Molecular Microbiology (2014) 92(4), 641–647 ■
doi:10.1111/mmi.12604 First published online 17 April 2014
MicroCommentary Lies and deception in bacterial gene regulation: the roles of nucleic acid decoys Yvonne Göpel and Boris Görke* Max F. Perutz Laboratories, Department of Microbiology, Immunobiology and Genetics, Center of Molecular Biology, University of Vienna, 1030 Vienna, Austria.
Summary Bacteria use intricately interconnected mechanisms acting at the transcriptional and post-transcriptional level to adjust gene expression to their needs. An intriguing example found in the chitosugar utilization systems of Escherichia coli and Salmonella is uncovered in a study by Plumbridge and colleagues. Three transcription factors (TFs), a small regulatory RNA (sRNA) and a sRNA trap cooperate to set thresholds and dynamics in regulation of chitosugar utilization. Specifically, under inducing conditions a decoy site on the polycistronic chitobiose (chbBCARFG) mRNA sequesters sRNA ChiX, which represses synthesis of the separately encoded chitoporin ChiP. Base-pairing of ChiX with its decoy has no role for the chb genes themselves when the mRNA is in excess. In the absence of substrate, however, this base-pairing tightly represses chbC encoding a subunit of the chitosugar transporter. Thus, one and the same sRNA/ mRNA interaction serves different regulatory functions under different environmental conditions. The employment of RNA decoys to control the activities of post-transcriptional regulators themselves is an increasingly recognized mechanism in gene regulation. Another observation in the current study highlights the possibility that decoy sites might even exist on the DNA controlling the availability of TFs for their target promoters.
Accepted 3 April, 2014. *For correspondence. E-mail boris.goerke@ univie.ac.at; Tel. (+43) 1 4277 54603; Fax (+43) 1 4277 9146.
© 2014 John Wiley & Sons Ltd
Intricate teamwork of three transcriptional regulators, a small RNA and a conditional small RNA decoy for sophisticated control of chitosugar utilization Access to carbon sources is a prerequisite for survival of any living organism. When present in a mixture, bacteria often select between compounds and preferentially use the carbon source that is most easily accessible and allows for fastest growth. Accordingly, carbohydrate metabolic pathways are extensively regulated at all levels (Görke and Stülke, 2008; Bobrovskyy and Vanderpool, 2013). A current study reveals a surprisingly complex regulatory circuit, in which three transcription factors and a small RNA (sRNA) intricately cooperate to regulate utilization of the chitin degradation products chitobiose and chitotriose (Plumbridge et al., 2014). This mechanism further involves an mRNA that, depending on its concentration, functions either as target or trap for the sRNA, representing a novel principle in bacterial gene regulation. Uptake of chitosugars in E. coli and Salmonella involves transport by chitoporin ChiP across the outer membrane followed by translocation into the cytoplasm via the tri-partite PTS-transporter ChbBCA (Keyhani et al., 2000; Figueroa-Bossi et al., 2009). Transcription initiation at the chbBCARFG operon is controlled by a triumvirate of TFs (Plumbridge and Pellegrini, 2004). In the absence of chitosugars, transcription is dually repressed by NagC, which is a global regulator of amino-sugar metabolism, and the operon-specific TF ChbR. In the presence of chitosugars, NagC is released from its operators and ChbR switches its role and stimulates transcription initiation in concert with the cAMP receptor protein. As a peculiarity, the inducers for ChbR and NagC are generated consecutively by catabolism of chitosugars via enzymes ChbG and ChbF respectively (Verma and Mahadevan, 2012). Expression of the separately encoded chiP gene is regulated by sRNA ChiX (Figueroa-Bossi et al., 2009; Rasmussen et al., 2009). ChiX inhibits translation of chiP by base-pairing and causes premature transcription termination by uncoupling translation from transcription (Bossi et al., 2012). In contrast to many other base-pairing
642 Y. Göpel and B. Görke ■
sRNAs, ChiX is constitutively expressed and not codegraded with chiP. Previous elegant studies revealed that activity of ChiX is controlled by an unorthodox mechanism involving an RNA trap. The chb mRNA also binds ChiX through a site in the chbB–chbC intercistronic
region, which dramatically accelerates decay of the sRNA by RNase E. Hence, transcription of chb sequesters and destabilizes ChiX, leading to de-repression of the chiPQ mRNA (Fig. 1A; Figueroa-Bossi et al., 2009; Overgaard et al., 2009). © 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
Roles of nucleic acid decoy sites 643
Fig. 1. Mechanisms employed by nucleic acid decoys in gene regulation. A. Trapping small RNA regulators by RNA decoys. In the absence of the decoy RNA (red) the sRNA regulator (blue) is highly abundant leading to repression of the target RNA (black). Following its induction the decoy RNA accumulates and sequesters the regulatory RNA, which has little effect on the decoy RNA itself due to its vast excess over the sRNA. As a result, the target RNA escapes repression by the sRNA regulator. Notably, the sRNA may also control expression of the decoy RNA itself as exemplified by repression of chbC by sRNA ChiX in the absence of chitosugars (Plumbridge et al., 2014). B. Trapping RNA-binding proteins by RNA decoys. RNA-binding proteins govern stability and/or translational efficiency of specific transcripts through their binding. Accumulation of decoy RNAs in response to a specific stimulus sequesters the protein from the target RNA. Accordingly, repression of the target RNA is relieved. C. Decoy binding sites on the DNA attune gene expression responses to the copy number of the TF (Brewster et al., 2014). The absence of decoy sites generates a linear relation (gray curve) between the number of repressor molecules and the fold-change in gene expression triggered by a promoter (black arrow) with a single operator site (red box). In the presence of decoy binding sites, the response requires a higher number of repressor molecules due to the buffering effect of the decoys. Regulation occurs more gradual, when the affinity of the decoy sites is lower (blue curve) or equal (green curve) as compared with the regulatory binding site. In contrast, a sharp transition between both regimes is observed (red curve), when the decoy sites possess a higher affinity for the TF. The diagram has been adapted from (Brewster et al., 2014).
