RNA-binding protein Hfq is a key component of the adaptive responses of many proteobacterial species including Escherichia coli, Salmonella enterica and Vibrio cholera. In these organisms, the importance of Hfq largely stems from its participation to regulatory mechanisms involving small non-coding RNAs. In contrast, the function of Hfq in Gram-positive bacteria has remained elusive and somewhat controversial. In the present study, we have further addressed this point by comparing growth phenotypes and transcription profiles between wild-type and an hfq deletion mutant of the model Gram-positive bacterium, Bacillus subtilis. The absence of Hfq had no significant consequences on growth rates under nearly two thousand metabolic conditions and chemical treatments. The only phenotypic difference was a survival defect of B. subtilis hfq mutant in rich medium in stationary phase. Transcriptomic analysis correlated this phenotype with a change in the levels of nearly one hundred transcripts. Albeit a significant fraction of these RNAs (36%) encoded sporulation-related functions, analyses in a strain unable to sporulate ruled out sporulation per se as the basis of the hfq mutant’s stationary phase fitness defect. When expressed in Salmonella, B. subtilis hfq complemented the sharp loss of viability of a degP hfq double mutant, attenuating the chronic σE-activated phenotype of this strain. However, B. subtilis hfq did not complement other regulatory deficiencies resulting from loss of Hfq-dependent small RNA activity in Salmonella indicating a limited functional overlap between Salmonella and B. subtilis Hfqs. Overall, this study confirmed that, despite structural similarities with other Hfq proteins, B. subtilis Hfq does not play a central role in post-transcriptional regulation but might have a more specialized function connected with stationary phase physiology. This would account for the high degree of conservation of Hfq proteins in all 17 B. subtilis strains whose genomes have been sequenced.
Citation: Rochat T, Delumeau O, Figueroa-Bossi N, Noirot P, Bossi L, Dervyn E, et al. (2015) Tracking the Elusive Function of Bacillus subtilis Hfq. PLoS ONE 10(4): e0124977. https://doi.org/10.1371/journal.pone.0124977
Academic Editor: Lennart Randau, Max-Planck-Institute for Terrestrial Microbiology, GERMANY
Received: January 21, 2015; Accepted: March 20, 2015; Published: April 27, 2015
Copyright: © 2015 Rochat et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work was supported by the Agence Nationale pour la Recherche (http://www.agence-nationale-recherche.fr/) (ANR-2010-BLAN-1602-01 ‘Duplex-Omics’ and ANR-12-BSV6-0008 'ReadRNA') to PB and by the European Commission–funded BaSysBio project (http://ec.europa.eu/research/fp6/index_en.cfm) (LSHG-CT-2006-037469) to PN. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Hfq is a RNA-binding protein that plays a crucial role in post-transcriptional regulation in many bacteria (reviewed in refs [1–3]). Escherichia coli Hfq (HfqEc) was first shown to be required for RNA phage Qß replication , but its function in uninfected host cells remained unknown for a long time. In 1994, the discovery of phenotypes associated with hfqEc insertion mutants revealed that HfqEc was important for bacterial physiology , but the origin of these phenotypes was only progressively uncovered through the discovery of Hfq-dependent regulatory RNAs.
Base-pairing association of regulatory RNAs with RNA-partner molecules is a conserved mechanism to regulate gene expression. In bacteria, regulatory RNAs, usually small and noncoding, affect mRNA translation and stability to modulate numerous processes, including plasmid replication, envelope homeostasis, iron homeostasis, virulence and central metabolism (reviewed in ref. ). Small RNAs (sRNAs) that are expressed from genetic regions unlinked to their targets are referred to as trans-encoded RNAs; in these instances, base-pairing is often imperfect with a limited nucleotide complementarity.
In many bacteria including E. coli, Salmonella and Vibrio, Hfq is required for the activity of most, if not all, trans-encoded RNAs. Hfq protects sRNAs against degradation by ribonucleases and is thought to stimulate pairing with their targets . As a result, Hfq alterations in these organisms typically produce highly pleiotropic effects. While most of these affects can be ascribed to the loss of sRNA-mediated regulation, some evidence suggests that Hfq can also affect gene expression directly, via sRNA-independent pathways .
Orthologs of the hfq gene are found in about half of the bacterial genomes ; however, their involvement in RNA regulatory mechanisms is sometimes unclear, in particular within the Bacilli class (i.e., Staphylococcus aureus, Listeria monocytogenes and Bacillus subtilis). While the Staphylococcal Hfq (HfqSa) structure has been known for more than ten years , there is little information about its activity. In several pathogenic isolates, HfqSa is either poorly or not expressed [11–13] and consistently, deletion of its gene does not impact the physiology of these isolates . On the other hand, in strains where HfqSa has been detected, deletion of hfqSa reportedly affected strain toxicity and virulence . Although HfqSa was shown to associate with RNAs in vitro, no discernible effect on sRNAs mediated translational repression has been demonstrated in vivo [12,14–16], therefore questioning the role of HfqSa in this type of regulation. In addition, HfqSa cannot complement for the absence of its homolog in Salmonella, HfqSTM , indicating that the two proteins are not functionally equivalent.
Deletion of the L. monocytogenes hfq gene (hfqLm) did not affect growth, except upon salt, ethanol or Triton X-100 exposure . Three RNAs binding HfqLm were identified . One of them, LhrA, down-regulates expression of three genes and its stability is affected in an Hfq-dependent manner [20,21]. However, HfqLm-dependent sRNA stability does not seem to be a general feature, since the abundance of other twelve sRNAs was not affected by HfqLm , and a comparative transcriptome analysis of the hfqLm mutant with its isogenic parental strain also indicated that none of the identified sRNAs were affected by the absence of Hfq . Recent structural studies indicate that HfqLm-RNA interactions differ from those established by the Hfq proteins of other Gram-positive bacteria [24,25].
Several recent studies point to a minor role of HfqBs in B. subtilis physiology. The growth rates of hfqBs mutants in glucose-supplemented minimal media are identical to those of the corresponding wild-type strains [26,27]. HfqBs associates in vivo with sRNAs and the 5’ leader regions of some mRNAs and its absence affect the abundance of few sRNAs . In at least one case, HfqBs was implicated in the translation of an mRNA (ahrC) [29,30]. However, the regulatory activities of several sRNAs were found to be HfqBs-independent [26,29–32]. Thus, despite the extensive knowledge on the structure of the Hfq protein from B. subtilis and other Gram-positive bacteria, the physiological role(s) of these proteins remain(s) poorly understood.
