Event Timing Statistics of Stochastic Gene Expression

Biswas, Kuheli (2021) Event Timing Statistics of Stochastic Gene Expression. PhD thesis, Indian Institute of Science Education and Research Kolkata.

Text (PhD thesis of Kuheli Biswas (15IP016)) 15IP016.pdf - Submitted Version Restricted to Repository staff only Download (11MB)

Abstract

The functionality of various biological processes depends on synergy among different constituents. Various cellular processes like cell division, differentiation, development, migration, death are results of temporally coherent microscopic interactions. It is thus imperative for a cell to regulate temporal precision in the most fundamental step, i.e., gene expression. Gene regulation involves several biochemical reactions such as transcription, translation, degradation, and complex interactions of mRNA and protein molecules and is an inherently stochastic process. Gene expression levels are subjected to unavoidable fluctuations because of their presence in small copy numbers generated by biochemical reactions occurring after random intervals. However, despite these fluctuations, cells have developed several efficient regulatory pathways to maintain their functionality. It is then pertinent to ask how these regulatory strategies maximize the temporal precision of molecular events? This thesis investigates this problem by studying the first-passagetime (FPT) statistics of proteins due to different transcriptional and post-transcriptional regulatory mechanisms. The FPT coefficient of variation can be considered an indicator of temporal fluctuations of regulatory pathways, and we obtain several cellular conditions for which temporal efficiency can be improved. We only consider the intrinsic noise arising from the stochastic kinetics of gene expression and model the system as a linear Markov process. An approximate solution of the master equation and a geometric simplification of the protein dynamics allow us to obtain a general form of the FPT coefficient of variation applicable to many regulatory scenarios. Another crucial consequence of stochasticity in gene expression levels is phenotypic differences for genetically identical cells. Due to noise, a cell can hop between different stable phenotypes, and our motivation is to determine the preferable phenotype for the cell. In this thesis, we consider the core gene regulatory network for epithelial-mesenchymal transition (EMT) during cancer metastasis and investigate the stochastic dynamics of gene expression levels in the presence of random fluctuations. The cells can explore different stable phenotypes - epithelial, mesenchymal, hybrid E/M, and we estimate the mean residence time (MRT), i.e., average time spent in each of these phenotypes. Our numerical analysis indicates that the MRT can successfully capture the stability of cell phenotype. We predict several factors and molecular mechanisms responsible for the stability of the hybrid E/M state, which plays an aggressive role in disease progression. Our results thus identify several potential targets for pushing cells out of a hybrid E/M state and halting metastatic progression.

Actions (login required)

View Item