Modelling and Experimental Results on Stochastic Model Reduction, Protein Maturation, Macromolecular Crowding, and Time-varying Gene Expression.

• Main Author: Dong, Guangqiang
• Other Authors: McMillen, David
• Language: en_ca
• Published: 2009
• Subjects: Gene expression; Noise and dynamics; 0786
• Online Access: http://hdl.handle.net/1807/19264

Gene expression, which connects genomic information to functional units in living cells, has received substantial attention since the completion of The Human Genome Project. Quantitative characterization of gene expression will provide valuable information for understanding the behavior of living cells, and possibilities of building synthetic gene circuits to control or modify the behavior of naturally occurring cells. Many aspects of quantitative gene expression have been studied, including gene expression dynamics and noise in E. coli. The gene expression process itself is stochastic, and modelling approaches have been broadly used to study gene expression noise; however, stochastic gene expression models are usually large and time intensive to simulate. To speed up simulations, we have developed a systematic method to simplify gene expression models with fast and slow dynamics, and investigated when we can ignore the gene expression from the background genome when modelling the gene expression from plasmids. When modelling the noise in gene expression, one usually neglected aspect is the slow maturation process of fluorescent proteins, necessary for the protein to give out fluorescence after it is produced. By modelling, we show that the maturation steps can bring large changes to both the mean protein number and the noise in the model. An unstudied aspect of gene expression dynamics is the time dependent gene expression behavior in E. coli batch culture. Contrary to the usual assumption, we have found, in E. coli batch culture gene expression, that there is no steady state in terms of both the mean number of proteins and the noise. Negative feedback is thought to be able to reduce the noise in a system, and experiments have shown that negative feedback indeed suppresses the noise in gene expression, but the modelling shows that negative feedback will increase the noise. We have found that the increase of noise by feedback is due to the exclusion of extrinsic noise from the model, and that negative feedback will suppress the extrinsic noise while increasing the intrinsic noise. Living cells are crowded with macromolecules, which will, predicted by modelling, make the reaction constant time dependent. Our experimental observation has confirmed this prediction.