Principles of metabolome conservation in animals

Metabolite concentrations shape cellular physiology and disease susceptibility, yet the general principles governing metabolome evolution are largely unknown. Here we introduce a measure of conservation of individual metabolite concentrations among related species. By analysing multispecies metabolome datasets in mammals and fruit flies, we show that conservation varies extensively across metabolites. Three major functional properties, metabolite abundance, essentiality and association with human diseases predict conservation, highlighting a striking parallel between the evolutionary forces driving metabolome and protein sequence conservation. Metabolic network simulations recapitulated these general patterns, and revealed that abundant metabolites are highly conserved due to their strong coupling to key metabolic fluxes in the network. This study uncovers simple rules governing metabolic evolution in animals and implies that most metabolome differences between species are permitted, rather than favored by selection. More broadly, our work paves the way towards using evolutionary information to discover biomarkers, as well as to detect pathogenic metabolome alterations in individual patients.