In this study, we have carried out a systematic and comprehensive analysis of the physiology of an HfqBs deletion mutant as compared to the wild type strain. We discovered that the main consequence of the hfqBs deletion is a decreased fitness in stationary phase. This defect correlated with a change in the levels of approximately 100 transcripts including transcripts related to sporulation and to Type-I toxin-antitoxin (TA) systems. However, we found the fitness defect to be independent of sporulation process and of the presence of txpA-RatA and bsrE-as-BsrE Type-I toxin-antitoxin (TA) systems. While this work was under way, an article describing a similar analysis of a B. subtilis hfq mutant appeared in the press . The results of latter study concord with ours in some aspects (stationary phase fitness phenotype) but differ in others (transcriptional profiling). We discuss possible sources of these discrepancies. Overall, however, the results from the two studies point to the conclusion that hfqBs is not a major player of sRNA-mediated regulation but its integrity is essential to ensure the bacterial survival under starvation conditions.
Patterns of hfqBs expression
In the course of a large scale transcriptome study, we observed that hfqBs is transcribed under all conditions tested, with initiation occurring at three distinct promoters: i) a σA-dependent promoter located upstream miaA, ii) a σH-dependent promoter upstream hfqBs active in stationary phase (see also ) and iii) an early stage sporulation promoter (σEF) leading to ymaF-miaA-hfqBs transcript (S1 File and ref. ). Two recent studies used translational gene fusions to the chromosomal hfqBs locus to measure HfqBs expression levels as a function of the growth phase in synthetic media. While the first study concluded that Hfq levels increase in cells in stationary phase , the second study reported no significant difference between all growth phases . We also independently constructed a strain with the hfqBs gene terminally fused to the peptide affinity (SPA) tag (containing three FLAG epitope repeats). The resulting strain, BSB1 hfqBs::SPA (TR229), expresses the HfqBs::SPA protein under the control of the hfqBs native promoter as a unique source of Hfq. When TR229 is grown in rich medium (LB), HfqBs::SPA significantly accumulates during the transition from exponential to stationary phase as revealed by Western blotting (Fig 1).
Cultures of strain carrying hfqBs-spa translational fusion (TR229) or wild-type strain (BSB1) were performed in LB medium at 37°C under vigorous agitation. Harvested cells from exponential, early and late transition and stationary phase cultures (OD600nm 1, 2.4, 3 and 4, respectively) were lysed by sonication and 2.75 μg of total protein extracts of each sample were separated by electrophoresis in 12.5% SDS-PAGE. HfqBsSPA was detected by immunoblotting using anti-FLAG M2 antibodies (Sigma).
High throughput phenotypic analysis of the ΔhfqBs mutant
In many bacteria, defective phenotypes resulting from Hfq inactivation can be revealed by exposing cells to stress conditions (e.g., oxidative stress, iron starvation, high temperature). Therefore, we sought to identify possible phenotypic alterations by comparing the growth of wild-type B. subtilis (BSB1, ref. ) and its isogenic Δhfq derivative (TR223, cf. Material and Methods) under numerous conditions. Nearly two thousand conditions were tested by performing a phenotype microarray analysis (cf. Material and Methods) . No growth differences were observed between the two strains when tested for the utilization of various carbon, nitrogen, phosphorus and sulfur sources, or when cells were subjected to nutrient upshifts or conditions affecting osmolarity or pH (S2 File). Beside a resistance to amphenicol of ΔhfqBs strain associated with presence of a chloramphenicol cassette in the mutant strain, very few effects related to chemicals were detected. The phenotype microarray analysis suggested that the Δhfq strain could be more resistant to compound 48/80 and more sensitive to the 2,4-Diamino-6,7-diisopropylpteridine and cetylpyridinium as compared to the wild-type strain. However, further tests showed these observations to be artifacts of the microarray analysis since no growth differences exist between the wild-type strain and its Δhfq derivative grown in the presence of these three compounds (S2 File). We concluded that HfqBs does not affect the growth rate or growth yield of B. subtilis under any of the tested conditions. Thus, HfqBs does not appear to have the same impact on adaptation responses as its homologs in enteric bacteria.
HfqBs improves survival in stationary phase
One of the three promoters transcribing the hfqBs gene is under the control of σH (S1 File, ref. ). This alternative sigma factor directs the transcription of genes required for cellular adaptation during transition from exponential to stationary phase, including induction of sporulation, genetic competence or biofilm development [36,37]. The observed increase in the amount of HfqBs in transition and stationary phases (Fig 1 and ref ) likely reflects activation of hfqBs transcription from the σH promoter. We wondered whether HfqBs activity could be linked to σH-dependent phenotypes. The ability of wild-type and ΔhfqBs strains to form biofilms or to uptake exogenous DNA was monitored, but no significant difference was observed (S3 File and Fig 2).
The apparition and the proportion of competent cells in cultures of wild-type (BSB1) and Δhfq mutant (TR223) strains were monitoring by calculating the number of cells able to integrate an antibiotic resistance gene in their chromosomal DNA during the competence process development.
We also considered that HfqBs could contribute to B. subtilis survival in stationary phase. To assess this possibility, we compared the viability of hfqBs+ and ΔhfqBs strains in co-cultures in rich medium. Each strain had a specific antibiotic resistance gene—either cat (chloramphenicol) or spc (spectinomycin)—inserted either in the hfqBs gene (alleles ΔhfqBs::cat and ΔhfqBs::spc) or in the intergenic region (igr) between hfqBs and ymzE (alleles igr::cat and igr::spc). The permutation of resistance markers allowed correcting for possible loss of fitness arising from the antibiotic resistance gene. Samples from the co-cultures were plated on selective media and survival was assayed counting the corresponding colony forming units (CFU). This analysis revealed that the loss of hfqBs results in a decreased survival in aging cultures (Fig 3A). This is in agreement with data recently reported and shows that hfqBs positively affects the survival of bacterial cells independently of the growth medium (i.e. minimal CS-glucose  vs LB [this study]).
