Plasmid 73 (2014) 10–15
Contents lists available at ScienceDirect
Plasmid journal homepage: www.elsevier.com/locate/yplas
Hfq protein deficiency in Escherichia coli affects ColE1-like but not k plasmid DNA replication Grzegorz M. Cech a,⇑, Bartosz Pakuła a, Dominika Kamrowska a, Grzegorz We˛grzyn a, Véronique Arluison b,c, Agnieszka Szalewska-Pałasz a a
´ sk, Poland Department of Molecular Biology, University of Gdansk, Wita Stwosza 59, 80-308 Gdan Univ Paris Diderot-Paris 7, Sorbone Paris Cité, 75013 Paris, France c Laboratoire Léon Brillouin UMR12 CEA CNRS, CEA-Saclay, 91191 Gif-sur-Yvette Cedex, France b
a r t i c l e
i n f o
Article history: Received 9 December 2013 Accepted 23 April 2014 Available online 5 May 2014 Communicated by Saleem Khan Keywords: Hfq Sm-like protein ColE1 and lambda plasmids DNA replication Nucleoid Small noncoding RNA
a b s t r a c t Hfq is a nucleic acid-binding protein involved in controlling several aspects of RNA metabolism. It achieves this regulatory function by modulating the translational activity and stability of different mRNAs, generally via interactions with stress-related small regulatory sRNAs. However, besides its role in the coordination of translation of bacterial mRNA, Hfq is also a nucleoid-associated DNA-binding protein. Motivated by the above property of Hfq, we investigated if hfq gene mutation has implications for the regulation of DNA replication. Efficiency of ColE1-like (pMB1- and p15A replicons) and bacteriophage k-derived plasmids’ replication has been investigated in wild-type strain and otherwise isogenic hfq mutant of Escherichia coli. Significant differences in plasmid amount and kinetics of plasmid DNA synthesis were observed between the two tested bacterial hosts for ColE1-like replicons, but not for k plasmid. Furthermore, ColE1-like plasmids replicated more efficiently in wild-type cells than in the hfq mutant in the early exponential phase of growth, but less efficiently in late exponential and early stationary phases. Hfq levels in the wildtype host, estimated by Western-blotting, were increased at the latter phases relative to the former one. Moreover, effects of the hfq mutation on ColE1-like plasmid replication were impaired in the absence of the rom gene, coding for a protein enhancing RNA I– RNA II interactions during the control of the replication initiation. These results are discussed in the light of a potential mechanism by which Hfq protein may influence replication of some, but not all, replicons in E. coli. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction The bacterial protein Hfq was discovered in Escherichia coli as a host factor for bacteriophage Qb RNA replication, providing its acronymic name (Franze de Fernandez et al., 1972). Later, the presence of Hfq-encoding genes, orthologues to the E. coli hfq gene, have been found in nearly half of known bacterial genome sequences and in ⇑ Corresponding author. Fax: +48 58 523 60 25. E-mail address:
[email protected] (G.M. Cech). http://dx.doi.org/10.1016/j.plasmid.2014.04.005 0147-619X/Ó 2014 Elsevier Inc. All rights reserved.
