Kinetic modeling of biochemical pathways: a computational systems biology study

• Main Authors: Hsih-Te Yang, 楊士德
• Other Authors: Ming-Jing Hwang
• Format: Others
• Language: en_US
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/19045275772410641041

博士 === 國立陽明大學 === 生化暨分子生物研究所 === 96 === Systems biology aims to study biology from a system’s point of view. Such studies usually involve creating a computational model to integrate relevant experimental data in such a way that the resulting model could be used to discover novel biological components and/or underlying working principles for the biological system of interest. This dissertation described work of computational systems biology on two biological systems, each studied by approaches with a different level of computational complexity, as briefly summarized below. (I) Kinetic modeling of gene transcription mediated by a dimeric transcription factor. To model gene transcription kinetics, empirical fitting with the Hill function or S-system is often used. In this study, we derived an analytical rate expression for gene transcription mediated by a dimeric transcription factor such as Gcn4p, a yeast gene regulator, in a manner similar to that developed for enzyme kinetics. We showed that the analytical rate expression and its parameters estimated from several sets of experimental data could accurately reproduce the experimentally measured promoter binding activity of Gcn4p. Furthermore, the analytical rate expression allowed us to derive analytically, rather than fit empirically, the parameters of the Hill function and S-system for use in modeling transcription kinetics. We found that a plot of gene transcription rate against Gcn4p concentration gave a sigmoidal dose-response curve with a positive cooperativity, Hill coefficient (~1.25), in accordance with previous experimental findings on the promoter binding of dimeric transcription factors. Furthermore, the characteristics of the dose-response curve around the estimated cellular Gcn4p concentration suggest that transcription regulation is efficiently controlled under physiological conditions. This work is a useful initial step towards analytically modeling and simulating complicated gene transcription networks. (II) Kinetic modeling of cellular heterogeneity. Isogenic cells can exhibit greatly different gene expression patterns, which underlie the observed cellular heterogeneity. This phenomenon gave us impetus to investigate the origins and controls of cellular heterogeneity by theoretical modeling and simulations. In this work, a stochastic algorithm is implemented to simulate the kinetics of the galactose (GAL) pathway on the level of individual cells. We showed that the stochastic model could account for both the induction kinetics of individual cells with ensemble averages in accord with results of deterministic models. Furthermore, we showed that our stochastic model was able to reproduce the experimentally observed uni- or bi-modal modes of gene expressions, i.e. cellular heterogeneity, of an isogenic cell population. By tracing the promoter state transitions of individual cells and experimenting with transcription factor (TF) binding/unbinding rate constants, we showed that the uni- or bi-modal expressions of the GAL genes are mainly dominated by a slow rate of the promoter binding. This simulation finding supports a suggestion made previously based on experimental observations that varying levels of gene expression are caused by the stability of transcriptional complex formation. This study provides a useful framework for theoretical simulations of complex biochemical pathways, and demonstrates that such simulations can be used to probe the physical and biological underpinnings of interesting cellular phenotypes, including those of a stochastic nature. The results of these studies reinforce the notion that, by carefully utilizing different sources of experimental data, and properly employing relevant simulation techniques, one can successfully predict and explore new biological characteristics manifested by a biochemical pathway using mathematical models.