RT Journal Article SR Electronic T1 Controlling E. coli gene expression noise JF bioRxiv FD Cold Spring Harbor Laboratory SP 013797 DO 10.1101/013797 A1 Kyung Hyuk Kim A1 Kiri Choi A1 Bryan Bartley A1 Herbert M. Sauro YR 2015 UL http://biorxiv.org/content/early/2015/01/14/013797.abstract AB Intracellular protein copy numbers show significant cell-to-cell variability within an isogenic population due to the random nature of biological reactions. Here we show how the variability in copy number can be controlled by perturbing gene expression. Depending on the genetic network and host, different perturbations can be applied to control variability. To understand more fully how noise propagates and behaves in biochemical networks we developed stochastic control analysis (SCA) which is a sensitivity-based analysis framework for the study of noise control. Here we apply SCA to synthetic gene expression systems encoded on plasmids that are transformed into Escherichia coli. The objective of the study was to show that we could differentially control the noise and mean levels of molecular concentrations in biological networks. We show that (1) dual control of transcription and translation efficiencies provides the most efficient way of noise-vs.-mean control. (2) The expressed proteins follow the gamma distribution function as found in chromosomal proteins. (3) Bursting size and frequency are strongly correlated, implying that transcription efficiency can affect transcript lifetimes and/or translation efficiency. (4) Lastly, genetic encoding in plasmids amplifies intrinsic noise of gene expression, showing that the two-promoter state model, commonly used to describe chromosomal gene expression, may need to be modified.