Quantitative modelling predicts the impact of DNA methylation on RNA polymerase II traffic

Significance We introduce an interdisciplinary approach to understanding global modulation of gene expression in mammalian cells. Conventional transcription factors target a limited subset of genes, whereas global modulators bind the genome broadly. An example of the latter is MeCP2, which is mutated in neurological disorders. MeCP2 has millions of genomic binding sites, but its effects on gene expression are mostly small scale and incompletely understood at a mechanistic level. Using datasets from genetically modified human neurons, our mathematical approach rigorously distinguishes global effects from experimental noise. This allows us to integrate theory with experiments to discriminate competing mechanistic models. The results indicate that MeCP2 creates “roadblocks” in gene bodies that slow down elongating RNA polymerase II, leading to polymerase queueing. Patterns of gene expression are primarily determined by proteins that locally enhance or repress transcription. While many transcription factors target a restricted number of genes, others appear to modulate transcription levels globally. An example is MeCP2, an abundant methylated-DNA binding protein that is mutated in the neurological disorder Rett syndrome. Despite much research, the molecular mechanism by which MeCP2 regulates gene expression is not fully resolved. Here, we integrate quantitative, multidimensional experimental analysis and mathematical modeling to indicate that MeCP2 is a global transcriptional regulator whose binding to DNA creates “slow sites” in gene bodies. We hypothesize that waves of slowed-down RNA polymerase II formed behind these sites travel backward and indirectly affect initiation, reminiscent of defect-induced shockwaves in nonequilibrium physics transport models. This mechanism differs from conventional gene-regulation mechanisms, which often involve direct modulation of transcription initiation. Our findings point to a genome-wide function of DNA methylation that may account for the reversibility of Rett syndrome in mice. Moreover, our combined theoretical and experimental approach provides a general method for understanding how global gene-expression patterns are choreographed.