Adaptive landscapes of transcription factors and their in vivo binding sites in 1135 Arabidopsis thaliana genomes

点击翻译按钮获取中文摘要

Gene expression is a crucial factor for determining organismal phenotypes and changes in the regulation of gene expression are consequently major forces in adaptation and evolution. The expression of a gene is controlled by several molecular processes, including the binding of a transcription factor (TF) to specific binding sites (TFBSs) on DNA. One source of variation in gene expression are point mutations in cis-regulatory sequences that lead to altered TF binding affinities. To understand how point mutations affect the binding affinity of a TF, protein binding microarrays (PBM) consisting of all possible double stranded DNA fragments with a length of ten base pairs were developed. Such arrays allow to assign binding affinities of TFs to all contiguous and gapped 8-mers. One beauty of such combinatorically complete data sets is that they uncover the effect of each individual mutation in all genetic backgrounds. Using binding affinities as a proxy for adaptation, the results of PBM experiments are therefore ideal for mapping phenotypes to genotypes and for building adaptive landscapes. Adaptive landscapes are well suited to investigate a central question in biology: which mutational paths are (not) followed by a population and why. The proposed study will for the first time elucidate the evolution of transcription factor binding sites in actual population genomics data of Arabidopsis thaliana. Such studies are now possible, because the proposed work will employ recently published data sets. These include 1,135 genome sequences obtained from a worldwide collection of A. thaliana, binding affinities of 313 TFs determined in PBM experiments, a collection of 529 in vivo determined TF binding sites and transcriptomes of 1,203 plants. Specifically, the project is centered on two main aspects: what are the roles of selection, neutrality and epistasis during the evolution of TFBSs? How repeatable and constraint is the evolution? To address these questions, genotype networks will be employed. These are networks where nodes represent the sequence of one TFBS in the genome. Two nodes are connected if their underlying sequences differ by exactly one nucleotide. By considering binding affinities for each binding site, adaptive landscapes can be reconstructed. This approach allows to take advantage of the rich tool set available for network analysis and was successfully used for investigating empirical data. In summary, the project will advance our understanding of the evolution of TFBSs in natural populations.