The current study amends our understanding of this regulatory circuit by adding two novelties (Plumbridge et al., 2014). First, it is shown that transcription initiation at chiPQ is controlled by NagC and not by ChbR as previously speculated. The requirement of ChbR for chiPQ expression is only indirect as it activates chb transcription, which generates the decoy for ChiX and the enzymes producing the inducer for NagC. Second, the effect of ChiX on expression of the chb genes themselves has now been clarified. As expected for a decoy, ChiX has no impact on chb expression under inducing conditions, when chb mRNA levels are likely in excess. In the absence of chitosugars however, ChiX tightly represses translation of chbC setting the basal level of the chitobiose transporter. Thus, depending on the stoichiometry of the partners, one and the same sRNA/mRNA basepairing generates different outputs. When highly transcribed, the chb mRNA serves to trap and destabilize ChiX, but at low transcription rates it becomes a ‘bona fide’ regulatory target of ChiX. Thus, the hunter and the hunted switch their roles (Plumbridge et al., 2014). Taking the novel findings into account, an integral model of regulation of chitosugar utilization is described, which likely includes a multi-tier induction process (Plumbridge et al., 2014). When chitosugars become available, the Chb proteins present at a basal level may catalyse formation of the inducer for ChbR, causing initial increase in chb transcription. However, a corresponding increase of the chitobiose transporter might only occur when chb transcription rates are sufficient to overcome repression by ChiX. Thus, ChiX likely sets the chitobiose threshold concentration and the delay time required for activation of the system. Persistently high chb expression levels lead to destabilization of ChiX and consequently to a first level of chiPQ expression. Continuous synthesis of ChiP and the Chb proteins will boost accumulation of the inducer for NagC. Relief of both operons from repression by NagC may finally amplify chitosugar utilization to maximum levels. © 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
Tight control of chitosugar utilization might prevent induction of the corresponding genes when only low amounts of chitosugars are available. Thus, many Enterobacteriaceae, in which the above described regulatory circuit is likely to be conserved, are perhaps only opportunistic utilizers of chitosugars when these compounds are available in large quantities as sporadic part of the hosts’ diet. This idea is reinforced by the observation that E. coli and Salmonella rely on excreted chitinases of other bacteria for degradation of chitin to the chitosugars that can be utilized. In line, both operons, chiPQ and chb, are subject to strong catabolite repression, i.e. presence of preferred sugars efficiently switches off chitosugar utilization (Plumbridge and Pellegrini, 2004; Plumbridge et al., 2014). In contrast, the role of chitin as carbon source might be more important for aquatic bacteria such as Vibrio species, reflecting that a major part of chitin is produced in aquatic environments. For Vibrio chitin has roles even beyond nutrition, by triggering complex developmental responses such as colonization of host cell tissues (Kremer et al., 2013) and genetic competence as highlighted by another study published earlier this year in Molecular Microbiology (Yamamoto et al., 2014).
RNA decoys: dawn of a novel paradigm in post-transcriptional regulation The employment of decoy RNAs for regulation of distinct RNAs either via sequestration of a shared RNA regulator (Fig. 1A) or by titration of an RNA-binding protein (Fig. 1B) is a hot topic, shifting current paradigms in the understanding of gene regulation. A plethora of cases reminiscent of the role of the chb decoy for regulation of (and by) sRNA ChiX was also reported for eukaryotic microRNAs (miRNA). Transcripts can sequester miRNAs through decoy sites termed microRNA recognition elements (MREs), thereby inhibiting the miRNA response at the true targets (Banks et al., 2012; de Giorgio et al., 2013; Kartha
644 Y. Göpel and B. Görke ■
and Subramanian, 2014). The first of such RNA decoys, with a role in the phosphate starvation response, was discovered in Arabidopsis, where the long non-coding RNA (lncRNA) IPS1 was shown to sequester miR399 by base-pairing, resulting in upregulation of the primary target of miR399 (Franco-Zorrilla et al., 2007). Meanwhile various types of transcripts called competing endogenous RNAs (ceRNAs) are believed to act as miRNAs sponges. These ceRNAs may include pseudogenes, long noncoding RNAs, circular RNAs or even mRNAs. mRNAs may communicate extensively with each other via MREs for shared miRNA. This is exemplified by studies in mammalian cells demonstrating an important role of target mimicry for cancer initiation and progression. In this case, intricate regulatory networks of coding ceRNAs establish cross-talk between different oncogenic pathways by titrating shared miRNA regulators. Similar to the bacterial chb transcript, many of these ceRNAs can act as targets as well as regulators of the cognate miRNA (Sumazin et al., 2011; Kartha and Subramanian, 2014). The concept of RNA mimicry was initially inspired by the paradox that a multitude of conserved target sites for a given miRNA can be predicted by bioinformatics approaches. Yet, only a few of these potential interactions generated detectable phenotypes. Consequently, a decoy function was proposed for these ‘pseudotargets’ (Seitz, 2009). A similar scenario is encountered in bacteria. Even improved search algorithms generate long lists of candidate target sites for a given sRNA. However, only a few of these sites are actual ‘bona fide’ targets (Wright et al., 2013). The remaining hits are usually considered as false-positives, but are they? In fact they could act as sponges for sRNA sequestration, but this possibility has been neglected so far. A different regulatory principle making use of RNA mimicry employs RNAs competing for binding to a common protein (Fig. 1B). A paradigmatic system is provided by the bacterial carbon storage regulatory Csr system. Protein CsrA controls translation and/or stability of its target RNAs by binding to GGA motifs. The activity of CsrA can be counteracted through its sequestration by two cognate sRNAs CsrB and CsrC or even by other mRNAs carrying binding sites for CsrA, recapitulating the hypothesis of ceRNAs (Romeo et al., 2013; Sterzenbach et al., 2013). A comparable mechanism has been reported in human sarcoma cell lines where miRNA miR-29 prevents binding of protein HuR to its cognate target mRNA A20. Since HuR is required for mRNA decay, this mechanism protects A20 from degradation (Balkhi et al., 2013). Competition for a shared protein regulator can also occur between sRNAs themselves as exemplified by the GlmY/ GlmZ system in E. coli (Göpel et al., 2014). Normally, sRNA GlmZ is recruited to degradation by adaptor protein RapZ. However, under conditions that require integrity of
GlmZ, sRNA GlmY acts as decoy and counters this process by sequestering RapZ (Göpel et al., 2013).
