Article Information

• PDF (127 kb)
• Full Text with Large Figures
• Supporting Material
• Cited by in Scopus (2)

PubMed

• Articles by Nicholas J. Guido
• Articles by J.J. Collins

Related Articles

• New horizons for (p)ppGpp in bacterial and plant physiology

  New horizons for (p)ppGpp in bacterial and plant physiology Trends in Microbiology, Volume 14, Issue 1, 1 January 2006, Pages 45-54 Kristien Braeken, Martine Moris, Ruth Daniels, Jos Vanderleyden and Jan Michiels Abstract A hyperphosphorylated guanosine nucleotide, (p)ppGpp, was initially identified as the effector molecule responsible for the stringent response in . However, a rapidly growing number of reports proves that (p)ppGpp-mediated regulation is conserved in many bacteria and even in plants. It is now clear that (p)ppGpp acts as a global regulator during physiological adaptation of the organism to a plethora of environmental conditions. Adaptation is not only essential for surviving periods of stress and nutrient exhaustion but also for the interaction of bacteria with their eukaryotic host, as observed during pathogenesis and symbiosis, and for bacterial multicellular behaviour. Recently, there have been several new discoveries about the effects of (p)ppGpp levels, balanced by RelA–SpoT homologue proteins, in diverse organisms. Abstract | Full Text | PDF (207 kb)

• Control of rRNA Expression by Small Molecules Is Dynamic and...

  Control of rRNA Expression by Small Molecules Is Dynamic and Nonredundant Molecular Cell, Volume 12, Issue 1, 1 July 2003, Pages 125-134 Heath D. Murray, David A. Schneider and Richard L. Gourse Summary The control of ribosomal RNA transcription is one of the most enduring issues in molecular microbiology, having been subjected to intense scrutiny for over 50 years. Rapid changes in rRNA expression occur during transitions in the bacterial growth cycle and following nutritional shifts during exponential growth. Genetic approaches and measurements of initiating nucleoside triphosphate (iNTP) and guanosine 5′-diphosphate, 3′-diphosphate (ppGpp) concentrations and of rRNA promoter activities showed that rapid changes in the concentrations of iNTPs and ppGpp account for the rapid changes in rRNA expression. The two regulatory signals are nonredundant: changes in iNTP concentration dominate regulation during outgrowth from stationary phase, whereas changes in ppGpp concentration are responsible for regulation following upshifts and downshifts during exponential phase. The results suggest a molecular logic for the use of two homeostatic regulatory mechanisms to monitor different aspects of ribosome activity and provide general insights into the nature of overlapping regulatory circuits. Summary | Full Text | PDF (453 kb)

• DksA

  DksA Cell, Volume 118, Issue 3, 6 August 2004, Pages 311-322 Brian J. Paul, Melanie M. Barker, Wilma Ross, David A. Schneider, Cathy Webb, John W. Foster and Richard L. Gourse Summary Ribosomal RNA (rRNA) transcription is regulated primarily at the level of initiation from rRNA promoters. The unusual kinetic properties of these promoters result in their specific regulation by two small molecule signals, ppGpp and the initiating NTP, that bind to RNA polymerase (RNAP) at all promoters. We show here that DksA, a protein previously unsuspected as a transcription factor, is absolutely required for rRNA regulation. In Δ mutants, rRNA promoters are unresponsive to changes in amino acid availability, growth rate, or growth phase. In vitro, DksA binds to RNAP, reduces open complex lifetime, inhibits rRNA promoter activity, and amplifies effects of ppGpp and the initiating NTP on rRNA transcription, explaining the requirement in vivo. These results expand our molecular understanding of rRNA transcription regulation, may explain previously described pleiotropic effects of , and illustrate how transcription factors that do not bind DNA can nevertheless potentiate RNAP for regulation. Summary | Full Text | PDF (558 kb)

• …more

Copyright © 2007 The Biophysical Society. All rights reserved.
Biophysical Journal, Volume 93, Issue 11, L55-L57, 1 December 2007

doi:10.1529/biophysj.107.118687

Biophysical Letters

A Pathway and Genetic Factors Contributing to Elevated Gene Expression Noise in Stationary Phase

Nicholas J. Guido^*, Philina Lee^*, Xiao Wang^*, Timothy C. Elston^† and J.J. Collins^*, ,  

^* Department of Biomedical Engineering and Center for BioDynamics, Boston University, Boston, Massachusetts
^† Department of Pharmacology, University of North Carolina, Chapel Hill, North Carolina

Address reprint requests and inquiries to J. J. Collins.

In a previous study, we predicted and measured the level of gene expression noise in a synthetic gene circuit under different experimental conditions 1. We also carried out stochastic simulations with a molecular kinetic model designed to represent the behavior of our synthetic gene network. Interestingly the stochastic model predicted that when cell division is stopped, fluctuations in protein levels expressed from high copy number plasmids increases. This increase in gene expression noise occurs because the random partitioning of plasmids between daughter cells during cell division tends to reduce plasmid copy number variability within the cell population. This prediction was validated by comparing noise in gene expression from cells undergoing exponential growth to that of cells grown in minimal media or cells in stationary phase. Surprisingly, the gene expression noise measured from cells in stationary phase was even higher than that predicted by our stochastic model. We established that stationary phase gene expression is noisier than exponential phase gene expression, and found that gene expression from our network was even noisier in stationary phase than was predicted by the model. Specifically, there was an increase of 0.1 in the coefficient of variation (CV, standard deviation divided by the mean) in our experimental measurements in stationary phase, which we could not account for with our model. We thus sought to address the source of this noise by identifying genetic factors that contribute to this variation.

