Current Opinion in Microbiology Volume 12, Issue 5, October 2009, Pages 531-535 Antimicrobials Genomics

doi:10.1016/j.mib.2009.08.006 | How to Cite or Link Using DOI

Permissions & Reprints

Direct methods for studying transcription regulatory proteins and RNA polymerase in bacteria
David C Grainger1, David J Lee2, Stephen JW Busby2,
1 2

Department of Biological Sciences, University of Warwick, Coventry CV4 7AL, UK School of Biosciences, University of Birmingham, Birmingham B15 2TT, UK

Available online 15 September 2009. Transcription factors and sigma factors play a major role in bacterial gene regulation by guiding the distribution of RNA polymerase between the promoters of different transcription units in response to changes in the environment. For 40 years Escherichia coli K-12 has been the paradigm for investigating this regulation and most studies have focused on small numbers of promoters studied by a combination of genetics and biochemistry. Since the first complete sequence for a bacterialgenome was reported, the emphasis has switched to studying transcription on a global scale, with transcriptomics and bioinformatics becoming the methods of choice. Here we discuss two complementary direct experimental methods for studying transcription factors and sigma factors and we outline their potential use in rapidly establishing the regulatory networks in newly sequencedbacteria.

Article Outline

Introduction Chromatin immunoprecipitation Applications of ChIP to study transcription factors Applications of ChIP to study RNA polymerase DNA sampling: a complementary method Conclusions and perspectives

References and recommended reading Acknowledgements References

Introduction
Gene expression in all bacteria is tightly controlled, with transcription initiation being the principal point of regulation for many genes. Since the beginning of molecular biology, Escherichia coli K12 has been the organism of choice for the study of transcriptional regulation and it is clear that this regulation is due to a complex network of transcription factors and sigma factors that control the expression of 1800 transcription units in response to changes in the environment [ [1] and [2] ]. TheE. coli genome encodes over 250 gene regulatory proteins that range from highly specifictranscription factors such as the lactose operon repressor (Lac repressor), which controls a single transcription unit, through to global regulatory proteins, such as the cyclic AMP receptor protein, which controls scores of transcription units. In addition, the nucleoidassociated proteins, which are needed for maintaining chromosome folding and compaction, play important roles in transcriptional regulation. Many of these are present in large quantities that vary according to growth conditions, and they play key roles in upregulating or downregulating specific promoters [ [3] and [4] ]. Over 50 years, the network of E. coli gene regulatory proteins has been established by integrating information from studies on individual promoters and transcription factors. Before the arrival of whole genome sequences, most investigators would begin with their favourite promoter or factor and exploit a toolbox of genetic tricks to select and characterise mutants in which the activity or regulation of the promoter or factor was altered. The advent of cloning gave access to a battery of biochemical approaches for studying protein binding at specific promoters, and bioinformatic approaches, based on establishing and exploiting consensus sequences, were widely applied. Up to the arrival of large-scale shotgun sequencing, only a small number of gene regulatory regions were studied in depth. However, whole genome sequences, in combination with transcriptomics and bioinformatics, opened the way to pangenome viewing of transcriptional regulation. It was quickly established that sometranscription factors regulated scores or more of transcription units. This led to the ideas of transcriptional regulatory networks and transcription factor hierarchies, and it is from these studies, taken together with the data from years of laborious effort, that we have now such a comprehensive view of transcription in E. coli K-12 [ [1] and [3] ]. This, of course, begs the question of whether we can now find more rapid easier routes to establish transcription networks in other bacteria. We want to argue that the rate limiting step in the post-genomic era has been the dearth of direct methods to detect what is happening at any gene regulatory region. Thus, any

transcriptomics or proteomics experiment depends on measuring the consequences of the actions of gene regulatory proteins (that is, production of RNA or protein) rather than their interactions directly, and disentangling direct from indirect effects is not trivial. Furthermore, following up such experiments, or following up bioinformatic predictions, using genetic or biochemical methods is time consuming and may not be possible for some bacteria. Here we present chromatin immunoprecipitation (ChIP) as a method of choice for the rapid analysis of binding targets for gene regulatory proteins in any bacterium for which the genome sequence is known. We also discuss a newly developed complementary method, DNA sampling, which permits an audit of the regulatory proteins interacting at any locus.