Decoy sites on the DNA for sequestration of transcription factors? The principle to employ decoy binding sites for regulation may not only apply to RNA but perhaps even to DNA. This intriguing possibility is emphasized by another interesting observation described in the current study (Plumbridge et al., 2014). A sequence matching the ChbR consensus site is present upstream of the Salmonella chiP promoter and DNA footprinting demonstrates tight and specific interaction of ChbR with this site. Paradoxically, this interaction has no role for regulation of chiPQ expression. Other species including E. coli carry a GalR repressor binding site at the corresponding position, which is conserved to a higher degree as compared with surrounding sequences. Doubtlessly, this site is bound by GalR and its isorepressor GalS with high affinity (Plumbridge et al., 2014). A role of GalR or GalS for regulation of chiP expression remains elusive, once again recapitulating the findings for ChbR and chiP in Salmonella. However, these observations are not unprecedented. Global approaches based on chromatin immunoprecipitation (ChIP) allow mapping of TF binding sites on a global scale. When combined with transcriptome profiling this approach can identify all regulatory relevant binding sites of a TF. One surprising result of these studies is that many TFs occupy specific sites on the chromosome without generating an apparent regulatory output on colocalizing genes. This finding not only applies to globally acting bacterial TFs such as FNR, LexA or σ32, but even for TFs controlling smaller regulons such as RutR, PurR or AraC (Wade et al., 2007; Shimada et al., 2008; Cho et al., 2011; Stringer et al., 2014). Similar observations were reported from archaea and eukaryotes as well (MacQuarrie et al., 2011; Nguyen-Duc et al., 2013). The ‘non-regulatory’ binding sites of TFs can even reside within genes as exemplified by RutR (Shimada et al., 2008). The high number of ‘non-functional’ sites detected in bacteria is highly controversial: It was argued that they represent false-positive artefacts generated by inherent experimental bias (Waldminghaus and Skarstad, 2010). However, based on results obtained by different methods other researchers claim that decoys are genuine binding sites for TFs in vivo (Bonocora et al., 2013). For statistical reasons, sequences similar to the canonical recognition motif of any TF are expected to occur by chance in the chromosome. In fact, LexA was found to bind virtually all sequences in the genome resembling its consensus site (Wade et al., 2007). Are the ‘non-functional’ sites indeed irrelevant and only exist because their random appearance is difficult to avoid during evolution? One counter© 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
Roles of nucleic acid decoy sites 645
argument is that spurious binding sites can be eliminated by negative selection when causing an adverse fitness effect (Hahn et al., 2003). In addition, the non-cognate sites are often conserved across species (MacQuarrie et al., 2011), as also recapitulated by the GalR/GalS site identified in the current study (Plumbridge et al., 2014), arguing against their functional irrelevance. Yet, what is their function? Several hypotheses were proposed to explain noncanonical TF-DNA interactions. First, an involvement of these sites in transcriptional regulation could be masked by redundantly acting TFs or when an additional TF is required. Moreover, a TF could initiate transcription at so far unannotated sites directing expression of e.g. novel non-coding RNA genes. Recently, it was shown that small RNAs may also be expressed from within genes, either in sense or antisense direction (Dornenburg et al., 2010; Chao et al., 2012). In agreement, antisense transcription seems to be pervasive in E. coli and a recent study identified 316 antisense RNAs forming duplexes with their cognate mRNAs indicating that these transcripts are functional in regulation (Lybecker et al., 2014). Second, the non-canonical binding sites could serve biological functions beyond transcriptional control. Operator-bound GalR molecules were proposed to interact with each other causing intra-chromosomal connections. This three-dimensional structure also includes GalR bound to non-canonical sites, which are scattered along the chromosome (Qian et al., 2012). This may contribute to compaction of the chromosome or serve to increase the local concentration of GalR at its target promoters thereby co-ordinating regulation. Strikingly, one of the GalR target genes, galP, is additionally repressed by NagC (El Qaidi et al., 2009). Hence, one possibility is that the GalR binding site observed in the current study (Plumbridge et al., 2014) serves to increase the local concentration of NagC at the galP and chiP promoters via interaction of bound GalR. Similar roles for interaction of distant chromosomal loci or for reorganization of the chromatin structure were also assigned to non-regulatory TF binding sites in higher organisms (MacQuarrie et al., 2011). Apart from these specialized functions, non-canonical binding sites might indeed play an important role in gene regulation by generating complex dosage responses to their cognate TFs and may therefore be termed ‘decoys’ (Burger et al., 2010; Brewster et al., 2014; Rydenfelt et al., 2014). It is commonly assumed that the copy number of a TF exceeds the number of its binding sites in a cell. Consequently, the correlation between the foldchange in transcription and the number of repressors will be linear, when considering a simple repression architecture i.e. a promoter with a single repressor binding site (Fig. 1C; Brewster et al., 2014). That is, the system will behave as if decoy sites are absent. However, several © 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
TFs in E. coli are present in less than 10 copies per cell suggesting that their number is equal or lower than the number of respective binding sites (Taniguchi et al., 2010; Rydenfelt et al., 2014). Under this regime the competing decoy sites provide a buffer and out-compete the regulatory site as long as the TF is limiting. Thus, there will be no major response in fold-changes until all decoy sites are occupied. Once the repressor copy number exceeds the number of decoys, the repressor gets access to the regulatory site. This results in a sharp transition to a scenario which allows regulation of the target gene (Fig. 1C). The steepness of the dose–response curve during this transition depends on the strength of the regulatory binding site. When the affinity of the decoys is higher as compared with the regulatory site, more repressor molecules are required to switch between both regimes (Fig. 1C). Alterations in the dose–response relationship by decoy sites have also been observed for transcriptional activators (Lee and Maheshri, 2012). In sum, decoy sites tune the regulatory response to TF availability in a non-trivial but predictable manner. This provides the basis for the controlled manipulation of transcriptional regulatory responses by using artificial decoys, without any need for alteration of the genome (Brown et al., 2013).
Acknowledgements Research in our laboratory is supported by DFG grant GO1355/7-1 and by FWF grant P26681-B22 to B.G.
References Balkhi, M.Y., Iwenofu, O.H., Bakkar, N., Ladner, K.J., Chandler, D.S., Houghton, P.J., et al. (2013) miR-29 acts as a decoy in sarcomas to protect the tumor suppressor A20 mRNA from degradation by HuR. Sci Signal 6: ra63. Banks, I.R., Zhang, Y., Wiggins, B.E., Heck, G.R., and Ivashuta, S. (2012) RNA decoys: an emerging component of plant regulatory networks? Plant Signal Behav 7: 1188– 1193. Bobrovskyy, M., and Vanderpool, C.K. (2013) Regulation of Bacterial Metabolism by Small RNAs Using Diverse Mechanisms. Annu Rev Genet 47: 227–250. Bonocora, R.P., Fitzgerald, D.M., Stringer, A.M., and Wade, J.T. (2013) Non-canonical protein-DNA interactions identified by ChIP are not artifacts. BMC Genomics 14: 254. Bossi, L., Schwartz, A., Guillemardet, B., Boudvillain, M., and Figueroa-Bossi, N. (2012) A role for Rho-dependent polarity in gene regulation by a noncoding small RNA. Genes Dev 26: 1864–1873. Brewster, R.C., Weinert, F.M., Garcia, H.G., Song, D., Rydenfelt, M., and Phillips, R. (2014) The transcription factor titration effect dictates level of gene expression. Cell 156: 1312–1323. Brown, A.J., Mainwaring, D.O., Sweeney, B., and James, D.C. (2013) Block decoys: transcription-factor decoys designed for in vitro gene regulation studies. Anal Biochem 443: 205–210.
