Comparative analysis of mitochondrial proteomes across the tree of life.

Abstract

Mitochondria arose from the endosymbiosis of a bacterium with an archaea-related host cell about 2 billion years ago. To understand their origins and evolution, we compared experimentally defined mitoproteomes from the MitoCarta Tree of Life project. Across eight organisms, we identified 8,619 distinct mitochondrial proteins within 3,199 families, of which 43% lack Pfam domains. We report 33 protein families conserved in eukaryotic pathogens yet absent in humans, representing promising candidate targets for protozoan infectious diseases. Leveraging our experimentally defined mitoproteomes, we retrained a classifier based on a protein language model to predict mitoproteomes of ∼200 eukaryotes. From this expanded set, we detail the evolutionary trajectories of mitochondria, ranging from clade-specific gene family expansions to extreme mitoproteome reductions seemingly en route to complete organelle loss. Finally, we reconstruct the last eukaryotic common ancestor (LECA) mitoproteome, revealing that LECA possessed a complex mitochondrion capable of both aerobic and anaerobic metabolism.