archeon (Mura et al., 2013). Subsequent studies indicated its involvement in many other metabolic pathways. Indeed, Hfq is now understood to be a pleiotropic regulator of bacterial gene expression that functions to modulate mRNA translation via its interactions with small regulatory noncoding RNAs (sRNAs). Even if only a part of Hfq-dependent mechanisms of regulation have been characterized thus far, most of them were shown with sRNA acting as a modulator of translation by base pairing around the mRNA ribosome-binding site (rbs), with Hfq facilitating interactions between the sRNA and its mRNA target (Vogel and
G.M. Cech et al. / Plasmid 73 (2014) 10–15
Luisi, 2011). For this reason, pleiotropic effects of hfq mutations have been described, which include change in cell morphology, osmosensitivity, sensitivity to UV-light and oxidants, decrease in growth rate as well as virulence defects in Salmonella (Sittka et al., 2007; Tsui et al., 1994). Moreover, Hfq deficiency also results in decreased negative supercoiling of plasmid DNA in stationary phase (Tsui et al., 1994). In E. coli, the hfq gene codes for a thermostable 102 amino acid-residue protein, which belongs to an evolutionarily conserved family of proteins, the Sm-like family (Arluison et al., 2002; Brennan and Link, 2007; Wilusz and Wilusz, 2013). Interestingly, as its Sm eukaryotic counterparts which participate in pre-mRNA splicing, mRNA decapping and decay, bacterial Hfq is particularly important in RNA metabolism (Tharun, 2009; Vogel and Luisi, 2011). Moreover, similarly to that found in Sm proteins, Hfq forms an OB-like fold in its N-terminal domain that consists in five b-strands that form a strongly bent antiparallel b-sheet, capped by an a-helix. In Hfq, the b-sheets from six monomers interact with each other to assemble in a toroidal hexameric structure (Brennan and Link, 2007). Even if the mechanism by which Hfq binds RNA is not completely understood, it is established now that the N-terminal Sm domain of the protein (about 65 amino acids) binds the RNA (Arluison et al., 2004), that uridinerich RNA sequences are bound to one face of the Sm-ring (called the ‘‘proximal’’ face), while the RNA A-rich sequences bind to the ‘‘distal’’ face of the tore (Link et al., 2009). Furthermore, in addition to the well characterized Sm domain, the E. coli Hfq protein also contains a C-terminal domain of about 35 amino acids that appears mainly non-structured, and which function has not yet been identified (Vogel and Luisi, 2011). In the cell, Hfq is an abundant protein (with intracellular concentration of 10 lM) mainly located in the cytoplasm where it can be associated to ribosomes (Vassilieva and Garber, 2002). Moreover, Hfq interacts with proteins involved in mRNA decay, such as poly A polymerase PAP I, polynucleotide phosphorylase PNP and RNase E (Ikeda et al., 2011; Mohanty et al., 2004). Similarly to some components of the RNA processing machinery of E. coli which have been shown to localize around the periphery of the cell (Taghbalout and Rothfield, 2008), a significant part of the Hfq protein is also found in the close proximity to bacterial membrane (Arluison et al., 2006; Taghbalout et al., 2014). Besides, Hfq has already been described as capable to bind DNA. Even if the protein tends to prefer RNA, its affinity for DNA is significant (Kd 250 nM) especially because Hfq is an abundant protein. This property was notably emphasized in different recent biophysical studies (Geinguenaud et al., 2011; Ohniwa et al., 2013; Updegrove et al., 2010) and in cellular localization experiments demonstrating that roughly 20% of Hfq was associated with DNA (Azam and Ishihama, 1999; Diestra et al., 2009). The protein has no apparent sequence specificity except a preference for A-rich sequences (Geinguenaud et al., 2011). Moreover, the expression of hfq increases during the stationary cell growth phase, relative to the exponential growth of bacterial culture (Diestra et al., 2009;
