Systems biology approach to stringent response

Bacterial cells constantly need to monitor their environment and act accordingly.

The trouble is, bacteria are very small and when you are so very small, all the effects of being quantized in terms of molecule numbers are becoming very strong: number of mRNA molecules for a certain gene is an integer value, and not a very high at that, events of receptor getting activated or RNA polymerase binding to the promoter are stochastic in nature, and since not too many of the individual events occur at a time, they are not averaged out due to the law of large numbers.

All this results in bacteria gambling all the time: some react to stimulus, some don't, some produce more proteins in response to it, some less. This leads to so called phenotypic heterogeneity, when otherwise (genetically) identical bacteria become very different in terms of their responses.

This could be a good thing and also could be a bad thing. Having a collection of different bugs instead of a clone army will provide certain versatility: some are ready for one conditions, and some are ready for others. For instance, some are ready to grow and divide right away and some are slower and more cautious. Both types of cells can be beneficial in different conditions: the active ones will drive the population growth, but will be sensitive to the antibiotic treatment, and the passive ones will wait until the treatment is over and then they will come to life. Sounds like a good strategy (and it has a name, this strategy - "bed hedging") and I guess it is exactly the reason why clone armies never caught on.

On the other hand, this noise makes it really hard for bacteria to make an educated guess and respond to a stimulus in the best possible way - there is simply too much background to filter through!

One of the widely used cellular response systems used by bacteria is so called stringent response. This one is mediated by a family of proteins called RSH (RelA SpoH Homologue, with RelA and SpoT being the first members to be discovered). These proteins sense different cues (aminoacid starvation, fatty acid starvation, heat shock and so on) and translate these input signals into modulation of the intracellular concentration of ppGpp - a modified G nucleotide which acts as a second messenger modulating numerous cellular processes, such as transcription, translation, replication and lots more.

The RSH-mediated system in M. tuberculosis was shown to be  vital for prolonged life in the host, linking it to the phenotypic heterogeneity. This bug has one RSH protein, Rel, which is capable of both producing and degrading ppGpp. The logical question is then - how is Rel activity and abundance regulated, and what is its distribution in the bacterial population?

In the recent two papers (1 and 2) exactly that was done using M. smegmatis as a non-pathogenic model for M. tuberculosis. Rel promotor was fused with a GFP ORF so that its activity can be monitored on the single cell level using flow cytometry. And indeed, they saw that cells are very strongly heterogeneous in respect to the GFP level, which reflects activity of the rel promoter. This heterogeneity turned out to be brought about by the positive feedback loop feeding Rel expression back to its promoter via activity of the MprAB and SigE proteins.

Positive feed back loops are a common way of creating this sort of bistability in the biological systems and is a very common motif in the gene networks. What would be interesting is to see how common is this positive feedback among different bacteria, especially the ones that have different RSH system, such as the most commonly studied E. coli, which has two RSH proteins instead of one in M. tuberculosis: RelA, which is mostly producing ppGpp, and SpoT, which is mostly degrading it.

Effects of noise in protein expression are also different in E. coli: here synthetic and hydrolytic components are split into two independently produced proteins, thus resulting in a system with more degrees of freedom. Sure enough, it is well documented that intrinsic noise in protein expression (that would be roughly variation in the protein expression between the two independent genes in one bacterial cell) is much less than extrinsic one (roughly - variation between different cells), but still - two genes is more than one!

And what about bugs having more RSHs? Streptococcus mutans has three! From classic microbiological studies it seems that similar regulation could be present there as well, linking stress tolerance and genetic competence.

However fruitful, systems biology investigations of the stringent response are suffering from one big limitation: using current approaches, only a few readouts can be followed in the same cell at the same time. Stringent response, however, involves many different systems being rewired, and microarrays in the combination with metabolomics can follow many, many things happening in the cell (though lacking in the single cell resolution!).

References:

Potrykus K, & Cashel M (2008). (p)ppGpp: still magical? Annual review of microbiology, 62, 35-51 PMID: 18454629

Seaton K, Ahn SJ, Sagstetter AM, & Burne RA (2011). A transcriptional regulator and ABC transporters link stress tolerance, (p)ppGpp, and genetic competence in Streptococcus mutans. Journal of bacteriology, 193 (4), 862-74 PMID: 21148727

Mittenhuber G (2001). Comparative genomics and evolution of genes encoding bacterial (p)ppGpp synthetases/hydrolases (the Rel, RelA and SpoT proteins). Journal of molecular microbiology and biotechnology, 3 (4), 585-600 PMID: 11545276

Lemos JA, Lin VK, Nascimento MM, Abranches J, & Burne RA (2007). Three gene products govern (p)ppGpp production by Streptococcus mutans. Molecular microbiology, 65 (6), 1568-81 PMID: 17714452

Dahl JL, Kraus CN, Boshoff HI, Doan B, Foley K, Avarbock D, Kaplan G, Mizrahi V, Rubin H, & Barry CE 3rd (2003). The role of RelMtb-mediated adaptation to stationary phase in long-term persistence of Mycobacterium tuberculosis in mice. Proceedings of the National Academy of Sciences of the United States of America, 100 (17), 10026-31 PMID: 12897239

Sureka K, Ghosh B, Dasgupta A, Basu J, Kundu M, & Bose I (2008). Positive feedback and noise activate the stringent response regulator rel in mycobacteria. PloS one, 3 (3) PMID: 18335046

Ghosh S, Sureka K, Ghosh B, Bose I, Basu J, & Kundu M (2011). Phenotypic heterogeneity in mycobacterial stringent response. BMC systems biology, 5 PMID: 21272295

Elowitz MB, Levine AJ, Siggia ED, & Swain PS (2002). Stochastic gene expression in a single cell. Science (New York, N.Y.), 297 (5584), 1183-6 PMID: 12183631