Chromatin immunoprecipitation
ChIP represents a powerful tool, since it identifies proteinDNA interactions in vivo directly, independent of the biological consequences of binding (different technical aspects are reviewed in [[5] , [6] and [7] ]). Briefly, bacterial cells are exposed to formaldehyde, thereby instantly crosslinkingDNA binding proteins to the chromosome. After cell lysis and shearing of chromosomal DNA by sonication, the protein of interest is immunoprecipitated with specific antibodies, together with crosslinked DNA fragments. After reversal of the crosslinks and purification, the immunoprecipitated DNA is analysed in order to detect enrichment of the sequences bound by the protein of interest. Using ChIP in conjunction with DNA microarray analysis (ChIP-on-chip) permits DNA binding to be measured on a chromosome-wide scale (Figure 1).

High-quality image (294K) Figure 1.

Schematic outline of ChIP-on-chip analysis. The coloured dots represent different proteins bound to DNA in a bacterium. ChIP and ChIP-on-chip were first developed for eukaryotic cells and have since found applications in organisms as diverse as bacteriophage, yeast and mammals, where the binding of transcription factors, as well as other DNA-associated proteins, was investigated. For bacteria, ChIP-on-chip was first used with Caulobacter crescentus [8] but has been most applied to E. coli [9]. In principle, it can be used in any bacterium, such as Bacillus subtilis [ [10] and [11] ], Salmonella enterica [12],Helicobacter pylori [13], Mycobacterium bovis BCG [14] and Mycobacterium tuberculosis [15]. A crucial point is that protein binding locations can be identified without the need for a lot of prior knowledge of the bacterium under study. Indeed, because it is now possible to identify immunoprecipated DNA by sequencing (ChIP-seq), in principle, it is possible to derive information from bacteria whose chromosome sequence has not been determined.

Applications of ChIP to study transcription factors

The most straightforward use of ChIP in bacterial systems is in the location of transcription factors. Following the global analysis of the C. crescentus CtrA regulon [8] and the B. subtillis Spo0A and CodY regulators [ [10] and [11] ], ChIP-on-chip has now been applied to many other bacterial systems, including several pathogens. For instance, the H. pylori Fur protein has been studied and found to bind at about 200 genomic loci in an iron-dependent manner, supporting the idea that this protein acts as a pleiotropic regulator [13]. The previously uncharacterised BlaI transcription factor from M.tuberculosis was recently shown to regulate five DNA loci including the blaI gene itself, and others involved in resistance to -lactam antibiotics [15]. Unexpectedly, BlaI was found to bind upstream of the operon encoding ATP synthase, suggesting links between cell wall damage and ATP production. In the case of the Salmonella enterica PhoP transcription regulator, ChIP experiments were used to understand the hierarchy with which different promoters are served as concentrations of the trigger ligand, magnesium ions, change [12]. Studies performed in E. coli K-12 represent the paradigm for ChIP-on-chip studies of transcription factors. The distribution of ArcA [16], CRP [17], FNR [18], LexA [19], Lrp [20], MelR [21], NsrR [ [22] and[23] ] and RutR [24] has been determined. These studies show that some factors recognise single binding sites (as in the case of MelR) whilst others have more complex distributions (for example, CRP and LexA). There have also been some surprises. Thus, CRP binds throughout the E. colichromosome to nearly 1000 sites and, at many targets, appears to have no effect on transcription [17]. With RutR, most of its 20 binding sites mapped within coding regions, suggesting that it may play some other, as yet undiscovered, role [24]. It

is plausible that the binding of transcription factorsto specific sites without any function is a byproduct of evolution, perhaps due to inadequate purging after horizontal gene transfer. For any bacterial chromosome, we can now aspire to obtaining a complete description of the location of all the different transcription factors. This is likely to be realised for E. coli K-12 in the next few years [25] and this will provide the essential base for constructing regulatory networks and understanding the extent of different regulons. An interesting sequel comes from the study of nucleoid-associated proteins, and data sets for IHF, Fis, H-NS and StpA have been published [ [26] , [27] , [28] and [29] ]. The striking result from these studies is that these proteins, which are sometimes referred to as bacterialhistones, are located preferentially at intergenic regulatory regions. As well as underscoring their pervasive role in transcriptional regulation, this raises the possibility that bacterial chromosomes are organised with gene regulatory regions as the foci. Parallel studies of the distribution of H-NS inSalmonella have pointed to its role in silencing expression from blocks of recently acquired AT-rich sequences [ [30] and [31] ].