A) Co-cultures were performed with hfqBs-expressing strain and ΔhfqBs mutant in LB medium and incubated at 37°C during 5 days. An antibiotic resistance gene (cat or spc) was inserted in the intergenic region hfq-ymzE (control, black) or in replacement of the hfqBs coding sequence (mutant, white). Each population was numbered on LB plates supplemented with spectinomycin or chloramphenicol. To number spores, samples were incubated 15 min at 80°C before plating (dot lines). Two combinations of co-cultures were performed using two sets of isogenic strains (●) TR247 (hfqBs+ igr::cat) and (◊) TR232 (ΔhfqBs::spc igr+) or strains (○) TR223 (ΔhfqBs::cat igr+) and (♦) TR241 (hfqBs+ igr::spc). B) The same experiments were performed with sporulation-deficient derivative strains (sigE::erm). The two co-cultures were performed using strains (●) TR243 and (◊) TR237 or strains (○) TR235 and (♦) TR245. C) Competition experiments were performed with TR259 and TR255 strains which are deleted of the bsrE-asBsrE type 1 TA and express (●) or not (◊) HfqBs. D) Competition experiments were performed with the TF8A (txpA-RatA deleted) derivative strains using ΔhfqBs TR248 (◊) and control TR252 (●) strains.
Transcriptome analysis of the hfqBs mutant: RNA changes in stationary phase
In Enterobacteria, loss of Hfq results in a reduced stability of many sRNAs and the concomitant deregulation of the corresponding mRNA targets under unperturbed conditions (reviewed in refs. [1–3]). Differences in RNA patterns between the wild-type and Δhfq strains of B. subtilis could reveal potential HfqBs-targets and possibly provide insight into the molecular basis for the observed stationary phase phenotype. Therefore, the transcriptomes of these two strains were analyzed by comparing tiling array hybridization profiles of RNAs extracted from cells cultured in rich medium (LB) at two different growth stage: i) exponential phase (OD600nm = 0.5) and ii) 5 hours after the onset of stationary phase. Transcripts were positioned on our B. subtilis 168 structural annotation map  and changes were investigated by differential expression analysis (cf. Material and Methods).
Somewhat surprisingly, the RNA profiles from exponentially growing cells were identical in the wild-type and ΔhfqBs strains, except for the obvious absence of the hfqBs RNA in the ΔhfqBs mutant (S4 and S5 Files). Many regulatory sRNAs were detected, including FsrA  (S512), RatA  (S976), RoxS (Ncr22 or S415) [32,38,39] (alias RsaE in S. aureus [15,40]), as-bsrE (Ncr1019, S718) as well as 16 other sRNAs of unknown function. None of these sRNAs, nor their known mRNA targets were affected by the absence of HfqBs (S5 File and S1 Table). Thus, HfqBs does not appear to influence B. subtilis RNA patterns during the exponential phase to any significant extent, at least in cells grown in rich medium. The situation might be different in minimal CS-glucose-grown cells where 68 mRNAs and one sRNA (ncr1670; S357) were reported to be affected by HfqBs inactivation .
Unlike the results from the exponential samples, 97 transcription units (representing 134 genes) were found significantly different between the wild-type and the ΔhfqBs strains in the stationary cultures (S6 File). Fold-change values and functional information are available in S1 Table. Many of the affected RNAs are from genes under the control of stationary phase-specific transcriptional regulators suggesting that Spo0A was activated in ΔhfqBs strain as exemplified by the deregulation of 18 genes of its regulon . Overall, 48 out of the 134 genes are linked to sporulation. Portions of the AbrB, SigH and SigD regulons also show their expression changed in the absence of Hfq. Some genes are related to respiration and anaerobiosis: the rex-ndh operon appears upregulated in the ΔhfqBs strain whereas the ctaDEF operon together with the cccA and qoxB loci are down-regulated. The envelope stress response controlled by SigM is also slightly activated in the absence of Hfq (S1 Table). Besides the modifications observed for protein-coding genes, the absence of Hfq affected the amount of i) two sRNAs: S1022, a SigD-induced RNA , and S1495 (unknown function); ii) four riboswitches (upstream of thiC, trpS, ileS and thrZ); and iii) two type-I toxin-antitoxin systems (namely the antisense RNAs as-bsrE and ratA as well as the toxin txpA). We considered that the changes in RNA profiles might arise from the upregulation of HfqBs in stationary phase (Fig 1) and might help explaining the mechanism responsible for HfqBs contribution to survival under these conditions (Fig 3A).
The survival advantage conferred by HfqBs is independent of sporulation
Sporulation is an energy and time consuming process that is used by B. subtilis as a last resource for survival when other adaptation programs like exogenous DNA-uptake or cannibalism have failed . Based on the transcriptomic profiles, we hypothesized that under nutrient deprivation, cells lacking hfq may enter sporulation earlier than wild-type cells, thus losing the chance to use the residual nutrients available. To test if the survival defect observed in the hfq mutant was dependent of sporulation, the competition experiment (see above) was repeated in cells carrying the combinations of hfq and igr alleles in the background of a sporulation-deficient (ΔsigE::erm) mutant. Results showed the ΔhfqBs deletion to still confer a survival defect in this background (Fig 3B) suggesting that the defect is maintained in vegetative cells and thus unlikely to relate to sporulation.
The survival advantage conferred by Hfq is independent of BsrE or TxpA type 1 Toxin-Antitoxin systems
Among the RNAs affected by the hfqBs deletion in stationary phase are antitoxin as-bsrE and RatA antisense RNAs (S1 Table). A third antitoxin RNA, as-bsrG (ncr1932), was found in reduced amounts in a separate study . Altogether these findings raised the possibility that alteration in the regulation of toxin/antitoxin systems might be responsible for the decreased fitness of the hfqBs mutant. However, further analyses ruled out this possibility as well. Deleting the entire bsrE/as-BsrE type-1 TA system did not relieve the survival defect of the hfqBs mutant (Fig 3C). Furthermore, the survival defect associated with the ΔhfqBs deletion was still observed in a B. subtilis strain cured for the SKIN and SP-β, prophages, which, combined, contribute four different TA systems, namely TxpA and BsrH (SKIN) and BsrG and YonT (SP-β) (Fig 3D).
Complementation tests in Salmonella
In Salmonella and E. coli, Hfq-regulated genes provide simple, yet sensitive assays for monitoring Hfq function in vivo. To take advantage of this system, we replaced the coding portion of the Salmonella hfq gene (hfqSTM) with the corresponding segment of the B. subtilis (hfqBs). Using a strategy previously reported , we demonstrated that hfqBs and hfqSTM were both expressed in a similar amount (Fig 4A). We then tested the effects of the gene exchange on the regulation of chiP and yifK genes, regulated by Hfq-dependent ChiX and GcvB sRNAs, respectively [44,45]. Deleting the hfqSTM gene in Salmonella (by replacement with an antibiotic resistance cassette) leads to a marked derepression of lacZ gene fusions to the two genes. Replacing the endogenous hfqSTM gene with B. subtilis hfqBs does not correct the regulatory defect of chiP and yifK fusions to a significant extent (Fig 4B and 4C), indicating that HfqBs cannot substitute for the endogenous protein in ChiX- and GcvB-mediated regulation.