11
Tsui et al., 1997). It was also proposed that the Hfq protein could be transferred from the nucleoid to the cytoplasm during the stationary phase (Azam and Ishihama, 1999). Motivated by the above properties of Hfq, we asked whether Hfq–DNA interactions might potentially have implications for the regulation of DNA replication. As plasmids are the simplest bacterial replicons, we aimed to analyze whether the lack of a functional Hfq protein in E. coli cell could influence plasmid DNA replication. 2. Materials and methods 2.1. Bacterial strains, plasmids and growth conditions E. coli CF7968 (MG1655 but rph+ DlacIZ mluI) strain was used in this work (Sze et al., 2002). For the construction of the hfq mutant strain (ASP 875), phage P1vir was used for transduction (Lennox, 1955) of the hfq::kan allele from the JW4130-1 strain (Baba et al., 2006) into CF7968 (the wild type strain). Plasmids used in this work were as follows: pET22b – ori pMB1 (from pBR322), 40 copies per cell (Studier et al., 1990); pACYC117 – p15A replicon, 10–12 copies per cell (Rose, 1988); pUC18 – ori pMB1 but devoid of the rom gene, 150–200 copies per cell (Yanisch-Perron et al., 1985), and pKBamp – k replicon, 20 copies per cell. The pKBamp plasmid, constructed by ligation of the AsuI– AatI fragment of the pKB2 plasmid (Kur et al., 1987) containing k origin with the AsuI–AatI fragment of pACYC177 (Rose, 1988) containing the ampicillin-resistance marker, was kindly provided by Monika Macia˛g-Dorszyn´ska (Department of Molecular Biology, University of Gdansk). All strains were cultured at 37 °C with shaking in LB – Lysogeny Broth (Sigma Aldrich), supplemented with antibiotics (when required) as follows: ampicillin at 50 lg/ ml, kanamycin at 50 lg/ml (Sigma Aldrich). 2.2. Preparation of competent cells and efficiency of transformation Bacteria were grown at 37 °C to OD600 = 0.2. A sample of the culture was pelleted (4000 rpm, 5 min). Supernatant was discarded and the pellet was resuspended in 100 mM CaCl2 (0.5 ml of the CaCl2 solution per equivalent of 1 ml of bacterial culture). Cells were incubated on ice for 30 min, then spin down (4000 rpm, 5 min). Next, fresh 100 mM CaCl2 was added (100 ll of the CaCl2 solution equivalent of 1 ml of bacterial culture). Cells were incubated on ice for 1 h before proceeding to transformation. During the transformation procedure, plasmids harboring different replication origins were introduced to bacterial cells in the competent state. The same amount of DNA was used for transformation (40 ng per 100 ll of competent cells). Cells were incubated with DNA on ice for 1 h. Then, bacteria were incubated at 42 °C for 3 min and then transferred quickly to ice-bath for 2 min. 900 ll of prewarmed (to 37 °C) LB was added and then cells were incubated at 37 °C for 5 min with shaking. Bacteria were spread on LB agar plates either supplemented with antibiotic or without any supplementation, and then incubated overnight at 37 °C. To eliminate the difference between the
12
G.M. Cech et al. / Plasmid 73 (2014) 10–15
growth rate of the mutant and the wild type strains, the efficiency of transformation was calculated as a number of transformants per number of viable competent cells. 2.3. Measurement of kinetics of plasmid DNA replication Overnight culture of bacteria bearing plasmids was diluted 100-times in a fresh LB medium, and cultivation was continued at 37 °C. Samples (1.5 ml) were withdrawn at indicated times, and [3H]thymidine was added to final concentration of 10 lCi/ml. Following incubation at 37 °C for 4 min, each sample was chilled on ice and pelleted by centrifugation (10 min, 4000 rpm, 4 °C). Plasmid DNA isolation was performed by using the GenElute™ Plasmid Miniprep Kit (Sigma–Aldrich), according to the manufacturer’s protocol, and radioactivity (shown as counts per minute – CPM) of each sample was measured in a scintillation counter (Perkin-Elmer).