Applications of ChIP to study RNA polymerase

An interesting and powerful application of ChIP is to study the distribution of RNA polymerase across an entire bacterial chromosome [17]. This can be exploited to identify different transcription units, and is especially informative in combination with transcriptome data [ [26] and [32] ]. Additionally, it can be extended to study sigma factor distribution [33] and changes in the composition of elongation complexes as RNA polymerase moves away from regulatory regions and traverses operons [ [32] and[34] ]. Thus, Mooney et al. [35] recently reported the distribution of E. coli RNA polymerase, sigma70, NusA, NusG and Rho factor across different transcription units, showing close association of the sigma factor with RNA polymerase at promoter regions, whereas Rho factor, NusA and NusG subsequently join the enzyme during transcript elongation. This illustrates how the intricate dynamics of formation and remodelling of macromolecular complexes can be followed. In another study [36], ChIP was exploited to show that the bacteriophage Q antiterminator protein, which is specific for the late transcript that initiates at the PR promoter, remains associated with elongation complexes during transcription of at least 22 000 bases of the late operon. One of the key applications of ChIP-on-chip directed to RNA polymerase is to measure its redistribution in response to environmental change. For example, the response of E. coli to rifampicinwas described in terms of re-localisation of RNA polymerase to promoter sequences [ [17] and [37] ]. A similar redistribution was observed as E. coli enters stationary phase [18]. Thus, this type of experiment, which measures changes in the distribution of RNA polymerase (RNA polymerase-omics), may eventually supplant RNA-based strategies (transcriptomics) for the study of global transcriptional responses.

DNA sampling: a complementary method

In ChIP experiments, the aim is to identify all of the binding locations on a chromosome for one specified protein. DNA sampling is a complementary protocol, aimed to identify all of the proteins that are bound at one specified chromosomal location [38]. Briefly, the DNA segment to be sampled is cloned into a low copy number plasmid, adjacent to an array of five binding targets for the Lac repressor, sandwiched between two recognition sequences for the yeast homing endonuclease,Sce-I. The plasmid is transformed into a bacterial host that carries a second plasmid encoding Sce-I under the control of an arabinose-inducible promoter. Induction of Sce-I expression triggers excision of a DNA fragment carrying the target DNA segment, together with the Lac repressor binding targets (Figure 2). Since Sce-I recognises an 18 base pair target sequence, excision is specific for the plasmid and the host chromosome is unaffected. In the sampling protocol, the fragment is immunopurified by exploiting a host that has been modified by recombineering so that the Lac repressor is fused to an epitope tag. By using such a host, it is possible to isolate the fragment, together with bound proteins, using magnetic beads to which antibodies directed to the epitope tag have been immobilised. Gel electrophoresis and mass spectrometry are then used to identify the different bound proteins (Figure 2). This methodology allows the comparison of proteins bound at a particular DNA location in cells grown in different conditions (e.g. the changes in bound RNA polymerase before and after induction of a promoter).

High-quality image (308K) Figure 2. Schematic outline of DNA sampling. The coloured dots represent different proteins bound to the DNA in a bacterium. The DNA target to be sampled is cloned into a low copy number plasmid.

Note that target sites for Sce-I endonuclease are located only in this plasmid. The yellow dot represents epitope-tagged Lac repressor, binding specifically to lac operator sequences in the plasmid. The principle behind DNA sampling is similar to that behind ChIP in that it exploits an affinity reagent to pull down a protein (Lac repressor) that is bound to a DNA target. In DNA sampling, target sites for the Lac repressor are engineered to be close to the DNA target that is to be sampled and the emphasis is on the analysis of co-purified proteins rather than analysis of the DNA target. To date, DNA sampling has been applied only to E. coli K-12, but in principle, it should be applicable in a variety of bacterial species.

