Systemic factors dominate mammal protein evolution.

Proteins encoded by highly expressed genes evolve more slowly. This correlation is thought to arise owing to purifying selection against toxicity of misfolded proteins (that should be more crucial for highly expressed genes). It is now widely accepted that this individual (by-gene) effect is a dominant cause in protein evolution. Here, I show that in mammals, the evolutionary rate of a protein is much more strongly related to the evolutionary rate of coexpressed proteins (and proteins of the same biological pathway) than to the expression level of its encoding gene. The complexity of gene regulation (estimated by the numbers of transcription factor targets and regulatory microRNA targets in the encoding gene) is another important cause, which is much stronger than gene expression level. Proteins encoded by complexly regulated genes evolve more slowly. The intronic length and the ratio of intronic to coding sequence lengths also correlate negatively with protein evolutionary rate (which contradicts the expectation from the negative link between expression level and evolutionary rate). One more important factor, which is much stronger than gene expression level, is evolutionary age. More recent proteins evolve faster, and expression level of an encoding gene becomes quite a minor cause in the evolution of mammal proteins of metazoan origin. These data suggest that, in contrast to a widespread opinion, systemic factors dominate mammal protein evolution.