3. Results and discussion Determination of efficiency of transformation of bacterial strains with plasmid DNA is the simplest way to roughly assess effectivity of plasmid replication and maintenance. We found that the E. coli Dhfq mutant, lacking the functional Hfq protein, was transformed by ColE1-like (pMB1- and p15A-derived replicons) and bacteriophage k-derived plasmids with significantly lower efficiency than otherwise isogenic wild-type strain (Fig. 1). However,
2.4. Plasmid DNA quantification assay Concentration of plasmid DNA, isolated as described above, was assessed using PicoGreenÒ dsDNA Quantitation Reagent (Life Technologies, USA), according to the manufacturer’s protocol. 2.5. Western-blotting Overnight culture of the E. coli CF7968 strain was diluted 100-fold in a fresh LB medium, and cultivated at 37 °C. Samples were withdrawn at indicated time points, and to ensure equal amount of bacteria, sample volume was calculated by formula 160/OD. The cells were harvested by centrifugation at 5000 g (10 min) and resuspended in the lysis buffer (120 mM Tris–HCl pH 6.8, 4% SDS, 20% glycerol, 10% b-mercaptoethanol, 0.002% bromophenol blue), and pre-heated at 100 °C for 3 min, followed by brief centrifugation. Total protein extract was separated on 12% SDS–PAGE. Proteins were transferred onto PVDF membranes, then the membranes were incubated in 3% nonfat milk in TBS buffer (50 mM Tris–Cl, pH 7.6, 150 mM NaCl). Primary rabbit antibodies against the C-terminal region of the Hfq protein (prepared by ProteoGenix, Schiltigheim, France) (diluted 1:2000 in 5% nonfat milk TBS buffer) were added and incubated at 4 °C overnight. The membranes were washed three times with TBS buffer and incubated with secondary antibodies, Anti-Rabbit IgG (whole molecule) – peroxidase (Sigma Aldrich), (1:3000) for 1 h at room temperature. The bands corresponding to the Hfq protein were visualized using Chemiluminescent Peroxidase Substrate-3 kit (Sigma Aldrich) and fluoro-imager Typhoon (GE Healthcare). Total protein loading control was estimated either by Ponceau Red membrane staining (0.2% Ponceau Red (Sigma Aldrich); 3% TCA) or by BCA Protein Assay Kit (Novagen). The purified Hfq protein (Arluison et al., 2006) was used as primary antibodies positive control, and protein extract from the hfq mutant strain (prepared as describe above) was used as negative control. Results from Western blotting were quantified by densitometry (Quantity One program).
Fig. 1. Efficiency of transformation of E. coli wild type (closed columns) and Dhfq (open columns) strains with ColE1-like (pET22b, a wild type pMB1 replicon; pUC18, a Drom derivative of pMB1 replicon; pACYC177, a p15A replicon) and k (pKBamp) plasmids. The value for wild type strain was set as 100% for each plasmid. The presented results are mean values from four independent experiments ± SD (standard deviation).
Fig. 2. Growth of E. coli wild type (closed circles) and Dhfq (open circles) strains in the LB medium at 37 °C. Presented results are mean values from five independent experiments with error bars representing SD.
G.M. Cech et al. / Plasmid 73 (2014) 10–15
13
Fig. 3. Efficiency of DNA synthesis (panels A, C, E, G) and amount of DNA (panels B, D, F, H) of pET22b (a wild type pMB1 replicon, panels A and B), pUC18 (a Drom derivative of pMB1 replicon, panels C and D), pACYC117 (a p15A replicon, panels E and F), and pKBamp (k phage-derived replicon, panels G and H) plasmids in E. coli wild-type (closed columns and circles) and Dhfq strains (open columns and circles) grown in the LB medium at 37 °C. Presented results are mean values from at least two independent experiments, with each measurement repeated at least twice. Error bars represent SD values.