Conclusions and perspectives

Historically, transcription regulation in bacteria has been studied by genetics, with deductions being made from intelligent thinking based on phenotypes, and experimental strategies primarily founded on direct measurements have been left to those interested in eukaryotes. The arrival of the bacterialgenome tsunami now prompts the adoption of previously shunned methods and here we have outlined two direct approaches, one (ChIP) that is well established, and the other (DNA sampling) that is in its infancy. ChIP-on-chip represents a powerful and accessible technique for bacteriologists, with many applications both in axenic cultures and in more complex environments, including natural hosts. The use of ChIP is likely to become more widespread as DNA microarrays are superceded by high throughput DNA sequencing platforms [39]. By analysing immunoprecipitated DNA by sequencing, complications arising from array hybridisation chemistry, non-specific probe-DNA interactions and interference from secondary structures can be avoided. The DNA sampling technique is more cumbersome than ChIP and certainly still has some technical limitations, such as the identification of gene regulatory proteins whose binding to the sampled target is weak or ligand-dependent. The current protocol for DNA sampling [38] does not employ a crosslinking step, in contrast to ChIP protocols, but it may be possible, in future, to exploit crosslinking strategies to circumvent these limitations. Additionally, with the development of more sensitive detection methods and the application of novel imaging technologies, DNA sampling may eventually be applicable on a whole genome scale.

References and recommended reading

Papers of particular interest, published within the period of review, have been highlighted as:

 of special interest

Acknowledgements
Work from the authors laboratory is supported by a Wellcome Trust Programme Grant to SJWB and a Wellcome Trust Research Career Development Fellowship to DCG.

References
1 P.D. Karp, I.M. Keseler, A. Shearer, M. Latendresse, M. Krummenacker, S.M. Paley, I. Paulsen, J. Collado-Vides, S. Gama-Castro and M. Peralta-Gil, et al. Multidimensional annotation of theEscherichia coli K-12 genome. Nucleic Acids Res, 35 (2007), pp. 7577 7590. 2 S.Y. Lee, Systems Biology and Biotechnology of Escherichia coli, Springer Science (2009). 3 E. Balleza, L.N. Lopez-Bojorquez, A. Martinez-Antonio, O. Resendis-Antonio, I. LazadaChavez, Y.I. Balderas-Marinez, S. Encarnacion and J. Collado-Vides, Regulation by transcription factors inbacteria: beyond description. FEMS Microbiol Rev, 33 (2009), pp. 133151. 4 A. Martinez-Antonio, S.C. Janga and D. Thieffry, Functional organisation of Escherichia colitranscriptional regulatory network. J Mol Biol, 381 (2008), pp. 238247. 5 D.C. Grainger and S.J. Busby, Methods for studying global patterns of DNA binding by transcription factors and RNA polymerase. Biochem Soc Trans, 36 (2008), pp. 754 757. 6 D.C. Grainger and S.J. Busby, Global regulators of transcription in Escherichia coli: mechanisms of action and methods for study. Adv Appl Microbiol, 65 (2008), pp. 93113. 7 C. Sala, D.C. Grainger and S.T. Cole, Dissecting regulatory networks in host-pathogen interaction using ChIP-on-chip technology. Cell Host Microbe, 5 (2009), pp. 430437. 8 M.T. Laub, S.L. Chen, L. Shapiro and H.H. McAdams, Genes directly controlled by CtrA, a master regulator of the Caulobacter cell cycle. Proc Natl Acad Sci U S A, 99 (2002), pp. 4632 4637. 9 J.T. Wade, K. Struhl, S.J. Busby and D.C. Grainger, Genomic analysis of protein-DNA interactions inbacteria: insights into transcription and chromosome organization. Mol Microbiol, 65 (2007), pp. 2126.