(A). S. enterica strains carrying hfqST-flag (MA11054) or hfqBs-flag (MA12275) translational fusion were grown in LB medium at 37°C under vigorous agitation. Harvested cells were lysed and crude extracts were used for western blot analysis using anti-FLAG M2 antibodies. ß-galactosidase activity was measured in exponentially growing LB cultures (OD600 ≈ 0.3) (B and C) or in early stationary phase cultures (OD600 ≈1.5) (D). Strains used were: B. MA9132, MA10744 and MA11214; C. MA8020, MA8021, and MA11215; D. MA9591, MA9603, and MA11216 (see Table 1 for full genotypes). (E) Cultures from strains in D were incubated 24 hours in stationary phase, serially diluted, and spotted on LB agar. hfqSTM (top row); ΔhfqSTM (middle row); ΔhfqSTM:: hfqBs (bottom row).
In enteric bacteria, the σE–driven envelope stress response is activated in hfq mutants due to the over-accumulation of several outer membrane proteins (OMP) [46–48]. The σE-controlled, Hfq-dependent MicA and RybB sRNAs are thought to be mainly responsible and ensure a negative feedback control [17,47,49]. We also tested the effects of the hfqSTM / hfqBs gene exchange on the regulation of degP, a member of the σE regulon that is chronically upregulated in an hfq defective background. The σE alteration is significantly alleviated in cells expressing HfqBs (activation ratio dropping from more than 20-fold to less than 8-fold; Fig 4D). In the course of these experiments, it became apparent that the Δhfq ΔdegP double mutant suffers a dramatic loss of viability in stationary phase, presumably reflecting the lack of DegP-mediated processing of the toxic products that are responsible for σE activation. We thus tested whether the strain with the replaced hfq showed a similar survival defect. Expression of hfqBs suppresses most of the stationary phase lethality, resulting in a 100-fold increase of viability in the double mutant (Fig 4E). Thus, it appears that HfqBs can, at least partially, perform the function(s) that avoid(s) the gratuitous induction of the σE-dependent envelope stress response (see Discussion).
The absence of hfqBs i) did not affect the strain growth rate and yield in nearly two thousand tested conditions and, ii) had no effect on the transcriptome of B. subtilis growing exponentially in a rich medium despite the fact that numerous small non-coding RNAs were expressed during the conditions of these experiments. Similar studies performed in close species led to the same results, namely no growth defect for S. aureus in more than a thousand tested conditions  and no transcriptome variations for L. monocytogenes . Nevertheless, Hfq is required in sustaining B. subtilis optimal survival upon starvation. To try dissecting this phenotype at the molecular level, we initially focused on the 171 RNA fragments reported to co-immunoprecipitate (CoIP) with Hfq . Surprisingly, only seven mRNAs and two antisense RNAs associated with Hfq were found affected by the hfq deletion in our study. Additional transcriptomic data from a B. subtilis hfq deleted strain became available while our study was under way . In this case only five mRNAs found altered in the mutant  were in common with the RNAs identified by CoIP . Note that only two of these RNAs (ctaD and yebD) correspond to those identified in our study (S1 Table). While the latter discrepancy could be ascribed to differences in the growth conditions, the lack of correlation between the HfqBs CoIP and the two independent HfqBs transcriptomic data suggests that HfqBs associates with RNA molecules but has little effect on their stability. Hfq may have only a limited number of specific partners or may influence regulatory RNA network by translational regulation.
The comparison between the transcriptomic data from cells grown either in rich medium (this study) or minimal CS-glucose medium  shows extensive differences. On the one hand, no sporulation-linked RNAs were affected by the Δhfq deletion in minimal medium. On the other hand, ResD/Rex (adaptation of anaerobiosis), GerE (germination) and ComA (competence) regulons, reportedly activated in the hfq mutant in minimal medium , were unaffected in our study. Only seven protein-coding genes were found in common but all of them, except one being affected in the opposite ways. Despite these discrepancies, both studies showed a decreased fitness of the hfq mutant in stationary phase regardless the medium used. We tested by competition experiments the possible involvement of yebD or ctaD (as their corresponding mRNAs were also found associated with Hfq ) and concluded these two genes are not associated with the hfq survival phenotype (S7 File). To date, the molecular mechanism responsible for this phenotype remains unidentified.
Hammerle et al proposed that the hfq-dependent phenotype was related to competence or sporulation . However, we showed that the hfq deletion did not affect competence efficiency and that the survival advantage conferred by Hfq remained in a sporulation deficient strain (Figs 2 and 3). As several Type-I TA systems had been reported to have their expression level modified in the absence of Hfq or were coIP with Hfq (this study and refs. [27,28]), one can speculate the existence of a functional link that could explain the survival advantage phenotype. We showed here that the fitness conferred by Hfq was independent of the presence of several Type-I TA systems (Fig 3). Altogether, these data suggest that as observed for HfqLm and HfqSa, HfqBs has a minor or no role on RNA stability in almost all conditions and are likely required only for starvation adaptation of vegetative cells.
hfq-ortholog genes are found in most sequenced proteobacteria and their role is well established in some of them. The Firmicutes division includes the Bacilli and Clostridia classes and within the last class, Clostridium difficile has an Hfq protein recently shown to be involved in sRNA mediated-regulations and sporulation gene expression [50,51]. In contrast, within the Bacilli, members of the Bacillales class (e.g., B. subtilis, L. monocytogenes, S. aureus) have a conserved hfq gene with an unclear function, while the Lactobacillales do not contain an hfq homolog. The selective pressure for hfq maintenance within the Bacillales is not strict, possibly because Hfq had a limited number of targets. In agreement with this proposal, the sRNA-dependent regulations reported within the Bacilli class are usually Hfq-independent, with the noticeable exception of LhrA in L. monocytogenes .