646 Y. Göpel and B. Görke ■
Burger, A., Walczak, A.M., and Wolynes, P.G. (2010) Abduction and asylum in the lives of transcription factors. Proc Natl Acad Sci USA 107: 4016–4021. Chao, Y., Papenfort, K., Reinhardt, R., Sharma, C.M., and Vogel, J. (2012) An atlas of Hfq-bound transcripts reveals 3’ UTRs as a genomic reservoir of regulatory small RNAs. EMBO J 31: 4005–4019. Cho, B.K., Federowicz, S.A., Embree, M., Park, Y.S., Kim, D., and Palsson, B.O. (2011) The PurR regulon in Escherichia coli K-12 MG1655. Nucleic Acids Res 39: 6456–6464. Dornenburg, J.E., Devita, A.M., Palumbo, M.J., and Wade, J.T. (2010) Widespread antisense transcription in Escherichia coli. MBio 1. pii: e00024-10. doi: 10.1128/mBio.0002410. El Qaidi, S., Allemand, F., Oberto, J., and Plumbridge, J. (2009) Repression of galP, the galactose transporter in Escherichia coli, requires the specific regulator of Nacetylglucosamine metabolism. Mol Microbiol 71: 146–157. Figueroa-Bossi, N., Valentini, M., Malleret, L., Fiorini, F., and Bossi, L. (2009) Caught at its own game: regulatory small RNA inactivated by an inducible transcript mimicking its target. Genes Dev 23: 2004–2015. Franco-Zorrilla, J.M., Valli, A., Todesco, M., Mateos, I., Puga, M.I., Rubio-Somoza, I., et al. (2007) Target mimicry provides a new mechanism for regulation of microRNA activity. Nat Genet 39: 1033–1037. de Giorgio, A., Krell, J., Harding, V., Stebbing, J., and Castellano, L. (2013) Emerging roles of competing endogenous RNAs in cancer: insights from the regulation of PTEN. Mol Cell Biol 33: 3976–3982. Göpel, Y., Papenfort, K., Reichenbach, B., Vogel, J., and Görke, B. (2013) Targeted decay of a regulatory small RNA by an adaptor protein for RNase E and counteraction by an anti-adaptor RNA. Genes Dev 27: 552–564. Göpel, Y., Khan, M.A., and Görke, B. (2014) Ménage à trois: post-transcriptional control of the key enzyme for cell envelope synthesis by a base-pairing small RNA, an RNase adaptor protein and a small RNA mimic. RNA Biol 11. doi: 10.4161/rna.28301 Görke, B., and Stülke, J. (2008) Carbon catabolite repression in bacteria: many ways to make the most out of nutrients. Nat Rev Microbiol 6: 613–624. Hahn, M.W., Stajich, J.E., and Wray, G.A. (2003) The effects of selection against spurious transcription factor binding sites. Mol Biol Evol 20: 901–906. Kartha, R.V., and Subramanian, S. (2014) Competing endogenous RNAs (ceRNAs): new entrants to the intricacies of gene regulation. Front Genet 5: 8. Keyhani, N.O., Wang, L.X., Lee, Y.C., and Roseman, S. (2000) The chitin disaccharide, N,N’-diacetylchitobiose, is catabolized by Escherichia coli and is transported/phosphorylated by the phosphoenolpyruvate:glycose phosphotransferase system. J Biol Chem 275: 33084–33090. Kremer, N., Philipp, E.E., Carpentier, M.C., Brennan, C.A., Kraemer, L., Altura, M.A., et al. (2013) Initial symbiont contact orchestrates host-organ-wide transcriptional changes that prime tissue colonization. Cell Host Microbe 14: 183–194. Lee, T.H., and Maheshri, N. (2012) A regulatory role for repeated decoy transcription factor binding sites in target gene expression. Mol Syst Biol 8: 576.
Lybecker, M., Zimmermann, B., Bilusic, I., Tukhtubaeva, N., and Schroeder, R. (2014) The double-stranded transcriptome of Escherichia coli. Proc Natl Acad Sci USA 111: 3134–3139. MacQuarrie, K.L., Fong, A.P., Morse, R.H., and Tapscott, S.J. (2011) Genome-wide transcription factor binding: beyond direct target regulation. Trends Genet 27: 141–148. Nguyen-Duc, T., van Oeffelen, L., Song, N., HassanzadehGhassabeh, G., Muyldermans, S., Charlier, D., and Peeters, E. (2013) The genome-wide binding profile of the Sulfolobus solfataricus transcription factor Ss-LrpB shows binding events beyond direct transcription regulation. BMC Genomics 14: 828. Overgaard, M., Johansen, J., Moller-Jensen, J., and Valentin-Hansen, P. (2009) Switching off small RNA regulation with trap-mRNA. Mol Microbiol 73: 790–800. Plumbridge, J., and Pellegrini, O. (2004) Expression of the chitobiose operon of Escherichia coli is regulated by three transcription factors: NagC, ChbR and CAP. Mol Microbiol 52: 437–449. Plumbridge, J., Bossi, L., Oberto, J., Wade, J.T., and Figueroa-Bossi, N. (2014) Interplay of transcriptional and small RNA-dependent control mechanisms regulates chitosugar uptake in Escherichia coli and Salmonella. Mol Microbiol 92: 648–658. Qian, Z., Dimitriadis, E.K., Edgar, R., Eswaramoorthy, P., and Adhya, S. (2012) Galactose repressor mediated intersegmental chromosomal connections in Escherichia coli. Proc Natl Acad Sci USA 109: 11336–11341. Rasmussen, A.A., Johansen, J., Nielsen, J.S., Overgaard, M., Kallipolitis, B., and Valentin-Hansen, P. (2009) A conserved small RNA promotes silencing of the outer membrane protein YbfM. Mol Microbiol 72: 566–577. Romeo, T., Vakulskas, C.A., and Babitzke, P. (2013) Posttranscriptional regulation on a global scale: form and function of Csr/Rsm systems. Environ Microbiol 15: 313–324. Rydenfelt, M., Cox, R.S., Garcia, H., and Phillips, R. (2014) Statistical mechanical model of coupled transcription from multiple promoters due to transcription factor titration. Phys Rev E Stat Nonlin Soft Matter Phys 89: 012702. Seitz, H. (2009) Redefining microRNA targets. Curr Biol 19: 870–873. Shimada, T., Ishihama, A., Busby, S.J., and Grainger, D.C. (2008) The Escherichia coli RutR transcription factor binds at targets within genes as well as intergenic regions. Nucleic Acids Res 36: 3950–3955. Sterzenbach, T., Nguyen, K.T., Nuccio, S.P., Winter, M.G., Vakulskas, C.A., Clegg, S., et al. (2013) A novel CsrA titration mechanism regulates fimbrial gene expression in Salmonella typhimurium. EMBO J 32: 2872–2883. Stringer, A.M., Currenti, S., Bonocora, R.P., Baranowski, C., Petrone, B.L., Palumbo, M.J., et al. (2014) Genome-scale analyses of Escherichia coli and Salmonella enterica AraC reveal noncanonical targets and an expanded core regulon. J Bacteriol 196: 660–671. Sumazin, P., Yang, X., Chiu, H.S., Chung, W.J., Iyer, A., Llobet-Navas, D., et al. (2011) An extensive microRNAmediated network of RNA-RNA interactions regulates established oncogenic pathways in glioblastoma. Cell 147: 370–381. Taniguchi, Y., Choi, P.J., Li, G.W., Chen, H., Babu, M., Hearn, © 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
Roles of nucleic acid decoy sites 647
J., et al. (2010) Quantifying E. coli proteome and transcriptome with single-molecule sensitivity in single cells. Science 329: 533–538. Verma, S.C., and Mahadevan, S. (2012) The chbG gene of the chitobiose (chb) operon of Escherichia coli encodes a chitooligosaccharide deacetylase. J Bacteriol 194: 4959– 4971. Wade, J.T., Struhl, K., Busby, S.J., and Grainger, D.C. (2007) Genomic analysis of protein-DNA interactions in bacteria: insights into transcription and chromosome organization. Mol Microbiol 65: 21–26. Waldminghaus, T., and Skarstad, K. (2010) ChIP on Chip:
© 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
surprising results are often artifacts. BMC Genomics 11: 414. Wright, P.R., Richter, A.S., Papenfort, K., Mann, M., Vogel, J., Hess, W.R., et al. (2013) Comparative genomics boosts target prediction for bacterial small RNAs. Proc Natl Acad Sci USA 110: E3487–E3496. Yamamoto, S., Mitobe, J., Ishikawa, T., Wai, S.N., Ohnishi, M., Watanabe, H., and Izumiya, H. (2014) Regulation of natural competence by the orphan two-component system sensor kinase ChiS involves a non-canonical transmembrane regulator in Vibrio cholerae. Mol Microbiol 91: 326– 347.
doi:10.1111/mmi.12604 First published online 17 April 2014
MicroCommentary Lies and deception in bacterial gene regulation: the roles of nucleic acid decoys Yvonne Göpel and Boris Görke* Max F. Perutz Laboratories, Department of Microbiology, Immunobiology and Genetics, Center of Molecular Biology, University of Vienna, 1030 Vienna, Austria.
