions. MCH: methenyltetrahydrofolate cyclohydrolase; FTL: formyltetrahydrofolate synthetase; HC: human commensals; HP: human pathogens. 10-Formyltetrahydrofolate acts as a formyl donor in purine biosynthesis, and for formylation of methionyl-tRNA required for producing fMet-tRNA – a molecule required in most bacterial species for initiating protein synthesis. All human commensals (7/7) contains two alternative routes for the production of 10-formyltetrahydrofolate. Only 28 out of 73 human pathogens which have MCH contain the alternative route, making MCH essential in the remaining 45 organisms. These 45 pathogenic organisms include several Shigella, Salmonella and Mycobacterium species (the full list of species and the essentiality of MCH and FTL is provided in Text S1 Note 13 and in Table S5). The approach presented here can easily be generalized for highlighting essentiality in other groups of medical, ecological or agricultural interest.