The only evidence for a molecular phenotype associated with HfqBs is its requirement for the mRNA ahrC translation [29,30]. Despite that the mRNA ahrC is the target of SR1 sRNA, the Hfq effect is SR1 independent. HfqBs acts as a specific post-transcriptional regulator, possibly by refolding the mRNA to promote its translation. As AhrC controls arginine metabolism, activating the catabolism pathway when arginine is available, we hypothesized that a lack of translation could explain the survival defect of the hfq mutant upon starvation. A translational fusion ahrC::spa was constructed at the locus in wild-type and Δhfq strains and the quantity of AhrC was measured by Western-blot in cells grown in LB rich medium. No difference of AhrC protein level was observed in exponential phase and we failed to detect the protein in stationary phase (S8 File). A strong accumulation of SR1, which represses ahrC expression, has been reported in stationary phase  and it probably explains the observed strong repression of ahrC. These data do not support a major role of AhrC in the survival defect observed in the absence of Hfq upon starvation.
As an alternative way for probing the biological activity of B. subtilis Hfq, we examined whether, and to what extent, it corrected some of the phenotypes resulting from Hfq inactivation in Salmonella enterica. We found HfqBs not to complement the regulatory defect of chiPQ and yifK loci to a relevant extent, suggesting that B. subtilis Hfq cannot interact with, or support the activity of, neither ChiX nor GcvB sRNAs. A third hfq mutant phenotype tested was the chronic activation of the Salmonella σE response. Interestingly, the expression of hfqBs significantly reduced the levels of Salmonella σE induction. At the same time, HfqBs suppressed the acute loss of viability suffered by a degP mutant lacking endogenous Hfq in stationary phase. HfqBs may affect Hfq-dependent sRNAs responsible for a negative feedback regulation, or additional uncharacterized components involved in the regulation . Alternatively Hfq could down-regulate OMP production directly through a sRNA-independent mechanism. The latter possibility might be relevant in the framework of the present study, as it would account for the observation that B. subtilis Hfq lacks the capacity of mediating sRNA activity, yet partially complements the σE-related phenotype of a Salmonella hfq mutant.
The hfqBs gene is found in the 49 sequenced genomes of the Bacillus genus (S9 File) indicating a strong requirement for its maintenance. HfqBs contribution to stationary phase survival gives a rational for its conservation within the specie. Interestingly, HfqBs accumulates in stationary phase and the only hfqBs-dependent previously reported phenotype was also observed in stationary phase . HfqBs clearly does not play the wide role attributed to its enteric bacteria counterparts. However, adaptation of B. subtilis to stationary phase represents a crucial process for this soil bacterium which encounters important and repeated variations in nutrient availability altering feast and starvation. In a competitive natural environment, HfqBs might play an essential role in promoting survival, possibly via stationary phase-dependent translational regulation.
Material and Methods
The B. subtilis strains BSB1 ΔhfqBs::cat (TR223) and ΔhfqBs::spc (TR232) were constructed by homologous replacement of the HfqBs coding sequence with a chloramphenicol (cat) or spectinomycin (spc) resistance gene, respectively, using a joining PCR technique . The cat and spc genes were amplified using 1151/1152 primers with pHV1610IR(-) and pIC156 as templates, respectively [53,54]. DNA fragments corresponding to the upstream and downstream hfq sequences were PCR amplified using 1147/1148 and 1149/1150 primers, respectively, with BSB1 chromosomal DNA as a template. DNA fragments were purified using NucleoSpin Gel clean-up (Macherey Nagel) and then joined by a second PCR using 1147 and 1150 primers. The hfq deletion mutant was obtained by transformation of the BSB1 strain with the joining PCR fragment and selection for chloramphenicol (5 μg/ml) or spectinomycin (100 μg/ml) resistance. The same procedure was used to construct a control strain carrying the cat or spc gene inserted into the intergenic region between hfq and ymzE genes. Briefly, DNA fragments upstream and downstream the insertion position were amplified using 1186/1187 and 1188/1154 primers, respectively, and then were joined with the cat or spc genes by a second PCR using 1186/1154 primers. Transformation of the BSB1 strain resulted in TR247 (Cmr) and TR241 (Spcr) strains.
The BSB1 sporulation-deficient strain was obtained by replacement of the sigE coding sequence by an erythromycin (erm) resistance gene amplified using 1151/1152 primers with pMUTIN as a template . Upstream and downstream DNA fragments were amplified using 1182/1183 and 1184/1185 primers, respectively. The joining PCR fragment was amplified using 1182/1185 primers and transformed into BSB1 strain resulting in BSB1 ΔsigE::erm (TR234). TR234 strain was transformed with chromosomal extracts of ΔhfqBs::cat, ΔhfqBs::spc, igr::spc or igr::cat resulting in the corresponding sporulation-deficient derivative strains TR235, TR237, TR243 and TR245, respectively. The ΔhfqBs::spc or igr::cat were also introduced by transformation with chromosomal extracts in four other genetic backgrounds: i) the TF8A strain (deleted of skin, PBSX and SP-β prophages, , resulting in TR248 and TR252 strains respectively; ii) BSB1-derivative strain deleted of the bsrE-asBsrE type 1 TA system, resulting in TR255 and TR259 strains; iii) the JJS-Din048 strain, a TF8A-derivative strain deleted of the genomic region from pycA to ctaG genes  resulting in EDJ1125 and EDJ1129 strains, and iv) a BSB1-derivative strain deleted of yebD gene, resulting in TR253 and TR257 strains (see Table 1). The yebD and bsrE-asBsrE mutants were obtained by gene replacement with erm gene amplified using TRO84/TRO85 primers. Upstream and downstream DNA fragments were amplified using TRO80/TRO81 and TRO82/TRO83 primers, respectively for yebD deletion, or using TRO86/TRO87 and TRO88/TRO89 primers, respectively for bsrE deletion. The joining PCR fragments were amplified using TRO80/TRO83 or TRO86/TRO89 primers for yebD::erm and bsrE::erm fragments respectively, and transformed into BSB1 strain.
A BSB1 strain expressing a C-terminal SPA-tagged HfqBs protein was constructed as previously described  by chromosomal integration of a translational fusion between the hfq coding sequence and the sequential peptide affinity (SPA) tag sequence  resulting in the TR229 strain expressing HfqBsSPA under the control of its native promoter. The pMUTIN-SPA plasmid was used as in ref. . A PCR amplification encompassing almost the entire hfq gene—except the ATG start codon—was obtained using the ODfw/ODrev primers. After digestion by Acc65I and NcoI the DNA fragment was ligated to the pMUTIN-SPA that was linearized by the same restriction enzymes. After establishment in E.coli, the plasmid was recovered by using a Plasmid purification kit (Qiagen) and used to transform competent B. subtilis BSB1 cells. Integration of the plasmid in the chromosome at the hfq locus was selected by plating on LB supplemented with 30 μg erythromycin and 0.5 mM IPTG. Correct insertion was checked by sequencing. The same procedure was used to construct ahrC::spa translational fusion at the locus resulting in strain Bas028 (Table 1). Bas028 was then transformed by genomic DNA from TRO223 (Δhfq::cat) and from TRO232 (Δhfq::spc) to produce strains Bas101 and Bas102 respectively.
