From DNA sequence to transcriptional behaviour: a quantitative approach

Abstract
This Review presents a unifying quantitative and conceptual framework for translating DNA sequences into transcriptional behaviours. Each DNA-binding molecule has specific DNA sequence preferences (affinities) and, thus, every regulatory sequence defines a unique affinity landscape for each molecule. At given concentrations of DNA-binding molecules, the unique affinity landscape of a regulatory sequence dictates a distinct distribution of molecule-binding configurations and, consequently, a distinct transcriptional output. Accurate models of the DNA sequence preferences of nucleosomes and of many transcription factors are now available. The intrinsic DNA sequence preferences of nucleosomes are major determinants of nucleosome organization in vivo, and partly account for the depletion of nucleosomes around the starts and ends of genes. Nucleosome depletion around transcription factor-binding sites is partly encoded in the nucleosome affinity landscape of the genome and might assist in directing factors to their appropriate genomic sites. Differences in the intrinsic nucleosome affinity landscapes in which factor-binding sites are embedded might allow the same factor to regulate its different targets with different activation dynamics. Two factors that have adjacent binding sites can show indirect binding cooperativity through competition with nucleosomes, allowing some factors to function as activators on some regulatory sequences but as inhibitors on others. The affinity landscape of a regulatory sequence dictates when factors must compete with nucleosomes for access to the DNA, partly explaining the differential requirement for chromatin remodellers at different loci. DNA sequence changes that directly alter the nucleosome affinity landscape of a regulatory sequence might help to drive phenotypic diversity across evolution. Variability in cell-to-cell expression and in DNA replication can be partly explained in terms of the affinity landscape of a DNA sequence for transcription factors and nucleosomes.