Grant: Discovery Grants Program - Individual

Delays in gene expression due to transcription and translation are known to be important in many fundamental biological processes, such as somitogenesis in which transcriptional delays are responsible for oscillations translated into spatial patterning by growth of the tissue. However, these delays pose a conundrum for the modeler. What delays must we include, and what properties of the delay distribution matter, keeping in mind that adding complexity to a model has conceptual, analytical and computational costs? Our recent work has shown that clearance delays, the time during which a promoter or ribosome binding site is occluded preventing the initiation of, respectively, new transcription or translation events, can have a surprisingly large stabilizing effect on the steady states of a model. At this point, we have only a few examples, so it is not clear whether this effect is general. I will systematically study this question using a hierarchy of models of increasing complexity, but constructed so that we can perform general analyses and therefore draw general conclusions. The vast majority of gene expression models in the literature use fixed delays, which is to say that they assume that the time taken to, e.g., transcribe a gene has a vanishingly narrow distribution. However, real delay distributions are not particularly narrow. Using a similarly hierarchy of models as in the project above, I will study the effects of a variety of realistic distributions of delays with varying statistical properties on experimental properties of interest: multistability (biologically: pluri- or multi-potency), switching times following stimulation of a signaling pathway, periods and amplitudes of biological oscillators, etc. Splicing is an important (possibly dominant) contributor to the mRNA expression time. Unfortunately, there has been very little mathematical modeling of splicing. Moreover, there are a number of contradictory results in the experimental literature with respect to even very basic questions such as how long splicing takes, or whether it occurs co- or post-transcriptionally. In order to help resolve these controversies, we will build a detailed mechanistic model of splicing. We will use this model to examine some of the contradictory results in the literature, which will greatly enhance our collective understanding of splicing. Moreover, we will use our model of splicing to develop expressions (probably using approximation methods) for the time required to synthesize a mature mRNA. All of these studies are designed to give modelers concrete recommendations with respect to the inclusion of delays in gene expression models, ranging from general questions such as which delays to include and how to represent them, to the specific issue of the contribution of splicing to expression delays. This will result in models whose conclusions contribute to our understanding of biology with a diminished need for caveats.