Salmonella enterica serovar Typhimurium strains used in this study are listed on Table 1. They are all derived from MA3409, a derivative of strain LT2 cured for the Gifsy-1 prophage . Salmonella strains carrying the structural portion of the hfq gene from B. subtilis were constructed with a two-step recombineering procedure as described . Strain MA10740, carrying a tetAR module inserted in the hfq gene of the Salmonella chromosome was used with a DNA fragment amplified from chromosomal DNA of B. subtilis BSB1 with primer pairs 1174/1175, the entire hfq::tetAR was crossed out selecting for the loss of tetracycline resistance.
Introduction of the 3xFLAG epitope at the 3’ ends of the coding sequence of hfqBs was carried out using DNA fragments amplified from plasmid pSUB11, as described .
Generalized transduction was carried out using the high frequency transducing mutant of phage P22, HT 105/1 int-201 . “λ Red”-mediated chromosomal recombineering was carried out as described . Constructs were verified by PCR and DNA sequence.
Bacterial growth conditions
Bacteria were cultured at 37°C under vigorous agitation in liquid media or in media solidified by the addition of 1.5% (w/v) Difco agar. LB broth [1% bacto tryptone (w/v), 0.5% Difco yeast extract (w/v), 0.5% NaCl (w/v)] was used as complex medium. When needed, LB medium was supplemented with 0.2% (w/v) arabinose and 0.5 μg/ml IPTG. When needed, Antibiotics were included at the following final concentrations: chloramphenicol, 10 μg/ml for S. Typhimurium and 5 μg/ml for B. subtilis; kanamycin monosulphate, 50 μg/ml; sodium ampicillin 100 μg/ml; tetracycline hydrochloride, 25 μg/ml; erythromycin, 0.6 μg/ml; spectinomycin, and 100 μg/ml.
Competition experiments were performed as follow: overnight cultures of hfqBs+ and ΔhfqBs strains were diluted in same flask containing 50 mL of medium without antibiotic with pairwise assembled as follows: TR241 (igr::spc) with TR223 (ΔhfqBs::cat), TR247 (igr::cat) with TR232 (ΔhfqBs::spc), TR245 (ΔsigE::erm igr::spc) with TR235 (ΔsigE::erm ΔhfqBs::cat), TR243 (ΔsigE::erm igr::cat) with TR237 (ΔsigE::erm ΔhfqBs::spc), TR252 with TR248, EDJ1125 with EDJ1129, TR259 with TR255, and TR257 with TR253. For each co-culture, viable cells of the two populations were numbered by plating serial dilutions on plates supplemented with either chloramphenicol or spectinomycin. For spore numbering, samples were heated at 80°C during 15 min before plating.
Biofilm formation was analyzed as previously described . Cells were grown overnight in LB medium and biofilm formation was tested in MSgg medium (5 mM potassium phosphate [pH 7], 100 mM morpholinepropane sulfonic acid [pH 7], 2 mM MgCl2, 700 μM CaCl2, 50 μM MnCl2, 50 μM FeCl3, 1 μM ZnCl2, 2 μM thiamine, 0.5% glycerol, 0.5% glutamate, 50 μg of tryptophan/ml, 50 μg of phenylalanine/ml). For that purpose, each overnight culture was diluted 200 times to inoculate either 1 ml of MSgg within a well of a 48-wells microtiter plate or 20 ml of MSgg in 100ml glass bottles. Cultures were incubated without shaking at 30°C, and their pellicles were analyzed by visual inspection during 72 h.
Competence development time course was analyzed as follows: overnight cultures in LB medium supplemented with 50 μg of tryptophan/ml and 3 mM MgSO4 were diluted in the same medium and grown at 37°C under vigorous agitation until O.D. 600nm around 1. These cultures were diluted in 10 ml of MD medium (10.7 g/l K2HPO4, 6 g/l KH2PO4, 1g/l trisodium citrate, 2% glucose, 0.1% tryptophan, 11 ug/ml ferric ammonium citrate, 3 mM MgSO4, 0.25 mg/ml potassium aspartate) to begin the time course competence assay. At different time, 1 ug of chromosomal DNA amyE::KnR (around 2 108 molecules) were added to 0.2 ml of cells, incubated 20 minutes at 37°C and then spread for transformants selection on LB plates supplemented by kanamycine 6 ug/ml.
HfqBs-3xFlag and HfqSTM-3xFlag were detected in Salmonella enterica strains as previously described . HfqSPA and AhrCSPA were detected as following: Cells grown in LB were taken at the end of the exponential phase and in the early stationary phase of growth for AhrCSPA, and in exponential, early and late transition and stationary phase (OD600nm 1, 2.4, 3 and 4, respectively) for HfqSPA. After disruption by sonication in 500 μL of 10 mM Tris-Cl pH 7.5, 150 mM NaCl and 1 mM EDTA, samples were centrifuged at 18000g for 30 min at 4°C. The protein concentrations of the samples were measured by the Bradford method and used to load equal quantities of sample proteins on a SDS-PAGE. Proteins were then transferred to a Amersham Hybond-P membrane. HfqSPA and AhrCSPA were detected by western blot using anti-FLAG M2 antibodies as primary antibody and an anti-mouse IgG-peroxidase antibody (SIGMA A2304) as the second antibody. ECL kit was used to enable the immunodetection and the signal was recorded by using Chemidoc from BioRad.
