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

Abstract

Gene expression patterns depend on the interaction of diverse transcription factors with their target genes. While many factors have a restricted number of targets, some appear to affect transcription globally. An example of the latter is MeCP2; an abundant chromatin-associated protein that is mutated in the neurological disorder Rett Syndrome. To understand how MeCP2 affects transcription, we integrated mathematical modelling with quantitative experimental analysis of human neurons expressing graded levels of MeCP2. We first used a model of MeCP2-DNA binding to demonstrate that changes in gene expression reflect MeCP2 density downstream of transcription initiation. We then tested five biologically plausible hypotheses for the effect of MeCP2 on transcription. The only model compatible with the data involved slowing down of RNA polymerase II by MeCP2, causing reduced transcript output due to polymerase queueing. Our general approach may prove fruitful in deciphering the mechanisms by which other global regulators choreograph gene expression.

MeCP2 Represses the Rate of Transcriptional Initiation of Highly Methylated Long Genes

“…Moreover, there is significant overlap between the genes misregulated in MeCP2 and Dnmt3a mutant mice, suggesting that mCA deposition is required for MeCP2-dependent gene repression in the mature brain. Recent mathematical modeling of MeCP2 function suggested that MeCP2 might function by slowing transcriptional elongation (Cholewa-Waclaw et al, 2019), but our genome-wide experimental measurements of transcriptional elongation and initiation rates in the brain led to the surprising observation that MeCP2 inhibits transcriptional initiation of highly methylated long genes through its interaction with the NCoR co-repressor, without having any detectable effect on transcriptional elongation. Consistent with this idea, decreased binding of the NCoR complex component Hdac3 at the TSS has been reported in MeCP2deficient hippocampus (Nott et al, 2016).…”

“…Given the broad distribution of MeCP2 binding to chromatin and the relatively subtle changes in mRNA levels when MeCP2 function is disrupted, it has been difficult to understand the mechanisms by which MeCP2 regulates gene expression. The previous finding that genes that are long and have high levels of gene body mCA are preferentially upregulated with loss of MeCP2 (Chen et al, 2015;Gabel et al, 2015;Kinde et al, 2016;Sugino et al, 2014) provided a potential inroad to address this issue and led to the hypothesis that MeCP2 binding to gene body mCA sites might inhibit transcriptional elongation (Cholewa-Waclaw et al, 2019;Kinde et al, 2016). However, these observations largely relied on a meta-analysis of previously published studies or small sample sizes.…”

“…In addition, the levels of mCA and MeCP2 protein increase significantly as RTT ensues in the postnatal period (Gabel et al, 2015;Lister et al, 2013;Skene et al, 2010). The preferential repression of long genes with high levels of gene body mCA and MeCP2 binding has led to the hypothesis that MeCP2 might function as a transcriptional speed bump, in which it acts as a weak repressor of elongating RNA polymerase (Pol II) (Cholewa-Waclaw et al, 2019;Kinde et al, 2016).…”

Molecular Systems Biology of Neurodevelopmental Disorders, Rett Syndrome as an Archetype

“…Techniques such as chromatin immunoprecipitation (ChIP) combined with RNA sequencing and/or quantitative proteomics, as we will discuss below, could solve this issue entirely. In fact, recent efforts pairing experimental design with mathematical modeling are heading in this direction (Cholewa-Waclaw et al, 2019).…”

FMRP attenuates activity dependent modifications in the mitochondrial proteome

“…This is evident in the case of gene defects in FMR1 and MECP2, causative of Fragile X (FXS) and Rett Syndrome, respectively, whose deficiency alters homeostatic plasticity responses [1,12,14,17,18]. FMR1 and MECP2 control the expression of a vast number of genes at the translational and transcriptional levels in neurons [19][20][21][22][23][24]. This suggests that FMR1-or MECP2-dependent gene expression mechanisms likely regulate multiple factors necessary for the establishment, modulation, and maintenance of homeostatic plasticity.…”