Summary Bacteria use intricately interconnected mechanisms acting at the transcriptional and post-transcriptional level to adjust gene expression to their needs. An intriguing example found in the chitosugar utilization systems of Escherichia coli and Salmonella is uncovered in a study by Plumbridge and colleagues. Three transcription factors (TFs), a small regulatory RNA (sRNA) and a sRNA trap cooperate to set thresholds and dynamics in regulation of chitosugar utilization. Specifically, under inducing conditions a decoy site on the polycistronic chitobiose (chbBCARFG) mRNA sequesters sRNA ChiX, which represses synthesis of the separately encoded chitoporin ChiP. Base-pairing of ChiX with its decoy has no role for the chb genes themselves when the mRNA is in excess. In the absence of substrate, however, this base-pairing tightly represses chbC encoding a subunit of the chitosugar transporter. Thus, one and the same sRNA/ mRNA interaction serves different regulatory functions under different environmental conditions. The employment of RNA decoys to control the activities of post-transcriptional regulators themselves is an increasingly recognized mechanism in gene regulation. Another observation in the current study highlights the possibility that decoy sites might even exist on the DNA controlling the availability of TFs for their target promoters.
Accepted 3 April, 2014. *For correspondence. E-mail boris.goerke@ univie.ac.at; Tel. (+43) 1 4277 54603; Fax (+43) 1 4277 9146.
© 2014 John Wiley & Sons Ltd
Intricate teamwork of three transcriptional regulators, a small RNA and a conditional small RNA decoy for sophisticated control of chitosugar utilization Access to carbon sources is a prerequisite for survival of any living organism. When present in a mixture, bacteria often select between compounds and preferentially use the carbon source that is most easily accessible and allows for fastest growth. Accordingly, carbohydrate metabolic pathways are extensively regulated at all levels (Görke and Stülke, 2008; Bobrovskyy and Vanderpool, 2013). A current study reveals a surprisingly complex regulatory circuit, in which three transcription factors and a small RNA (sRNA) intricately cooperate to regulate utilization of the chitin degradation products chitobiose and chitotriose (Plumbridge et al., 2014). This mechanism further involves an mRNA that, depending on its concentration, functions either as target or trap for the sRNA, representing a novel principle in bacterial gene regulation. Uptake of chitosugars in E. coli and Salmonella involves transport by chitoporin ChiP across the outer membrane followed by translocation into the cytoplasm via the tri-partite PTS-transporter ChbBCA (Keyhani et al., 2000; Figueroa-Bossi et al., 2009). Transcription initiation at the chbBCARFG operon is controlled by a triumvirate of TFs (Plumbridge and Pellegrini, 2004). In the absence of chitosugars, transcription is dually repressed by NagC, which is a global regulator of amino-sugar metabolism, and the operon-specific TF ChbR. In the presence of chitosugars, NagC is released from its operators and ChbR switches its role and stimulates transcription initiation in concert with the cAMP receptor protein. As a peculiarity, the inducers for ChbR and NagC are generated consecutively by catabolism of chitosugars via enzymes ChbG and ChbF respectively (Verma and Mahadevan, 2012). Expression of the separately encoded chiP gene is regulated by sRNA ChiX (Figueroa-Bossi et al., 2009; Rasmussen et al., 2009). ChiX inhibits translation of chiP by base-pairing and causes premature transcription termination by uncoupling translation from transcription (Bossi et al., 2012). In contrast to many other base-pairing
642 Y. Göpel and B. Görke ■
sRNAs, ChiX is constitutively expressed and not codegraded with chiP. Previous elegant studies revealed that activity of ChiX is controlled by an unorthodox mechanism involving an RNA trap. The chb mRNA also binds ChiX through a site in the chbB–chbC intercistronic
region, which dramatically accelerates decay of the sRNA by RNase E. Hence, transcription of chb sequesters and destabilizes ChiX, leading to de-repression of the chiPQ mRNA (Fig. 1A; Figueroa-Bossi et al., 2009; Overgaard et al., 2009). © 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
Roles of nucleic acid decoy sites 643
Fig. 1. Mechanisms employed by nucleic acid decoys in gene regulation. A. Trapping small RNA regulators by RNA decoys. In the absence of the decoy RNA (red) the sRNA regulator (blue) is highly abundant leading to repression of the target RNA (black). Following its induction the decoy RNA accumulates and sequesters the regulatory RNA, which has little effect on the decoy RNA itself due to its vast excess over the sRNA. As a result, the target RNA escapes repression by the sRNA regulator. Notably, the sRNA may also control expression of the decoy RNA itself as exemplified by repression of chbC by sRNA ChiX in the absence of chitosugars (Plumbridge et al., 2014). B. Trapping RNA-binding proteins by RNA decoys. RNA-binding proteins govern stability and/or translational efficiency of specific transcripts through their binding. Accumulation of decoy RNAs in response to a specific stimulus sequesters the protein from the target RNA. Accordingly, repression of the target RNA is relieved. C. Decoy binding sites on the DNA attune gene expression responses to the copy number of the TF (Brewster et al., 2014). The absence of decoy sites generates a linear relation (gray curve) between the number of repressor molecules and the fold-change in gene expression triggered by a promoter (black arrow) with a single operator site (red box). In the presence of decoy binding sites, the response requires a higher number of repressor molecules due to the buffering effect of the decoys. Regulation occurs more gradual, when the affinity of the decoy sites is lower (blue curve) or equal (green curve) as compared with the regulatory binding site. In contrast, a sharp transition between both regimes is observed (red curve), when the decoy sites possess a higher affinity for the TF. The diagram has been adapted from (Brewster et al., 2014).