14
G.M. Cech et al. / Plasmid 73 (2014) 10–15
opposite results were obtained with pUC18 plasmid, a pMB1-derived replicon lacking the rom gene which codes for a protein involved in enhancing interactions between RNA I (an antisense regulator) and RNA II (a pre-primer RNA) transcripts, the main step in the regulation of ColE1-type plasmid replication initiation. These results suggested an impaired replication and/or maintenance of some plasmids in the Dhfq mutant. However, due to multiple phenotypes caused by this mutation, the observed decreased or increased transformation efficiency could also arise, at least partially, from secondary effects of a lack of Hfq. Therefore, we measured a kinetics of plasmid DNA synthesis and plasmid DNA amount in bacterial cells. Growth curves of cultures of the hfq mutant and its wild-type counterpart are presented in Fig. 2. Samples of these cultures were withdrawn at indicated times, and plasmid DNA level, as well as incorporation of radioactive thymidine into plasmid DNA were measured. Surprisingly, we found that relative amount of pMB1derived wild-type plasmid was higher in the hfq mutant than in the wild-type host, especially in late exponential and early stationary phases of growth (Fig. 3). These differences were less pronounced for another ColE1-like plasmid, a p15A-derived replicon, and did not occur in k plasmid. When comparing efficiency of plasmid DNA synthesis, the general tendency was that the most efficient plasmid DNA synthesis occurred in the mid-exponential phase, with less intensive incorporation of the radioactive precursor in early exponential and early stationary phases. However, when comparing the hfq mutant and its wildtype counterpart, both ColE1-like plasmids replicated more intensively in the wild-type strain in early exponential phase, while synthesis of plasmid DNA was several fold more efficient in bacteria devoid of Hfq in late exponential and early stationary phases (Fig. 3). Interestingly, kinetics of k plasmid DNA synthesis was not significantly different in both tested strains (Fig. 3). When pUC18 plasmid (a Drom derivative of pMB1 replicon) was tested, the results resembled those obtained with k plasmid rather than the wild-type pMB1 replicon. Nevertheless, synthesis of pUC18 DNA was particularly efficient at the late exponential phase of growth of the hfq mutant which might explain the high efficiency of transformation of this host by this plasmid. The results described above and presented in Fig. 3 indicate that the Hfq protein may significantly influence replication of ColE1-like plasmids, but not that of k plasmids. This may suggest specificity of Hfq-mediated control of plasmid replication. Interestingly, ColE1-like plasmids replicated more efficiently in the wild-type host than in the hfq mutant in early exponential phase of bacterial culture growth, while opposite phenomenon (significantly more intensive incorporation of [3H]thymidine into plasmid DNA in the mutant cells) was observed in late exponential and early stationary phases. It is likely that such differences in effects of the Hfq deprivation may result from natural fluctuations of levels in this protein during bacterial cultivation, as it was reported previously that the intracellular Hfq amount increases in the stationary phase relative to the exponential one (Diestra et al., 2009; Tsui et al.,
1997). Indeed, we found that levels of this protein in the wild-type strain were significantly elevated in the late exponential and early stationary phases relative to early exponential growth (Fig. 4). Therefore, one may suggest that Hfq is a negative regulator of replication of ColE1-like plasmids, and that the Hfq-mediated control is pronounced effectively when the level of this protein in E. coli cell is sufficiently high. Presented results suggest the Hfq-dependent control of plasmid replication in the stationary phase. Whether Hfq-dependent control of ColE1-like plasmid DNA replication is direct or indirect remains to be elucidated. Indeed, this process is tightly controlled by a number of factors and may arise from: (i) putative specific interactions of Hfq with plasmid DNA, (ii) binding of Hfq to RNA molecules involved in the replication regulation, and/or (iii) the regulation of expression of gene(s) being under control of sRNA-pathway(s). This includes either the scenario when the absence of Hfq reverses the sRNAmediated arrest of expression of genes involved in plasmid DNA replication or when the lack of Hfq decreases stability of some sRNA transcripts, as it was shown for DsrA, which is unstable in hfq mutants (Guantes et al., 2012; McCullen et al., 2010). The impairment of the hfq mutation effects on ColE1-like plasmid replication in the absence of the rom gene favors the second option. However, since differences in pUC18 replication between wild-type and hfq hosts were still evident, other hypotheses can still be valid. It is worth noting that both ColE1-like and k replicons require transcription for DNA replication initiation. However, k replicons depend on production of mRNA for replication proteins and for the process of transcription proceeding at the orik region, called transcriptional activation of the origin (Mensa-Wilmot et al., 1989). On the other hand, specific interactions between RNA molecules (preprimer RNA II and inhibitory RNA I), and between the
Fig. 4. Levels of the Hfq protein in E. coli cells during bacterial growth, assessed by Western-blotting. Upper panel shows a representative Western-blotting membrane with Hfq protein visualization, and lower panel presents quantitative results (mean values from three independent experiments with error bars representing SD values).