This minireview presents a succinct overview of current ChIP technology and highlights some of its more surprising applications. See also citation [7]. 10 V. Molle, M. Fujita, S.T. Jensen, P. Eichenberger, J.E. Gonzalez-Pastor, J.S. Liu and R. Losick, The Spo0A regulon of Bacillus subtilis. Mol Microbiol, 50 (2003), pp. 16831701. 11 V. Molle, Y. Nakaura, R.P. Shivers, H. Yamaguchi, R. Losick, Y. Fujita and A.L. Sonenshein, Additional targets of the Bacillus subtilis global regulator CodY identified by chromatinimmunoprecipitation and genome-wide transcript analysis. J Bacteriol, 185 (2003), pp. 19111922. 12 J.C. Perez, D. Shin, I. Zwir, T. Latifi, T.J. Hadley and E.A. Groisman, Evolution of a bacterial regulon controlling virulence and Mg2+ homeostasis. PLoS Genet, 5 (2009), p. e1000428. 13 A. Danielli, D. Roncarati, I. Delany, V. Chiarini, R. Rappuoli and V. Scarlato, In vivo dissection of theHelicobacter pylori Fur regulatory circuit by genome-wide location analysis. J Bacteriol, 188 (2006), pp. 46544662. 14 S. Rodrigue, J. Brodeur, P.E. Jacques, A.L. Gervais, R. Brzezinski and L. Gaudreau, Identification of mycobacterial sigma factor binding sites by chromatin immunoprecipitation assays. J Bacteriol, 189 (2007), pp. 15051513. 15 C. Sala, A. Haouz, F.A. Saul, I. Miras, I. Rosenkrands, P.M. Alzari and S.T. Cole, Genomewide regulon and crystal structure of BlaI (Rv1846c) from Mycobacterium tuberculosis. Mol Microbiol, 71 (2009), pp. 11021116. This paper reports the unusual combination of a high resolution structure and a genomic analysis for a previously uncharacterised Mycobacterium tuberculosis transcription factor and illustrates the power of applying ChIP-on-chip in a difficult organism. 16 B.K. Cho, E.M. Knight and B.. Palsson, Transcriptional regulation of the fad regulon genes ofEscherichia coli by ArcA. Microbiology, 152 (2006), pp. 22072219. 17 D.C. Grainger, D. Hurd, M. Harrison, J. Holdstock and S.J. Busby, Studies of the distribution of E. coli cAMP-receptor protein and RNA polymerase along the E. coli chromosome. Proc Natl Acad Sci U S A, 102 (2005), pp. 1769317698.

18 D.C. Grainger, H. Aiba, D. Hurd, D.F. Browning and S.J. Busby, Transcription factor distribution inEscherichia coli: studies with FNR protein. Nucleic Acids Res, 35 (2007), pp. 269278. 19 J.T. Wade, N.B. Reppas, G.M. Church and K. Struhl, Genomic analysis of LexA binding reveals the permissive nature of the Escherichia coli genome and identifies unconventional target sites. Genes Dev, 19 (2005), pp. 26192630. 20 B.K. Cho, C.L. Barrett, E.M. Knight, Y.S. Park and B.. Palsson, Genome-scale reconstruction of the Lrp regulatory network in Escherichia coli. Proc Natl Acad Sci U S A, 105 (2008), pp. 1946219467. 21 D.C. Grainger, T.W. Overton, N. Reppas, J.T. Wade, E. Tamai, J.L. Hobman, C. Constantinidou, K. Struhl, G. Church and S.J. Busby, Genomic studies with Escherichia coli MelR protein: applications ofchromatin immunoprecipitation and microarrays. J Bacteriol, 186 (2004), pp. 69386943. 22 S. Efromovich, D. Grainger, D. Bodenmiller and S. Spiro, Genome-wide identification of binding sites for the nitric oxide-sensitive transcriptional regulator NsrR. Methods Enzymol, 437 (2008), pp. 211233. An excellent source of in-depth protocols for ChIP-on-chip analysis and the mathematical treatment of ChIP-on-chip data. 23 L.D. Rankin, D.M. Bodenmiller, J.D. Partridge, S.F. Nishino, J.C. Spain and S. Spiro, Escherichia coli NsrR regulates a pathway for the oxidation of 3-nitrotyramine to 4hydroxy-3-nitrophenylacetate. J Bacteriol, 190 (2008), pp. 61706177. 24 T. Shimada, A. Ishihama, S.J. Busby and D.C. Grainger, The Escherichia coli RutR transcription factor binds at targets within genes as well as intergenic regions. Nucleic Acids Res, 36 (2008), pp. 39503955. 25 N.E. Lewis, B.K. Cho, E.M. Knight and B.. Palsson, Gene expression profiling and the use of genome-scale in silico models of Escherichia coli for analysis: providing context for content. J Bacteriol, 191 (2009), pp. 34373444. 26 T. Oshima, S. Ishikawa, K. Kurokawa, H. Aiba and N. Ogasawara, Escherichia coli histonelike protein H-NS preferentially binds to horizontally acquired DNA in association with RNA polymerase.DNA Res, 13 (2006), pp. 141153.

