Evolutionary rate coevolution between mitochondria and mitochondria-associated nuclear-encoded proteins in insects

Abstract

The mitochondrion is a pivotal organelle for energy production, and includes components encoded by both the mitochondrial and nuclear genomes. How these two genomes coevolve is a long-standing question in evolutionary biology. Here we initially investigate the evolutionary rates of mitochondrial components (oxidative phosphorylation (OXPHOS) proteins and ribosomal RNAs) and nuclear-encoded proteins associated with mitochondria, across the major orders of holometabolous insects. There are significant evolutionary rate correlations (ERCs) between mitochondria and mitochondria-associated nuclear-encoded proteins, which is likely driven by different rates of mitochondrial sequence evolution and compensatory changes in the interacting nuclear-encoded proteins. The pattern holds after correction for phylogenetic relationships and considering protein conservation levels. Correlations are stronger for nuclear-encoded OXPHOS proteins in contact with mitochondrial-encoded OXPHOS proteins and nuclear-encoded mitochondrial ribosomal amino acids directly contacting the mitochondrial rRNA. Mitochondrial-associated proteins show apparent rate acceleration over evolutionary time, but we suspect this pattern to be due to artifacts (e.g. rate estimation or calibration bias). We find that ERC between mitochondrial and nuclear proteins is a strong predictor of nuclear proteins known to interact with mitochondria, and therefore ERCs can be used to predict new candidate nuclear proteins with mitochondrial function. Using this approach, we detect proteins with high ERCs but not with known mitochondrial function based on gene ontology (GO). Manual screening of the literature revealed potential mitochondrial function for some of these proteins in humans or yeast. Their holometabolous ERCs therefore indicate these proteins may have phylogenetically conserved mitochondrial function. Twenty three additional candidates warrant further study for mitochondrial function based on this approach, including ERC evidence that proteins in the minichromosome maintenance helicase (MCM) complex interact with mitochondria. We conclude that the ERC method shows promise for identifying new candidate proteins with mitochondrial function.