G.M. Cech et al. / Plasmid 73 (2014) 10–15
RNA II transcript and its DNA template, occur during initiation of ColE1-like plasmid replication (Allen et al., 2011). Therefore, one might suggest that since Hfq is an RNA chaperone, differences between effects of the hfq mutation on replication of both types of tested plasmids arise from specific involvement of Hfq in RNA–RNA and/or DNA– RNA interactions in ColE1-like replicons, which are absent or irrelevant in k plasmids. This hypothesis can be supported by impairment of effects of the hfq mutation on ColE1-like plasmid replication in the absence of the rom gene. Both Rom and Hfq affect RNA structure and may modulate RNA–RNA interactions. Therefore, interference between functions of these proteins in the RNA I–RNA II transactions is likely. 4. Conclusions We present evidence that a lack of the Hfq protein may affect plasmid DNA replication. ColE1-like plasmids replicate more efficiently in wild-type cells relative to the hfq mutant in the early exponential phase of growth, but less efficiently in late exponential and early stationary phases; this phenomenon was not observed for k plasmid. Whether Hfq-dependent control of ColE1-like plasmid DNA replication is direct or indirect remains to be further elucidated, nevertheless, influence of this protein on RNA I–RNA II interactions is likely, as effects on the hfq mutations were impaired in the absence of the rom gene, coding for an enhancer of these interactions. Acknowledgments We would like to thank Monika Macia˛g-Dorszyn´ska for providing of the pKBamp plasmid and Anna Kloska and _ ´ for many invaluable scientific Maja Kochanowska-Łyzen advices. The hfq mutant (JW4130-1) was obtained from the Keio collection (25, National BioResource Project (NIG, Japan): E. coli). This work was supported by the National Science Center (Poland) (project grant no. UMO2012/04/M/NZ1/00067 to G.W.), by the CNRS, CEA and University Paris Diderot (to V.A.), and by the POLONIUM program (task no. 534-L000-S013-12-1E) of Polish-French scientific collaboration (to G.W. and V.A.). References Allen, J.M. et al., 2011. Roles of DNA polymerase I in leading and laggingstrand replication defined by a high-resolution mutation footprint of ColE1 plasmid replication. Nucleic Acids Res. 39, 7020–7033. Arluison, V. et al., 2002. Structural modelling of the Sm-like protein Hfq from Escherichia coli. J. Mol. Biol. 320, 705–712. Arluison, V. et al., 2004. The C-terminal domain of Escherichia coli Hfq increases the stability of the hexamer. Eur. J. Biochem. 271, 1258– 1265. Arluison, V. et al., 2006. Three-dimensional structures of fibrillar Sm proteins: Hfq and other Sm-like proteins. J. Mol. Biol. 356, 86–96. Azam, T.A., Ishihama, A., 1999. Twelve species of the nucleoid-associated protein from Escherichia coli. Sequence recognition specificity and DNA binding affinity. J. Biol. Chem. 274, 33105–33113. Baba, T. et al., 2006. Construction of Escherichia coli K-12 in-frame, singlegene knockout mutants: the Keio collection. Mol. Syst. Biol. 2 (2006), 0008.