In narrowing the field of potential genetic factors that may alter gene expression noise, we considered that the transition to stationary phase is governed by activation of several global regulators that cause cell-wide changes in gene expression, in particular, the ribosome modulation factor RMF and guanosine 3′,5′-bis(diphosphate) (ppGpp). RMF accumulation causes sequestration and inactivation of ribosomal subunits, decreasing the translational capacity of the cell 2,3, while ppGpp accumulation causes downregulation of transcription, translation, and DNA synthesis (Figure 1a).

Display large version of this figure

Figure 1

Graphical representation of a stationary phase-related gene cascade and coefficient of variation (CV) results for deletion mutants and wild-type cells. (a) A simplified model of a stationary phase-related gene cascade. Proteins produced by the relA and spoT genes mediate the production of ppGpp, which in turn activates the rmf gene. RMF protein causes ribosomes to become inactivated. (b) CV versus arabinose levels for our reporter system with (blue) and without (red) the rmf gene. (c) CV versus arabinose levels for our reporter system with (blue) and without (red) the relA and spoT genes. (d) CV versus arabinose levels for our reporter system with (blue) and without (red) the galM gene.

We hypothesized that RMF activity, during stationary phase, increases gene expression noise by decreasing the level of translationally active ribosomes. Deleting rmf should thus lead to decreased gene expression noise.

Using the same gene circuit as in our previous study (Supplementary Material Fig. S1 ), we measured gene expression noise from the green fluorescent protein (GFP), indirectly induced with arabinose. We found that deleting the rmf gene led to decreased variation in gfp expression (an ∼0.05 decrease in CV) during stationary phase, consistent with the above hypothesis (Figure 1b). We suspect that a higher level of translational efficiency in the rmf deletion mutant is responsible for the decreased variation. In cells without RMF, more ribosomes are active during stationary phase which likely increases the translation of protein. When protein translation is increased and the level of transcription remains steady, the protein production CV will be reduced 4,5. We used mathematical modeling to provide further support for this relationship between translational efficiency and gene expression noise (Box 1). In addition, we observed a higher mean GFP fluorescence in the rmf deletion mutant (Supplementary Material Fig. S2 ), substantiating the notion that there is increased translational efficiency resulting from the rmf deletion.

Given our finding that RMF only contributes a 0.05 increase in CV, we suspected that there might be other factors in addition to RMF, such as ppGpp, that contribute to gene expression noise during stationary phase. ppGpp-mediated downregulation of transcription, in addition to translation, could lead to increased gene expression noise. Escherichia coli has two ppGpp synthetases, encoded by relA and spoT. The spoT gene product also has ppGpp degradase activity 6. Deleting both relA and spoT renders cells unable to make ppGpp, which should lead to decreased gene expression noise.

As shown in Figure 1c, deleting relA and spoT decreased gene expression noise substantially. The double mutant has a CV that is 0.1 lower than that of cells with relA and spoT. Part of this effect is likely due to RMF, which is stimulated by ppGpp accumulation; other ppGpp-stimulated factors likely also contribute. It is important to note that the increase in stationary phase noise in both the RMF and ppGpp deletion strains is beyond that which can be accounted for by the measured decrease in the mean expression level, indicating nontrivial contributions from stationary phase-associated factors that are in addition to those causing downregulation of transcription and translation.

We also created and tested a deletion strain of the galM gene as a control to show that the above findings are not the result of nonspecific artifacts of the gene deletions. The expression of galM is not dependent on growth phase, and it does not exert transcriptional or translational control on the genes involved in our reporter system. We measured GFP expression from the galM deletion and wild-type strains in stationary phase and found that there was no change in gene expression noise due to the galM deletion (Figure 1d). In addition, to demonstrate that the rmf and relA/spoT deletions are altering gene expression noise only in stationary phase, we measured the CV of GFP production in both deletion strains and the wild-type strain during exponential phase, and found little difference between them (Supplementary Material Fig. S3 ).

Previous work has shown, by directly manipulating mRNA sequences and DNA sequences regulating specific genes under study, that the biochemical processes of transcription and translation affect gene expression noise 4,7,8,9,10,11,12. Our present work provides evidence for a cellular pathway and associated genetic factors (Figure 1a) that globally affect gene expression noise in a growth-phase specific manner, through their influence on transcription and translation. Phenotypic diversity arising from such noise effects 13, could confer a survival advantage under extreme conditions, such as stationary phase 14,15,16,17.

In the rmf deletion mutant, there are more active ribosomes to carry out translation compared to wild-type cells; this increases the translation rate while transcription remains steady. We hypothesized that increased translational efficiency in protein production can decrease GFP expression noise in our system. To test this hypothesis, we used a model including mRNA transcription, translation, and degradation, similar to previous work 18:

(1)

(2)

(3)