Method of Gene Regulation Influences Intrinsic Noise Independently of Expression Levels


Promoter Sequence Determines the Relationship between Expression Level and Noise


Lucas B. Carey David van Dijk Peter M. A. Sloot Jaap A. Kaandorp Eran Segal Robert Singer


Published in: PLoS Biology
Volume: 11, Issue: 4, Pages: e1001528-
Published: 4/2/2013
DOI: 10.1371/journal.pbio.1001528

Altmetric Statistics:

Our Discussion:

Print Friendly

Cell-to-cell variability in gene expression (a.k.a. gene expression noise) in a genetically identical population of cells is the result of a combination of extrinsic and intrinsic “noise”. Extrinsic noise is the result of variations in the cellular environment (e.g. cells on the outer edge of a colony/biofilm vs. cells in the interior) and the abundance or state of intracellular signaling molecules and transcription factors. Intrinsic noise is the result of random fluctuations in molecular concentrations of mRNA and protein due to the stochastic nature of intermolecular interactions, especially when molecule numbers are very small. Because of the stochastic nature of transcription factor interactions with promoters and RNAPII, transcription also tends to occur in bursts in which multiple transcripts are produced from each transcriptional activation event.

It is generally agreed among experimental and computational biologists that genes that are expressed at low levels are prone to higher levels of expression noise. In a recent paper by Carey et al. in PLOS Biology, the authors show that, while this is true in the case of most genes, noise depends on the specifics of how transcription is regulated. The authors wanted to explore gene expression noise in the context of varying levels of transcription factor activity. They chose to look at the expression of 16 genes regulated by the Zap1 transcription factor, which is inhibited by zinc in a dose dependent-manner. 16 different constructs were made containing the  native promoters of these genes driving YFP expression. These constructs were integrated at the his3 locus. Experiments consisted of determining single cell fluorescence distributions in various concentrations of zinc using flow cytometry at numerous time points. The authors also measure expression in variants of the translation start site and then fit their results to a mathematical model for gene expression.

This paper makes several conclusions:

1) that increasing Zap1 activity results in an increased transcriptional burst frequency

2) translational efficiency determines burst size

3) genes at the same level of transcription can have different levels of noise

4) at low transcription levels, noise due to burst frequency dominates, while at high transcription levels, noise due to burst size dominates

5) a combination of high burst frequency and low burst size can reduced noise of very low expression genes, ensuring that all cells have at least some of the protein

6) at some promoters, Zap1 acts as both and activator and a repressor. This reduces the sensitivity of the target gene to fluctuations in Zap1 concentration (extrinsic noise)

The general implication of this paper is that promoter architecture plays a dominant role in determining the level of gene expression noise. This is a little disappointing because there are few similarly in-depth studies in eukaryotes (much more is known about promoter structure effects on gene expression noise in prokaryotes), so it’s not clear what the rules are for designing a promoter that will produce the desired behaviour. On the bright side though, this means that the effects of extrinsic noise on transcription can be minimized.