UC Irvine

Gene regulation is required for proper organismic development, functioning, and reproduction. Many gene regulatory systems focus on modulating distinct features such as transcriptional noise, fidelity to transcription factors (TFs), or ultrasensitivity in gene expression. These processes generally take place before the initiation of transcription and include the remodeling of chromatin topologies. Chromatin regulation partly involves unpacking the tightly bound heterochromatin as a prerequisite for transcriptional enzymes to access genetic regions. Levels of chromatin density respond ultrasensitively to TFs which results in sharp boundaries between heterochromatin and euchromatin. Viable mechanisms for this ultrasensitivity have been proposed but typically rely on cooperative assumptions which can lead to unbounded chromatin expansion. Downstream from this event, the chromatin needs to be properly structured for enhancers to be in close contact with their respective promoters. These groups of enhancers, commonly denoted as shadow enhancers, have been shown to buffer against environmental stress but the mechanisms underlying this function remain unclear. Here, we present a series of stochastic and deterministic chemical reaction network models that provide sufficient conditions for shadow enhancers and chromatin remodeling to achieve their regulatory targets. We developed a model of the Kruppel shadow enhancers to show that separation of TFs between enhancers is sufficient for achieving lower expression noise. Additional models were generated to determine how shadow enhancer numbers potentially modulate transcriptional noise and fidelity. Separately, a model of the chromatin that does not rely on cooperative interactions between nucleosomes is shown to achieve ultrasensitive gene expression. By limiting complexity, we show that our chemical reaction network models convey clear mechanisms by which TFs, shadow enhancers, and nucleosome interactions can be used to optimize these transcriptional properties.