The current study amends our understanding of this regulatory circuit by adding two novelties (Plumbridge et al., 2014). First, it is shown that transcription initiation at chiPQ is controlled by NagC and not by ChbR as previously speculated. The requirement of ChbR for chiPQ expression is only indirect as it activates chb transcription, which generates the decoy for ChiX and the enzymes producing the inducer for NagC. Second, the effect of ChiX on expression of the chb genes themselves has now been clarified. As expected for a decoy, ChiX has no impact on chb expression under inducing conditions, when chb mRNA levels are likely in excess. In the absence of chitosugars however, ChiX tightly represses translation of chbC setting the basal level of the chitobiose transporter. Thus, depending on the stoichiometry of the partners, one and the same sRNA/mRNA basepairing generates different outputs. When highly transcribed, the chb mRNA serves to trap and destabilize ChiX, but at low transcription rates it becomes a ‘bona fide’ regulatory target of ChiX. Thus, the hunter and the hunted switch their roles (Plumbridge et al., 2014). Taking the novel findings into account, an integral model of regulation of chitosugar utilization is described, which likely includes a multi-tier induction process (Plumbridge et al., 2014). When chitosugars become available, the Chb proteins present at a basal level may catalyse formation of the inducer for ChbR, causing initial increase in chb transcription. However, a corresponding increase of the chitobiose transporter might only occur when chb transcription rates are sufficient to overcome repression by ChiX. Thus, ChiX likely sets the chitobiose threshold concentration and the delay time required for activation of the system. Persistently high chb expression levels lead to destabilization of ChiX and consequently to a first level of chiPQ expression. Continuous synthesis of ChiP and the Chb proteins will boost accumulation of the inducer for NagC. Relief of both operons from repression by NagC may finally amplify chitosugar utilization to maximum levels. © 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
Tight control of chitosugar utilization might prevent induction of the corresponding genes when only low amounts of chitosugars are available. Thus, many Enterobacteriaceae, in which the above described regulatory circuit is likely to be conserved, are perhaps only opportunistic utilizers of chitosugars when these compounds are available in large quantities as sporadic part of the hosts’ diet. This idea is reinforced by the observation that E. coli and Salmonella rely on excreted chitinases of other bacteria for degradation of chitin to the chitosugars that can be utilized. In line, both operons, chiPQ and chb, are subject to strong catabolite repression, i.e. presence of preferred sugars efficiently switches off chitosugar utilization (Plumbridge and Pellegrini, 2004; Plumbridge et al., 2014). In contrast, the role of chitin as carbon source might be more important for aquatic bacteria such as Vibrio species, reflecting that a major part of chitin is produced in aquatic environments. For Vibrio chitin has roles even beyond nutrition, by triggering complex developmental responses such as colonization of host cell tissues (Kremer et al., 2013) and genetic competence as highlighted by another study published earlier this year in Molecular Microbiology (Yamamoto et al., 2014).
RNA decoys: dawn of a novel paradigm in post-transcriptional regulation The employment of decoy RNAs for regulation of distinct RNAs either via sequestration of a shared RNA regulator (Fig. 1A) or by titration of an RNA-binding protein (Fig. 1B) is a hot topic, shifting current paradigms in the understanding of gene regulation. A plethora of cases reminiscent of the role of the chb decoy for regulation of (and by) sRNA ChiX was also reported for eukaryotic microRNAs (miRNA). Transcripts can sequester miRNAs through decoy sites termed microRNA recognition elements (MREs), thereby inhibiting the miRNA response at the true targets (Banks et al., 2012; de Giorgio et al., 2013; Kartha
644 Y. Göpel and B. Görke ■
and Subramanian, 2014). The first of such RNA decoys, with a role in the phosphate starvation response, was discovered in Arabidopsis, where the long non-coding RNA (lncRNA) IPS1 was shown to sequester miR399 by base-pairing, resulting in upregulation of the primary target of miR399 (Franco-Zorrilla et al., 2007). Meanwhile various types of transcripts called competing endogenous RNAs (ceRNAs) are believed to act as miRNAs sponges. These ceRNAs may include pseudogenes, long noncoding RNAs, circular RNAs or even mRNAs. mRNAs may communicate extensively with each other via MREs for shared miRNA. This is exemplified by studies in mammalian cells demonstrating an important role of target mimicry for cancer initiation and progression. In this case, intricate regulatory networks of coding ceRNAs establish cross-talk between different oncogenic pathways by titrating shared miRNA regulators. Similar to the bacterial chb transcript, many of these ceRNAs can act as targets as well as regulators of the cognate miRNA (Sumazin et al., 2011; Kartha and Subramanian, 2014). The concept of RNA mimicry was initially inspired by the paradox that a multitude of conserved target sites for a given miRNA can be predicted by bioinformatics approaches. Yet, only a few of these potential interactions generated detectable phenotypes. Consequently, a decoy function was proposed for these ‘pseudotargets’ (Seitz, 2009). A similar scenario is encountered in bacteria. Even improved search algorithms generate long lists of candidate target sites for a given sRNA. However, only a few of these sites are actual ‘bona fide’ targets (Wright et al., 2013). The remaining hits are usually considered as false-positives, but are they? In fact they could act as sponges for sRNA sequestration, but this possibility has been neglected so far. A different regulatory principle making use of RNA mimicry employs RNAs competing for binding to a common protein (Fig. 1B). A paradigmatic system is provided by the bacterial carbon storage regulatory Csr system. Protein CsrA controls translation and/or stability of its target RNAs by binding to GGA motifs. The activity of CsrA can be counteracted through its sequestration by two cognate sRNAs CsrB and CsrC or even by other mRNAs carrying binding sites for CsrA, recapitulating the hypothesis of ceRNAs (Romeo et al., 2013; Sterzenbach et al., 2013). A comparable mechanism has been reported in human sarcoma cell lines where miRNA miR-29 prevents binding of protein HuR to its cognate target mRNA A20. Since HuR is required for mRNA decay, this mechanism protects A20 from degradation (Balkhi et al., 2013). Competition for a shared protein regulator can also occur between sRNAs themselves as exemplified by the GlmY/ GlmZ system in E. coli (Göpel et al., 2014). Normally, sRNA GlmZ is recruited to degradation by adaptor protein RapZ. However, under conditions that require integrity of
GlmZ, sRNA GlmY acts as decoy and counters this process by sequestering RapZ (Göpel et al., 2013).