RNA preparation and transcriptome analysis
Total RNA was isolated as previously described  except for cells lysis. Briefly, overnight cultures were diluted 2000 times in preheated LB medium and incubated at 37°C under shaking agitation (180 rpm) until OD600nm reached 0.5 (exponential growth phase transcriptome) or 5 hours after the onset of stationary phase (stationary phase transcriptome). Cultures were centrifuged; the pellets were frozen in liquid nitrogen and stored at -80°C until RNA extraction. Cells pellets were resuspended into 400 μL of Lysis buffer (4 M guanidine thiocyanate, 25 mM sodium acetate pH 5.2, 5 g/L N-laurylsarcosinate), transferred into FastPrep tubes containing 0.6 g of glass beads (G4649, Sigma) and 400 μL of acid phenol:chloroform:IAA (25:24:1). Bacteria were mechanically lysed by using the Fastprep apparatus (MP Biomedicals) with 3 cycles of 45 s at speed 6.0 separated by incubation on ice during 5 min. After lysis, tubes were centrifuged 15 min at 17,900 g at 4°C. The aqueous phase was acid phenol extracted and isopropanol precipitated as previously described . 40 μg RNA were treated using the Turbo DNase I (Ambion) and purified using the RNA Clean-Up and Concentration Micro Kit (Norgen). The quality of the RNA preparations was assessed using RNA Nano Chip with an Agilent 2100 Bioanalyzer (Agilent Technologies). Synthesis of Cy3-labeled DNA from the RNA samples with random priming, one-color hybridization on tiling arrays and signal acquisition were carried out as previously described . An aggregated expression value was computed for all transcripts according to our recently published B. subtilis 168 structural annotation which contains 1583 defined transcribed regions in addition to the 4292 previously annotated coding sequences . Expression values were quantile-normalized between experiments and a differential analysis was performed using the Limma R package . P-values were corrected for multiple testing using the Bonferroni-Holm method. Genes with adjusted p-value <0.05 were considered differentially expressed in the Δhfq mutant relative to wild-type.
The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus  and are accessible through GEO Series accession number GSE66893 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE66893).
Phenotype microarray analysis
A full array phenotype microarray analysis was performed by the Biolog Company (Hayward, CA, USA) according to their standard procedure. The principle is to compare isogenic pairs of strains (here BSB1 and TR223) for their growth on wells of microtiter plates (panels), each well containing a different growth medium. This high-throughput technology allows the testing of a large number of phenotypes (eight metabolic-array panels and twelve sensitivity-array panels). The experiment was run in duplicate (strains/conditions) and pairwise comparison was created to analyze the results. All putative phenotypes were independently verified using disk diffusion assays as follows: exponentially growing cultures of wild-type and ΔhfqBs strains were diluted 50 times in 10 ml of LB and used to inoculate three LB plates for each strain. The diameters of growth inhibition around disks containing vibriostatic agent 0.129 (Biorad #53872 0.5 mg), 5 μl of cetylpyridinium chloride 10% (Acros organics) or 5 μl of 48/80 compound (5 mg/ml, Sigma) were measured after 24 h of incubation at 37°C.
Activity of ß-galactosidase was measured in toluene-permeabilized cells as described  and is expressed in Miller units. Reported values were the average of at least two independent determinations, each involving duplicate or triplicate samples.
S2 File. BSB1 and BSB1 ΔhfqBs phenotype MicroArrays analysis.
S3 File. Biofilm formation of BSB1 and BSB1 ΔhfqBs.
S4 File. BSB1 and BSB1 ΔhfqBs transcriptome in exponential growth phase: Scatter plot and selected windows (hfqBs mRNA and sRNAs).
S5 File. BSB1 and BSB1 ΔhfqBs transcriptome in exponential growth phase.
Genome-wide representation of transcriptome profiles of wild-type and Δhfq strains in exponential phase of growth (Figure legend in S4 File).
S6 File. BSB1 and BSB1 ΔhfqBs transcriptome in stationary phase.
Genome-wide representation of transcriptome profiles of wild-type and Δhfq strains in stationary phase of growth (Figure legend in S4 File).
S7 File. BSB1 ΔhfqBs survival in competition with hfq-expressing cells in the absence of yebD or ctaD.
S8 File. ahrC expression in hfqBs-expressing strain and ΔhfqBs mutant.
S9 File. In silico analysis of the hfqBs gene conservation and synteny among the Bacillus genus.
S1 Table. Transcriptome quantile-normalized expression values of BSB1 and BSB1 ΔhfqBs and list of genes differentially expressed in stationary phase.
We thank Anne Aucouturier for strain constructions. We thank Chantal Bohn, Annick Jacq and R. MacGregor for technical assistance, helpful discussions, and warm support. We thank Cyprien Guérin for his help with tiling arrays signal processing, and Pierre Nicolas and Ulrike Mäder for the transition from Nimblegen-based BaSysBio Bsub T2 design to new BaSysBio Bsub T3 design using Agilent microarrays.
Conceived and designed the experiments: TR OD NFB PN LB ED PB. Performed the experiments: TR OD NFB LB ED. Analyzed the data: TR OD NFB LB ED PB. Contributed reagents/materials/analysis tools: LB PN PB. Wrote the paper: TR OD NFB LB ED PB.
- 1. Brennan RG, Link TM (2007) Hfq structure, function and ligand binding. Curr Opin Microbiol 10: 125–133. pmid:17395525
- 2. Valentin-Hansen P, Eriksen M, Udesen C (2004) The bacterial Sm-like protein Hfq: a key player in RNA transactions. Mol Microbiol 51: 1525–1533. pmid:15009882
- 3. Vogel J, Luisi BF (2011) Hfq and its constellation of RNA. Nat Rev Microbiol 9: 578–589. pmid:21760622
- 4. Franze de Fernandez MT, Hayward WS, August JT (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. pmid:4550762
- 5. Tsui HC, Leung HC, Winkler ME (1994) Characterization of broadly pleiotropic phenotypes caused by an hfq insertion mutation in Escherichia coli K-12. Mol Microbiol 13: 35–49. pmid:7984093
- 6. Storz G, Vogel J, Wassarman KM (2011) Regulation by small RNAs in bacteria: expanding frontiers. Mol Cell 43: 880–891. pmid:21925377
- 7. Aiba H (2007) Mechanism of RNA silencing by Hfq-binding small RNAs. Curr Opin Microbiol 10: 134–139. pmid:17383928
- 8. Bossi L, Maloriol D, Figueroa-Bossi N (2008) Porin biogenesis activates the sigma(E) response in Salmonella hfq mutants. Biochimie 90: 1539–1544. pmid:18585433
- 9. Sun X, Zhulin I, Wartell RM (2002) Predicted structure and phyletic distribution of the RNA-binding protein Hfq. Nucleic Acids Res 30: 3662–3671. pmid:12202750
- 10. Schumacher MA, Pearson RF, Moller T, Valentin-Hansen P, Brennan RG (2002) Structures of the pleiotropic translational regulator Hfq and an Hfq-RNA complex: a bacterial Sm-like protein. EMBO J 21: 3546–3556. pmid:12093755
- 11. Bohn C, Rigoulay C, Bouloc P (2007) No detectable effect of RNA-binding protein Hfq absence in Staphylococcus aureus. BMC Microbiol 7: 10. pmid:17291347
- 12. Geisinger E, Adhikari RP, Jin R, Ross HF, Novick RP (2006) Inhibition of rot translation by RNAIII, a key feature of agr function. Mol Microbiol 61: 1038–1048. pmid:16879652
- 13. Liu Y, Wu N, Dong J, Gao Y, Zhang X, Mu C, et al. (2010) Hfq is a global regulator that controls the pathogenicity of Staphylococcus aureus. PLoS One 5.