27 E. Uyar, K. Kurokawa, M. Yoshimura, S. Ishikawa, N. Ogasawara and T. Oshima, Differential binding profiles of StpA in wild-type and H-NS mutant cells: a comparative analysis of cooperative partners by chromatin immunoprecipitation-microarray analysis. J Bacteriol, 191 (2009), pp. 23882391. 28 B.K. Cho, E.M. Knight, C.L. Barrett and B.. Palsson, Genome-wide analysis of Fis binding inEscherichia coli indicates a causative role for A-/AT-tracts. Genome Res, 18 (2008), pp. 900 910. Fis is one of the most abundant nucleoid-associated proteins in rapidly growing E. coli. This study shows that Fis binds at hundreds of target sites, the majority of which are in gene regulatory regions. This has important consequences for the organisation of the folded chromosome and shows that the influence of Fis is more pervasive than previously thought. See also citation [29]. 29 D.C. Grainger, D. Hurd, M.D. Goldberg and S.J. Busby, Association of nucleoid proteins with coding and non-coding segments of the Escherichia coli genome. Nucleic Acids Res, 34 (2006), pp. 46424652. 30 S. Lucchini, G. Rowley, M.D. Goldberg, D. Hurd, M. Harrison and J.C. Hinton, H-NS mediates thesilencing of laterally acquired genes in bacteria. PLoS Pathog, 2 (2006), p. e81. 31 W.W. Navarre, S. Porwollik, Y. Wang, M. McClelland, H. Rosen, S.J. Libby and F.C. Fang, Selectivesilencing of foreign DNA with low GC content by the H-NS protein in Salmonella. Science, 313 (2006), pp. 236238. 32 N.B. Reppas, J.T. Wade, G.M. Church and K. Struhl, The transition between transcriptional initiation and elongation in E. coli is highly variable and often rate limiting. Mol Cell, 24 (2006), pp. 747757. 33 J.T. Wade, D.C. Roa, D.C. Grainger, D. Hurd, S.J. Busby, K. Struhl and E. Nudler, Extensive functional overlap between sigma factors in Escherichia coli. Nat Struct Mol Biol, 13 (2006), pp. 806814. 34 M. Raffaelle, E.T. Kanin, J. Vogt, R.R. Burgess and A.Z. Ansari, Holoenzyme switching and stochastic release of sigma factors from RNA polymerase in vivo. Mol Cell, 20 (2005), pp. 357 366.

35 R.A. Mooney, S.E. Davis, J.M. Peters, J.L. Rowland, A.Z. Ansari and R. Landick, Regulator trafficking on bacterial transcription units in vivo. Mol Cell, 33 (2009), pp. 97108. This remarkable paper presents a picture of the comings and goings of different factors throughout the bacterial transcription cycle. The work is a stunning technical feat illustrating the power of high-density arrays and rigorous data analysis. See also citation [32]. 36 P. Deighan and A. Hochschild, The bacteriophage Q anti-terminator protein regulates late gene expression as a stable component of the transcription elongation complex. Mol Microbiol, 63 (2007), pp. 911920. 37 C.D. Herring, M. Raffaelle, T.E. Allen, E.I. Kanin, R. Landick, A.Z. Ansari and B.. Palsson, Immobilization of Escherichia coli RNA polymerase and location of binding sites by use of chromatinimmunoprecipitation and microarrays. J Bacteriol, 187 (2005), pp. 61666174. 38 M. Butala, S.J. Busby and D.J. Lee, DNA sampling: a method for probing protein binding at specific loci on bacterial chromosomes. Nucleic Acids Res, 37 (2009), p. e37. Article | (493 K) | |View Record in Scopus | | Cited By in Scopus (32) 39 K.V. Voelkerding, S.A. Darnes and J.D. Durtschi, Next-generation sequencing: from basic research to diagnostics. Clin Chem, 55 (2009), pp. 641658. Article | | View Record in Scopus | | Cited By in Scopus (43) PDF (2256 K) | PDF