Decoy sites on the DNA for sequestration of transcription factors? The principle to employ decoy binding sites for regulation may not only apply to RNA but perhaps even to DNA. This intriguing possibility is emphasized by another interesting observation described in the current study (Plumbridge et al., 2014). A sequence matching the ChbR consensus site is present upstream of the Salmonella chiP promoter and DNA footprinting demonstrates tight and specific interaction of ChbR with this site. Paradoxically, this interaction has no role for regulation of chiPQ expression. Other species including E. coli carry a GalR repressor binding site at the corresponding position, which is conserved to a higher degree as compared with surrounding sequences. Doubtlessly, this site is bound by GalR and its isorepressor GalS with high affinity (Plumbridge et al., 2014). A role of GalR or GalS for regulation of chiP expression remains elusive, once again recapitulating the findings for ChbR and chiP in Salmonella. However, these observations are not unprecedented. Global approaches based on chromatin immunoprecipitation (ChIP) allow mapping of TF binding sites on a global scale. When combined with transcriptome profiling this approach can identify all regulatory relevant binding sites of a TF. One surprising result of these studies is that many TFs occupy specific sites on the chromosome without generating an apparent regulatory output on colocalizing genes. This finding not only applies to globally acting bacterial TFs such as FNR, LexA or σ32, but even for TFs controlling smaller regulons such as RutR, PurR or AraC (Wade et al., 2007; Shimada et al., 2008; Cho et al., 2011; Stringer et al., 2014). Similar observations were reported from archaea and eukaryotes as well (MacQuarrie et al., 2011; Nguyen-Duc et al., 2013). The ‘non-regulatory’ binding sites of TFs can even reside within genes as exemplified by RutR (Shimada et al., 2008). The high number of ‘non-functional’ sites detected in bacteria is highly controversial: It was argued that they represent false-positive artefacts generated by inherent experimental bias (Waldminghaus and Skarstad, 2010). However, based on results obtained by different methods other researchers claim that decoys are genuine binding sites for TFs in vivo (Bonocora et al., 2013). For statistical reasons, sequences similar to the canonical recognition motif of any TF are expected to occur by chance in the chromosome. In fact, LexA was found to bind virtually all sequences in the genome resembling its consensus site (Wade et al., 2007). Are the ‘non-functional’ sites indeed irrelevant and only exist because their random appearance is difficult to avoid during evolution? One counter© 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
Roles of nucleic acid decoy sites 645
argument is that spurious binding sites can be eliminated by negative selection when causing an adverse fitness effect (Hahn et al., 2003). In addition, the non-cognate sites are often conserved across species (MacQuarrie et al., 2011), as also recapitulated by the GalR/GalS site identified in the current study (Plumbridge et al., 2014), arguing against their functional irrelevance. Yet, what is their function? Several hypotheses were proposed to explain noncanonical TF-DNA interactions. First, an involvement of these sites in transcriptional regulation could be masked by redundantly acting TFs or when an additional TF is required. Moreover, a TF could initiate transcription at so far unannotated sites directing expression of e.g. novel non-coding RNA genes. Recently, it was shown that small RNAs may also be expressed from within genes, either in sense or antisense direction (Dornenburg et al., 2010; Chao et al., 2012). In agreement, antisense transcription seems to be pervasive in E. coli and a recent study identified 316 antisense RNAs forming duplexes with their cognate mRNAs indicating that these transcripts are functional in regulation (Lybecker et al., 2014). Second, the non-canonical binding sites could serve biological functions beyond transcriptional control. Operator-bound GalR molecules were proposed to interact with each other causing intra-chromosomal connections. This three-dimensional structure also includes GalR bound to non-canonical sites, which are scattered along the chromosome (Qian et al., 2012). This may contribute to compaction of the chromosome or serve to increase the local concentration of GalR at its target promoters thereby co-ordinating regulation. Strikingly, one of the GalR target genes, galP, is additionally repressed by NagC (El Qaidi et al., 2009). Hence, one possibility is that the GalR binding site observed in the current study (Plumbridge et al., 2014) serves to increase the local concentration of NagC at the galP and chiP promoters via interaction of bound GalR. Similar roles for interaction of distant chromosomal loci or for reorganization of the chromatin structure were also assigned to non-regulatory TF binding sites in higher organisms (MacQuarrie et al., 2011). Apart from these specialized functions, non-canonical binding sites might indeed play an important role in gene regulation by generating complex dosage responses to their cognate TFs and may therefore be termed ‘decoys’ (Burger et al., 2010; Brewster et al., 2014; Rydenfelt et al., 2014). It is commonly assumed that the copy number of a TF exceeds the number of its binding sites in a cell. Consequently, the correlation between the foldchange in transcription and the number of repressors will be linear, when considering a simple repression architecture i.e. a promoter with a single repressor binding site (Fig. 1C; Brewster et al., 2014). That is, the system will behave as if decoy sites are absent. However, several © 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
TFs in E. coli are present in less than 10 copies per cell suggesting that their number is equal or lower than the number of respective binding sites (Taniguchi et al., 2010; Rydenfelt et al., 2014). Under this regime the competing decoy sites provide a buffer and out-compete the regulatory site as long as the TF is limiting. Thus, there will be no major response in fold-changes until all decoy sites are occupied. Once the repressor copy number exceeds the number of decoys, the repressor gets access to the regulatory site. This results in a sharp transition to a scenario which allows regulation of the target gene (Fig. 1C). The steepness of the dose–response curve during this transition depends on the strength of the regulatory binding site. When the affinity of the decoys is higher as compared with the regulatory site, more repressor molecules are required to switch between both regimes (Fig. 1C). Alterations in the dose–response relationship by decoy sites have also been observed for transcriptional activators (Lee and Maheshri, 2012). In sum, decoy sites tune the regulatory response to TF availability in a non-trivial but predictable manner. This provides the basis for the controlled manipulation of transcriptional regulatory responses by using artificial decoys, without any need for alteration of the genome (Brown et al., 2013).
Acknowledgements Research in our laboratory is supported by DFG grant GO1355/7-1 and by FWF grant P26681-B22 to B.G.
References Balkhi, M.Y., Iwenofu, O.H., Bakkar, N., Ladner, K.J., Chandler, D.S., Houghton, P.J., et al. (2013) miR-29 acts as a decoy in sarcomas to protect the tumor suppressor A20 mRNA from degradation by HuR. Sci Signal 6: ra63. Banks, I.R., Zhang, Y., Wiggins, B.E., Heck, G.R., and Ivashuta, S. (2012) RNA decoys: an emerging component of plant regulatory networks? Plant Signal Behav 7: 1188– 1193. Bobrovskyy, M., and Vanderpool, C.K. (2013) Regulation of Bacterial Metabolism by Small RNAs Using Diverse Mechanisms. Annu Rev Genet 47: 227–250. Bonocora, R.P., Fitzgerald, D.M., Stringer, A.M., and Wade, J.T. (2013) Non-canonical protein-DNA interactions identified by ChIP are not artifacts. BMC Genomics 14: 254. Bossi, L., Schwartz, A., Guillemardet, B., Boudvillain, M., and Figueroa-Bossi, N. (2012) A role for Rho-dependent polarity in gene regulation by a noncoding small RNA. Genes Dev 26: 1864–1873. Brewster, R.C., Weinert, F.M., Garcia, H.G., Song, D., Rydenfelt, M., and Phillips, R. (2014) The transcription factor titration effect dictates level of gene expression. Cell 156: 1312–1323. Brown, A.J., Mainwaring, D.O., Sweeney, B., and James, D.C. (2013) Block decoys: transcription-factor decoys designed for in vitro gene regulation studies. Anal Biochem 443: 205–210.