Larson DR, Singer RH, & Zenklusen D (2009). A single molecule view of gene expression. Trends in cell biology, 19 (11), 630-7 PMID: 19819144

Alon U (2007). Network motifs: theory and experimental approaches. Nature reviews. Genetics, 8 (6), 450-61 PMID: 17510665

Ingolia NT, & Murray AW (2007). Positive-feedback loops as a flexible biological module. Current biology : CB, 17 (8), 668-77 PMID: 17398098

Raj A, & van Oudenaarden A (2009). Single-molecule approaches to stochastic gene expression. Annual review of biophysics, 38, 255-70 PMID: 19416069

Lidstrom ME, & Konopka MC (2010). The role of physiological heterogeneity in microbial population behavior. Nature chemical biology, 6 (10), 705-12 PMID: 20852608

Taniguchi Y, Choi PJ, Li GW, Chen H, Babu M, Hearn J, Emili A, & Xie XS (2010). Quantifying E. coli proteome and transcriptome with single-molecule sensitivity in single cells. Science (New York, N.Y.), 329 (5991), 533-8 PMID: 20671182

Balaban NQ, Merrin J, Chait R, Kowalik L, & Leibler S (2004). Bacterial persistence as a phenotypic switch. Science (New York, N.Y.), 305 (5690), 1622-5 PMID: 15308767

Mendeley group on stringent response

Kinetics of stringent response revealed by microarray analysis

Stringent response is immensely complex and many different proteins are targeted by ppGpp. Different targets have different susceptibility to ppGpp - K_D of binding or K_i of inhibition. Therefore as stringent response progresses and ppGpp concentration changes, different targets will be engaged sequentially, in this way orchestrating changes in the cellular physiology.

Studying the development of stringent response in time is a tricky business. Temporal resolution in one key component, ability following many parameters (many targets) is the other one. Microarrays seem to be the way (Traxler 2006, Traxler 2008, Durfee 2008, Traxler 2011 and Balsalobre 2011).

Here are the main points from Traxler 2011. They did microarray analysis of four strains: wt E. coli K-12-derivative MG1655, ppGpp⁰, Lrp- and RpoS- in MG1655 background. The nature of the wt is important - L12 derivatives are special in that sense that they exhibit stringent response upon L-Valine and L-isolucine overdose, and the latter is used to induce the stringent response in Traxler 2011.

In these strains they follow the effects of two global regulators, Transcription factor Lrp (leucine responsive protein) and RpoS AKA Sigma 38, a specialized σ-factor.

It turnes out that ppGpp regulates both the feed-back loop regulated by Lrp and the feed-forward loop regulated by RpoS (for discussion of how these motifs work see Alon 2007). Moreover, Lrp regulon is induced much earlier than RpoS.

So how is one molecule - ppGpp - evokes two temporally separated responses (early Lrp and late RpoS)? The obvious idea is that these two regulons have different sensitivity to ppGpp, thus the more sensitive one is induced earlier (Lrp) and the less sensitive one is induced later (RpoS). Combining microarray data with measurements of ppGpp concentrations in vivo confirmed this hypothesis.

Rendering Lrp operon inactive in turn disrupts normal progression of the stringent response: both wt and Lpr- strains rapidly accumulate ppGpp, but during the later stage wt cells lower the ppGpp concentration and resume growth, whereas Lrp- strain retains high ppGpp concentration is unable to adapt to stringent conditions. Functional RpoS (which works on the later stages of stringent response) does not help - you have to come prepared, i.e. Lrp regulon should do its job.

However powerful, microarray-based techniques inherently average out the cell-to-cell variability which plays important role in the stringent response as shown by the systems biology investigations.

References:

Matthew F. Traxler, Vineetha M. Zacharia, Stafford Marquard, Sean M. Summers, Huyen-Tran Nguyen, S. Elizabeth Stark and Tyrrell Conway. Discretely calibrated regulatory loops controlled by ppGpp partition gene induction across the ‘feast to famine’ gradient in Escherichia coli. Molecular Microbiology (2011) doi:10.1111/j.1365-2958.2010.07498.x

Matthew F. Traxler, Sean M. Summers, Huyen-Tran Nguyen, Vineetha M. Zacharia, G. Aaron Hightower, Joel T. Smith and Tyrrell Conway. The global, ppGpp-mediated stringent response to amino acid starvation in Escherichia coli. Molecular Microbiology (2008) 68(5), 1128–1148

Carlos Balsalobre. Concentration matters!! ppGpp, from a whispering to astrident alarmone. Molecular Microbiology (2011) doi:10.1111/j.1365-2958.2010.07521.x

Matthew F Traxler, Dong-Eun Chang, Tyrrell Conway. Guanosine 3',5'-bispyrophosphate coordinates global gene expression during glucose-lactose diauxie in Escherichia coli. PNAS 2006 vol. 103 (7) pp. 2374-9

Tim Durfee, Anne-Marie Hansen, Huijun Zhi, Frederick R. Blattner and Ding Jun Jin. Transcription Profiling of the Stringent Response in Escherichia coli. J. of Bacteriol. 2008, p. 1084-1096, Vol. 190, No. 3

R. I. Leavitt, H. E. Umbarger. Isoleucine and valine metabolism in Escherichia coli. XI. Valine inhibition of the growth of Escherichia coli strain K-12. J. Bacteriol. 1962 vol. 83 pp. 624-30

Uri Alon. Network motifs: theory and experimental approaches. Nat Rev Genet. 2007 vol. 8 (6) pp. 450-61

Brinkman AB, Ettema TJ, de Vos WM, van der Oost J. The Lrp family of transcriptional regulators. Mol Microbiol. 2003 48(2) pp. 287-94

Mendeley group on stringent response