Stochastic analysis of gene expression noise
Date
2021
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Stochastic variation in the level of gene product amongst cells of the same population is ubiquitous across cell types and organisms. The species involved in gene expression exist at low copy numbers which consequently amplifies the inherent probabilistic nature of the complex set of biochemical reactions associated with the gene expression process. Analytical expressions are developed to quantify noise in gene expression, allowing for systematic comparisons of different regimes and the role they play in noise attenuation. We incorporate stochastic time delays into gene expression models, where the delay is an independent and identically distributed random variable. After characterizing the effect of time delays, we further explore the effects of time delays when subject to extrinsic fluctuations. Here, we find counter-intuitive results in regards to the incorporation of an extrinsic factor. When it is incorporated, non-monotonic noise behavior emerges when the mRNA Fano factor is plotted as a function of increasing RNA transport time. We also find that for both low and high extrinsic factor timescale, gene expression noise is buffered, however, noise increases at mid-level extrinsic species timescale. The results in this thesis are obtained through moment dynamics and linear noise approximation to analyze noise at steady state. Next, we apply the analysis of gene expression noise to the specific context of the Human Immunodeficiency Virus (HIV). As previously mentioned, noise can have advantages and disadvantages for the cell. In the case of HIV, noise drives a key cell fate decision: between active replication and viral latency, known as a major obstacle in the eradication of HIV and HIV therapies. We study the bimodal distribution of protein level, corresponding to the alternate cell fate decisions, that result from three different circuits that incorporate feedback strategies. Using stochastic simulations from Gillespie algorithm, we ultimately reveal that Tat-mediated transcriptional positive feedback combined with precursor auto-depletion of nuclear RNA species results in the greatest stability in cell fate. Last, experimental data from live-cell imaging of gene switching in developmental enhancers is analyzed. Analysis of the active and inactive time intervals of the gene reveal memory in the time spent in the active state.
Description
Keywords
Gene expression, Noise suppression, Regulatory mechanisms, Stochastic analysis, HIV therapies