646 Y. Göpel and B. Görke ■
Burger, A., Walczak, A.M., and Wolynes, P.G. (2010) Abduction and asylum in the lives of transcription factors. Proc Natl Acad Sci USA 107: 4016–4021. Chao, Y., Papenfort, K., Reinhardt, R., Sharma, C.M., and Vogel, J. (2012) An atlas of Hfq-bound transcripts reveals 3’ UTRs as a genomic reservoir of regulatory small RNAs. EMBO J 31: 4005–4019. Cho, B.K., Federowicz, S.A., Embree, M., Park, Y.S., Kim, D., and Palsson, B.O. (2011) The PurR regulon in Escherichia coli K-12 MG1655. Nucleic Acids Res 39: 6456–6464. Dornenburg, J.E., Devita, A.M., Palumbo, M.J., and Wade, J.T. (2010) Widespread antisense transcription in Escherichia coli. MBio 1. pii: e00024-10. doi: 10.1128/mBio.0002410. El Qaidi, S., Allemand, F., Oberto, J., and Plumbridge, J. (2009) Repression of galP, the galactose transporter in Escherichia coli, requires the specific regulator of Nacetylglucosamine metabolism. Mol Microbiol 71: 146–157. Figueroa-Bossi, N., Valentini, M., Malleret, L., Fiorini, F., and Bossi, L. (2009) Caught at its own game: regulatory small RNA inactivated by an inducible transcript mimicking its target. Genes Dev 23: 2004–2015. Franco-Zorrilla, J.M., Valli, A., Todesco, M., Mateos, I., Puga, M.I., Rubio-Somoza, I., et al. (2007) Target mimicry provides a new mechanism for regulation of microRNA activity. Nat Genet 39: 1033–1037. de Giorgio, A., Krell, J., Harding, V., Stebbing, J., and Castellano, L. (2013) Emerging roles of competing endogenous RNAs in cancer: insights from the regulation of PTEN. Mol Cell Biol 33: 3976–3982. Göpel, Y., Papenfort, K., Reichenbach, B., Vogel, J., and Görke, B. (2013) Targeted decay of a regulatory small RNA by an adaptor protein for RNase E and counteraction by an anti-adaptor RNA. Genes Dev 27: 552–564. Göpel, Y., Khan, M.A., and Görke, B. (2014) Ménage à trois: post-transcriptional control of the key enzyme for cell envelope synthesis by a base-pairing small RNA, an RNase adaptor protein and a small RNA mimic. RNA Biol 11. doi: 10.4161/rna.28301 Görke, B., and Stülke, J. (2008) Carbon catabolite repression in bacteria: many ways to make the most out of nutrients. Nat Rev Microbiol 6: 613–624. Hahn, M.W., Stajich, J.E., and Wray, G.A. (2003) The effects of selection against spurious transcription factor binding sites. Mol Biol Evol 20: 901–906. Kartha, R.V., and Subramanian, S. (2014) Competing endogenous RNAs (ceRNAs): new entrants to the intricacies of gene regulation. Front Genet 5: 8. Keyhani, N.O., Wang, L.X., Lee, Y.C., and Roseman, S. (2000) The chitin disaccharide, N,N’-diacetylchitobiose, is catabolized by Escherichia coli and is transported/phosphorylated by the phosphoenolpyruvate:glycose phosphotransferase system. J Biol Chem 275: 33084–33090. Kremer, N., Philipp, E.E., Carpentier, M.C., Brennan, C.A., Kraemer, L., Altura, M.A., et al. (2013) Initial symbiont contact orchestrates host-organ-wide transcriptional changes that prime tissue colonization. Cell Host Microbe 14: 183–194. Lee, T.H., and Maheshri, N. (2012) A regulatory role for repeated decoy transcription factor binding sites in target gene expression. Mol Syst Biol 8: 576.
Lybecker, M., Zimmermann, B., Bilusic, I., Tukhtubaeva, N., and Schroeder, R. (2014) The double-stranded transcriptome of Escherichia coli. Proc Natl Acad Sci USA 111: 3134–3139. MacQuarrie, K.L., Fong, A.P., Morse, R.H., and Tapscott, S.J. (2011) Genome-wide transcription factor binding: beyond direct target regulation. Trends Genet 27: 141–148. Nguyen-Duc, T., van Oeffelen, L., Song, N., HassanzadehGhassabeh, G., Muyldermans, S., Charlier, D., and Peeters, E. (2013) The genome-wide binding profile of the Sulfolobus solfataricus transcription factor Ss-LrpB shows binding events beyond direct transcription regulation. BMC Genomics 14: 828. Overgaard, M., Johansen, J., Moller-Jensen, J., and Valentin-Hansen, P. (2009) Switching off small RNA regulation with trap-mRNA. Mol Microbiol 73: 790–800. Plumbridge, J., and Pellegrini, O. (2004) Expression of the chitobiose operon of Escherichia coli is regulated by three transcription factors: NagC, ChbR and CAP. Mol Microbiol 52: 437–449. Plumbridge, J., Bossi, L., Oberto, J., Wade, J.T., and Figueroa-Bossi, N. (2014) Interplay of transcriptional and small RNA-dependent control mechanisms regulates chitosugar uptake in Escherichia coli and Salmonella. Mol Microbiol 92: 648–658. Qian, Z., Dimitriadis, E.K., Edgar, R., Eswaramoorthy, P., and Adhya, S. (2012) Galactose repressor mediated intersegmental chromosomal connections in Escherichia coli. Proc Natl Acad Sci USA 109: 11336–11341. Rasmussen, A.A., Johansen, J., Nielsen, J.S., Overgaard, M., Kallipolitis, B., and Valentin-Hansen, P. (2009) A conserved small RNA promotes silencing of the outer membrane protein YbfM. Mol Microbiol 72: 566–577. Romeo, T., Vakulskas, C.A., and Babitzke, P. (2013) Posttranscriptional regulation on a global scale: form and function of Csr/Rsm systems. Environ Microbiol 15: 313–324. Rydenfelt, M., Cox, R.S., Garcia, H., and Phillips, R. (2014) Statistical mechanical model of coupled transcription from multiple promoters due to transcription factor titration. Phys Rev E Stat Nonlin Soft Matter Phys 89: 012702. Seitz, H. (2009) Redefining microRNA targets. Curr Biol 19: 870–873. Shimada, T., Ishihama, A., Busby, S.J., and Grainger, D.C. (2008) The Escherichia coli RutR transcription factor binds at targets within genes as well as intergenic regions. Nucleic Acids Res 36: 3950–3955. Sterzenbach, T., Nguyen, K.T., Nuccio, S.P., Winter, M.G., Vakulskas, C.A., Clegg, S., et al. (2013) A novel CsrA titration mechanism regulates fimbrial gene expression in Salmonella typhimurium. EMBO J 32: 2872–2883. Stringer, A.M., Currenti, S., Bonocora, R.P., Baranowski, C., Petrone, B.L., Palumbo, M.J., et al. (2014) Genome-scale analyses of Escherichia coli and Salmonella enterica AraC reveal noncanonical targets and an expanded core regulon. J Bacteriol 196: 660–671. Sumazin, P., Yang, X., Chiu, H.S., Chung, W.J., Iyer, A., Llobet-Navas, D., et al. (2011) An extensive microRNAmediated network of RNA-RNA interactions regulates established oncogenic pathways in glioblastoma. Cell 147: 370–381. Taniguchi, Y., Choi, P.J., Li, G.W., Chen, H., Babu, M., Hearn, © 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
Roles of nucleic acid decoy sites 647
J., et al. (2010) Quantifying E. coli proteome and transcriptome with single-molecule sensitivity in single cells. Science 329: 533–538. Verma, S.C., and Mahadevan, S. (2012) The chbG gene of the chitobiose (chb) operon of Escherichia coli encodes a chitooligosaccharide deacetylase. J Bacteriol 194: 4959– 4971. Wade, J.T., Struhl, K., Busby, S.J., and Grainger, D.C. (2007) Genomic analysis of protein-DNA interactions in bacteria: insights into transcription and chromosome organization. Mol Microbiol 65: 21–26. Waldminghaus, T., and Skarstad, K. (2010) ChIP on Chip:
© 2014 John Wiley & Sons Ltd, Molecular Microbiology, 92, 641–647
surprising results are often artifacts. BMC Genomics 11: 414. Wright, P.R., Richter, A.S., Papenfort, K., Mann, M., Vogel, J., Hess, W.R., et al. (2013) Comparative genomics boosts target prediction for bacterial small RNAs. Proc Natl Acad Sci USA 110: E3487–E3496. Yamamoto, S., Mitobe, J., Ishikawa, T., Wai, S.N., Ohnishi, M., Watanabe, H., and Izumiya, H. (2014) Regulation of natural competence by the orphan two-component system sensor kinase ChiS involves a non-canonical transmembrane regulator in Vibrio cholerae. Mol Microbiol 91: 326– 347.
