Agustino Martinez-Antonio, Sarath Chandra Janga, Heladia Salgado & Julio Collado-Vides

Summary

Individual cells need to have a constant monitoring of their environmental composition. They need to discern and synchronize transcriptional responses in according to variations in external and internal conditions. To achieve this level of coordination, metabolites and chemical compounds are used by the cell as messages to know the composition of these environments. These signals act as control elements for the activity of transcription factors (TF), which enable the cell to reciprocate with a specific transcriptional response to environmental changes. Here we outline how the E. coli TF repertoire is responding to external or internal conditions based on the cellular location of the signals affecting them; if the signal is outside the periplasm, then the TF affected is sensing an external condition. On the other hand, if the signal is produced or synthesized within the cell (inside the membrane), then the TF affected is sensing an internal condition. We classify 121 TFs according to their response to external, internal or both conditions. We find that the global transcriptional regulation is directed by the internal class of sensing since: i) The internal class controls most of the regulons, ii) TFs of external and hybrid classes are mostly regulated by TFs from the internal class, iii) The internal class has the highest frequency of co-regulation and iv) Most of the global regulators belong to the internal class of sensing. This simple classification of TFs will facilitate to study experimentally and theoretically how the cell partitions its regulatory network to sense different stimuli while coordinating its regulatory response to extracellular and intracellular changes.

Supplementary data accompanying the paper

1. Definitions used in the manuscript describing the classification of sensing systems are shown below in the supplementary figure 1. For detailed description of the classes, please look to figure 1 and the text accompanying the paper.

2. The classification of the regulatory network of E.coli in different systems of sensing as described in this paper

In these Tables, we show the systems involving TFs for sensing environment signals in Escherichia coli K-12. The file is divided into three parts(to obtain the file in flat file format click here); in the first part are those TFs sensing extracellular signals; E-TC and E-TM. In the second are those TFs for hybrid systems (H) whose signals are both,transported and synthesized by the cell. In the third part are those systems for sensing endogenous signals (I-SM) and those for DNA-bending (I-DB). Data obtained from RegulonDB³ , TransportDB⁴ and EcoCyC⁵ databases(References at the bottom).

*  External sensing

A) External-Two-component systems (E-TC), (29 TFs)

B). External sensing using transported metabolites ( E-TM ), (28 TFs)

*  Hybrid systems (H). Sensing external and internal signals (33 TFs)

*  TFs for internal sensing

*  Using internal synthesized signals (I-SM), (26 TFs)

  B) TFs for DNA-bending ( I-DB ), (4 TFs)

3. Supplementary Figure 2 shows the Escherichia coli transcriptional regulatory network for sensing the extracellular and intracellular environment.You can also view this figure in pdf format

In the upper part and from left to right, in green, are those TFs corresponding to the extracellular class of sensing; in light green, are those TFs from two-component systems (E-TC) and in dark green, are those TFs using exogenous metabolites transported into the cells by transport systems (E-TM). In light blue are those TFs corresponding to hybrid system of sensing (H); i.e those TFs using metabolites synthesized inside the cell and incorporated from the milieu, resembling mechanism from E-TM and I-SM sub-classes. For TFs for internal sensing; in dark blue are those TFs for DNA-bending or chromatin architectural TFs, they do not sense metabolites directly. In pink are those TFs for sensing intracellular conditions using endogenous signals synthesized inside the cell or sensing the internal cellular redox-state. Finally, in light orange are those TFs without metabolites or un known mechanisms to modulate their activities. Global TFs⁶ (ArcA, Lrp, Hns, IHF, FIS, FNR and CRP) are at the top level. The connections: green lines represent activation; red, repression; blue, dual (activation and repression). The loops in the TFs represent auto-regulation positive, negative or dual represented as by the line codes. In yellow (low) are those genes which do not code for TF products. Abbreviations: S, substrate; E, enzyme and P, product.

4. Figure 3 shows the overlap between regulated genes,TUs and Operons among the different classes of sensing-Internal,External and Hybrid.It can be seen that in each case more than 75% of the machinery is regulated by the internal sensing class. The numbers indicate the absolute number of elements in each class being regulated by each class of TFs (in parenthesis are the percentages of each class of elements). Abbreviations: I, internal class of sensing; E, external class of sensing; H, hybrid systems as described in Box 1.

5. Interaction of signal metabolites with Transcription Factors

We analyzed the TFs by the number of signal effectors in order to know the flexibility of the transcriptional machinery to adapt to environmental changes. Currently, there are 139 known TFs regulating 1058 genes with some experimental evidence in E. coli. of the Total of TFs, 120 (86%) could be classified into one of the three classes of sensing proposed in the paper and presented in the tables before. Of the classifiable TFs, in 77 TFs there is a one-to-one relation between TF and the metabolite (Figure 4a). When a TF responds to metabolite (13 TFs), these metabolites have structural similarity and the TF has a different transcriptional response depending on the metabolite used (Figure 4b). For instance, TyrR can interact with three aromatic amino acids and ATP, and it is able to acquire more than one homomultimeric conformation; it can bind to strong sites by itself, or with tyrosine bound, binds to weak sites for repression, while phenylalanine and tryptophan are used mostly to activate transcription by binding only to strong sites in the DNA⁷. On the other hand, some metabolites regulate the activity of more than one TF (15 TFs, Figure 4c). In such cases there could be a differential binding to the TFs depending on the availability of the metabolite; as is known for galactose, where GalS is active in high concentrations of galactose and GalR (higher affinity) is active when the availability of galactose is low in the milieu⁸.

Other 5 TFs do not respond precisely to one metabolite, such as FNR, SoxR, OxyR (modifying their conformation depending on the redox potential in the cell) and Ada and LexA (sensing the damage on DNA strands). There are 4 TFs; IHF, FIS, HU and HNS which do not respond directly to signal effectors (but through their regulation by TFs, those responding to signals). Finally, of 7 TFs forming part of two component systems (external class) we do not have knowledge about their sensed signals.

References

1. Wall, M.E. , et al. (2004) Design of gene circuits: lessons from bacteria. Nat Rev Genet 5, 34-42

2. Ulrich, L.E. , et al. (2005) One-component systems dominate signal transduction in prokaryotes. Trends Microbiol 13, 52-56

3. Salgado, H. , et al. (2004) RegulonDB (version 4.0): transcriptional regulation, operon organization and growth conditions in Escherichia coli K-12. Nucleic Acids Res 32 Database issue, D303-306

4. Ren, Q. , et al. (2004) TransportDB: a relational database of cellular membrane transport systems. Nucleic Acids Res 32 Database issue, D284-288

5. Keseler, I.M. , et al. (2005) EcoCyc: a comprehensive database resource for Escherichia coli. Nucleic Acids Res 33 Database Issue, D334-337

6. Martinez-Antonio, A., and Collado-Vides, J. (2003) Identifying global regulators in transcriptional regulatory networks in bacteria. Curr Opin Microbiol 6, 482-489

7. Pittard, J. , et al. (2005) The TyrR regulon. Mol Microbiol 55, 16-26

8. Geanacopoulos, M., and Adhya, S. (1997) Functional characterization of roles of GalR and GalS as regulators of the gal regulon. J Bacteriol 179, 228-234

For Questions/Comments/Data requests, please mail: [email protected] /[email protected] / [email protected]