Evolution of the Correlation between Expression Divergence and Protein Divergence in Mammals

Abstract

Divergence of protein sequences and gene expression patterns are two fundamental mechanisms that generate organismal diversity. Here, we have used genome and transcriptome data from eight mammals and one bird to study the positive correlation of these two processes throughout mammalian evolution. We demonstrate that the correlation is stable over time and most pronounced in neural tissues, which indicates that it is the result of strong negative selection. The correlation is not driven by genes with specific functions and may instead best be viewed as an evolutionary default state, which can nevertheless be evaded by certain gene types. In particular, genes with developmental and neural functions are skewed toward changes in gene expression, consistent with selection against pleiotropic effects associated with changes in protein sequences. Surprisingly, we find that the correlation between expression divergence and protein divergence is not explained by between-gene variation in expression level, tissue specificity, protein connectivity, or other investigated gene characteristics, suggesting that it arises independently of these gene traits. The selective constraints on protein sequences and gene expression patterns also fluctuate in a coordinate manner across phylogenetic branches: We find that gene-specific changes in the rate of protein evolution in a specific mammalian lineage tend to be accompanied by similar changes in the rate of expression evolution. Taken together, our findings highlight many new aspects of the correlation between protein divergence and expression divergence, and attest to its role as a fundamental property of mammalian genome evolution.

Genomics of Adaptation to Multiple Concurrent Stresses: Insights from Comparative Transcriptomics of a Cichlid Fish from One of Earth’s Most Extreme Environments, the Hypersaline Soda Lake Magadi in Kenya, East Africa

“…However, the relationship between gene expression and protein sequence evolution remains largely debated. Some studies support decoupling of gene expression from sequence evolution as is suggested by our study, whereas several other studies have suggested a positive correlation between the two evolutionary processes in several species including bacteria, yeast (Kim and Yi 2007), Drosophila (Lemos et al 2005), ants (Hunt et al 2013), and mammals (Khaitovich et al 2005;Warnefors and Kaessmann 2013). Several factors have been suggested that may influence the evolution of protein sequences and gene expression including organismic attributes (e.g., protein protein interactions), mutation rates, and the strength of selection, which in turn determine if the two processes are coupled or decoupled.…”

Evolutionary constraints shape caste-specific gene expression across 15 ant species

“…Such caution is especially important given the lack of a clear neutral baseline expectation (comparable to neutral expectation in sequence evolution) for expression evolution (Helanterä and Uller ), and our modest understanding of the prevalent evolutionary processes underlying expression evolution (Khaitovich et al. ; Warnefors and Kaessmann ; Rohlfs et al. ; Hodgins‐Davis et al.…”

“…In other words, not all genes showing low phylogenetic signal are necessarily ones where expression patterns are evolving under strong positive or divergent selection, or unconstrained. Such caution is especially important given the lack of a clear neutral baseline expectation (comparable to neutral expectation in sequence evolution) for expression evolution (Helanterä and Uller 2014), and our modest understanding of the prevalent evolutionary processes underlying expression evolution (Khaitovich et al 2006;Warnefors and Kaessmann 2013;Rohlfs et al 2014;Hodgins-Davis et al 2015), including the role of neutral drift in expression evolution. Nevertheless, the observed associations of connectivity, expression levels, total variation in expression levels, and phylogenetic signals give us some confidence in interpreting low phylogenetic signal as a measure of how flexible expression patterns of a given gene are.…”

A quantitative model for characterizing the evolutionary history of mammalian gene expression

“…Initial studies on an evolutionary model for gene expression have also been hampered by small datasets, leading to conflicting results. Indeed, some studies concluded that expression is diverging at a linear rate across time, suggestive of widespread neutral evolution (22)(23)(24)(25), whereas others determined that the evolutionary rate is constant, suggestive of strong purifying selection (20,(26)(27)(28)(29).…”