15
Brennan, R.G., Link, T.M., 2007. Hfq structure, function and ligand binding. Curr. Opin. Microbiol. 10, 125–133. Diestra, E. et al., 2009. Cellular electron microscopy imaging reveals the localization of the Hfq protein close to the bacterial membrane. PLoS One 4, e8301. Franze de Fernandez, M.T. et al., 1972. Bacterial proteins required for replication of phage Q ribonucleic acid. Purification and properties of host factor I, a ribonucleic acid-binding protein. J. Biol. Chem. 247, 824–831. Geinguenaud, F. et al., 2011. Conformational transition of DNA bound to Hfq probed by infrared spectroscopy. Phys. Chem. Chem. Phys. 13, 1222–1229. Guantes, R. et al., 2012. Positive regulatory dynamics by a small noncoding RNA: speeding up responses under temperature stress. Mol. Biosyst. 8, 1707–1715. Ikeda, Y. et al., 2011. Hfq binding at RhlB-recognition region of RNase E is crucial for the rapid degradation of target mRNAs mediated by sRNAs in Escherichia coli. Mol. Microbiol. 79, 419–432. Kur, J. et al., 1987. Escherichia coli dnaA initiation function is required for replication of plasmids derived from coliphage lambda. J. Mol. Biol. 198, 203–210. Lennox, E.S., 1955. Transduction of linked genetic characters of the host by bacteriophage P1. Virology 1, 190–206. Link, T.M. et al., 2009. Structure of Escherichia coli Hfq bound to polyriboadenylate RNA. Proc. Natl. Acad. Sci. USA 106, 19292–19297. McCullen, C.A. et al., 2010. Mechanism of positive regulation by DsrA and RprA small noncoding RNAs: pairing increases translation and protects rpoS mRNA from degradation. J. Bacteriol. 192, 5559–5571. Mensa-Wilmot, K. et al., 1989. Transcriptional activation of bacteriophage lambda DNA replication in vitro: regulatory role of histone-like protein HU of Escherichia coli. EMBO J. 8, 2393–2402. Mohanty, B.K. et al., 2004. The Sm-like protein Hfq regulates polyadenylation dependent mRNA decay in Escherichia coli. Mol. Microbiol. 54, 905–920. Mura, C. et al., 2013. Archaeal and eukaryotic homologs of Hfq: a structural and evolutionary perspective on Sm function. RNA Biol. 10, 636–651. Ohniwa, R.L. et al., 2013. Atomic force microscopy analysis of the role of major DNA-binding proteins in organization of the nucleoid in Escherichia coli. PLoS One 8, e72954. Rose, R.E., 1988. The nucleotide sequence of pACYC177. Nucleic Acids Res. 16, 356. Sittka, A. et al., 2007. The RNA chaperone Hfq is essential for the virulence of Salmonella typhimurium. Mol. Microbiol. 63, 193–217. Studier, F.W. et al., 1990. Use of T7 RNA polymerase to direct expression of cloned genes. Meth. Enzymol. 185, 60–89. Sze, C.C. et al., 2002. Integration of global regulation of two aromaticresponsive sigma(54)-dependent systems: a common phenotype by different mechanisms. J. Bacteriol. 184, 760–770. Taghbalout, A., Rothfield, L., 2008. New insights into the cellular organization of the RNA processing and degradation machinery of Escherichia coli. Mol. Microbiol. 70, 780–782. Taghbalout, A. et al., 2014. The Escherichia coli RNA processing and degradation machinery is compartmentalized within an organized cellular network. Biochem. J. 458, 11–22. Tharun, S., 2009. Lsm1-7-Pat1 complex: a link between 30 and 50 -ends in mRNA decay? RNA Biol. 6, 228–232. Tsui, H.C. et al., 1997. Negative regulation of mutS and mutH repair gene expression by the Hfq and RpoS global regulators of Escherichia coli K12. J. Bacteriol. 179, 7476–7487. Tsui, H.C. et al., 1994. Characterization of broadly pleiotropic phenotypes caused by an hfq insertion mutation in Escherichia coli K-12. Mol. Microbiol. 13, 35–49. Updegrove, T.B. et al., 2010. E. coli DNA associated with isolated Hfq interacts with Hfq’s distal surface and C-terminal domain. Biochim. Biophys. Acta 1799, 588–596. Vassilieva, I.M., Garber, M.B., 2002. The regulatory role of the Hfq protein in bacterial cells. Mol. Biol. 36, 785–791. Vogel, J., Luisi, B.F., 2011. Hfq and its constellation of RNA. Nat. Rev. Microbiol. 9, 578–589. Wilusz, C.J., Wilusz, J., 2013. Lsm proteins and Hfq: life at the 30 end. RNA Biol. 10, 592–601. Yanisch-Perron, C. et al., 1985. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33, 103–119.
