Mathematical modelling unravels a novel mechanism of gene expression regulation

In this husband-and-wife seminar we would like to present an example of how statistical physics can help to solve a biological mystery, and how biology can in turn inspire new statistical physics models. Specifically, we will discuss how a protein called MeCP2 regulates gene expression in human neurons. The protein is expressed (produced) mostly in the brain, and its lack or over-production are associated with severe neurological disorders such as Rett syndrome and MeCP2 duplication syndrome (both on the autism spectrum). The protein binds to methylated cytosine in the DNA. Since there are tens of millions of such binding sites distributed roughly uniformly in the genome, the protein does not behave as a typical transcription factor with well defined target gene(s). Many conflicting hypotheses have been proposed to explain what the protein does in the cell, but so far no explanation has been universally accepted.

We will show that the most likely mechanism of gene regulation by MeCP2 is actually quite simple: MeCP2 affects transcription by slowing down the RNA polymerase, the enzyme responsible for making mRNA (a template for protein synthesis). This process can be understood using a simple TASEP-like model with dynamical obstacles. However, to show that the mechanism is correct and to eliminate other plausible mechanisms, we need to use a combination of advanced molecular biology techniques, statistical analysis, and mathematical modelling of MeCP2-chromatin binding and high-throughput sequencing techniques to asses the state of chromatin in the cells with different levels of MeCP2.