Koch, E. L.; Rocabert, C.; Beeravolu Reddy, C.; Guillaume, F.

Changes in gene expression play a fundamental role in the process of adaptation. However, single gene expression levels cannot evolve independently but form highly correlated groups, often corresponding to functional modules. Any evolutionary response is therefore the product of direct selection acting on a respective gene and indirect selection acting on genetically correlated partners. Here, we studied the mechanisms behind evolution in expression in response to a change in climate conditions using whole transcriptomes and fitness data of the red flour beetle Tribolium castaneum before and after adaptation in seven independent selection lines. Genetic covariance between expression levels, as well as with fitness, could predict evolutionary responses and allowed us to quantify contributions from direct and indirect selection. These predictions were highly consistent with observed changes after 20 generations of adaptation and indicate a pivotal role of indirect selection. Consequently, genes with central positions in gene networks experienced the strongest selection and exhibited larger evolutionary changes in expression. Our genomic analysis revealed that selection on expression levels drives parallel allele frequency changes in the respective genes. More connected genes and those carrying expression quantitative trait loci (eQTLs) showed a higher degree of parallel evolution. We demonstrated how a quantitative genetics approach combined with transcriptomics and network analysis provides deep insights into the evolution of gene expression: Network structure can greatly enhance expression evolution through indirect selection when concerted selection pressures act on the functional properties of expression modules.