- 14. Boisset S, Geissmann T, Huntzinger E, Fechter P, Bendridi N, Possedko M, et al. (2007) Staphylococcus aureus RNAIII coordinately represses the synthesis of virulence factors and the transcription regulator Rot by an antisense mechanism. Genes Dev 21: 1353–1366. pmid:17545468
- 15. Geissmann T, Chevalier C, Cros MJ, Boisset S, Fechter P, Noirot C, et al. (2009) A search for small noncoding RNAs in Staphylococcus aureus reveals a conserved sequence motif for regulation. Nucleic Acids Res 37: 7239–7257. pmid:19786493
- 16. Chabelskaya S, Gaillot O, Felden B (2010) A Staphylococcus aureus small RNA is required for bacterial virulence and regulates the expression of an immune-evasion molecule. PLoS Pathog 6: e1000927. pmid:20532214
- 17. Rochat T, Bouloc P, Yang Q, Bossi L, Figueroa-Bossi N (2012) Lack of interchangeability of Hfq-like proteins. Biochimie 94: 1554–1559. pmid:22326874
- 18. Christiansen JK, Larsen MH, Ingmer H, Sogaard-Andersen L, Kallipolitis BH (2004) The RNA-binding protein Hfq of Listeria monocytogenes: role in stress tolerance and virulence. J Bacteriol 186: 3355–3362. pmid:15150220
- 19. Christiansen JK, Nielsen JS, Ebersbach T, Valentin-Hansen P, Sogaard-Andersen L, Kallipolitis BH (2006) Identification of small Hfq-binding RNAs in Listeria monocytogenes. RNA 12: 1383–1396. pmid:16682563
- 20. Nielsen JS, Lei LK, Ebersbach T, Olsen AS, Klitgaard JK, Valentin-Hansen P, et al. (2010) Defining a role for Hfq in Gram-positive bacteria: evidence for Hfq-dependent antisense regulation in Listeria monocytogenes. Nucleic Acids Res 38: 907–919. pmid:19942685
- 21. Nielsen JS, Larsen MH, Lillebaek EM, Bergholz TM, Christiansen MH, Boor KJ, et al. (2011) A small RNA controls expression of the chitinase ChiA in Listeria monocytogenes. PLoS One 6: e19019. pmid:21533114
- 22. Mandin P, Repoila F, Vergassola M, Geissmann T, Cossart P (2007) Identification of new noncoding RNAs in Listeria monocytogenes and prediction of mRNA targets. Nucleic Acids Res 35: 962–974. pmid:17259222
- 23. Toledo-Arana A, Dussurget O, Nikitas G, Sesto N, Guet-Revillet H, Balestrino D, et al. (2009) The Listeria transcriptional landscape from saprophytism to virulence. Nature 459: 950–956. pmid:19448609
- 24. Kovach AR, Hoff KE, Canty JT, Orans J, Brennan RG (2014) Recognition of U-rich RNA by Hfq from the Gram-positive pathogen Listeria monocytogenes. RNA 20: 1548–1559. pmid:25150227
- 25. Robinson KE, Orans J, Kovach AR, Link TM, Brennan RG (2014) Mapping Hfq-RNA interaction surfaces using tryptophan fluorescence quenching. Nucleic Acids Res 42: 2736–2749. pmid:24288369
- 26. Silvaggi JM, Perkins JB, Losick R (2005) Small untranslated RNA antitoxin in Bacillus subtilis. J Bacteriol 187: 6641–6650. pmid:16166525
- 27. Hammerle H, Amman F, Vecerek B, Stulke J, Hofacker I, Blasi U (2014) Impact of Hfq on the Bacillus subtilis transcriptome. PLoS One 9: e98661. pmid:24932523
- 28. Dambach M, Irnov I, Winkler WC (2013) Association of RNAs with Bacillus subtilis Hfq. PLoS One 8.
- 29. Heidrich N, Chinali A, Gerth U, Brantl S (2006) The small untranslated RNA SR1 from the Bacillus subtilis genome is involved in the regulation of arginine catabolism. Mol Microbiol 62: 520–536. pmid:17020585
- 30. Heidrich N, Moll I, Brantl S (2007) In vitro analysis of the interaction between the small RNA SR1 and its primary target ahrC mRNA. Nucleic Acids Res 35: 4331–4346. pmid:17576690
- 31. Gaballa A, Antelmann H, Aguilar C, Khakh SK, Song KB, Smaldone GT, et al. (2008) The Bacillus subtilis iron-sparing response is mediated by a Fur-regulated small RNA and three small, basic proteins. Proc Natl Acad Sci U S A 105: 11927–11932. pmid:18697947
- 32. Durand S, Braun F, Lioliou E, Romilly C, Helfer AC, Kuhn L, et al. (2015) A Nitric Oxide Regulated Small RNA Controls Expression of Genes Involved in Redox Homeostasis in Bacillus subtilis. PLoS Genet 11: e1004957. pmid:25643072
- 33. Britton RA, Eichenberger P, Gonzalez-Pastor JE, Fawcett P, Monson R, Losick R, et al. (2002) Genome-wide analysis of the stationary-phase sigma factor (sigma-H) regulon of Bacillus subtilis. J Bacteriol 184: 4881–4890. pmid:12169614
- 34. Nicolas P, Mader U, Dervyn E, Rochat T, Leduc A, Pigeonneau N, et al. (2012) Condition-dependent transcriptome reveals high-level regulatory architecture in Bacillus subtilis. Science 335: 1103–1106. pmid:22383849
- 35. Bochner BR (2009) Global phenotypic characterization of bacteria. FEMS Microbiol Rev 33: 191–205. pmid:19054113
- 36. Phillips ZE, Strauch MA (2002) Bacillus subtilis sporulation and stationary phase gene expression. Cell Mol Life Sci 59: 392–402. pmid:11964117
- 37. Branda SS, Gonzalez-Pastor JE, Ben-Yehuda S, Losick R, Kolter R (2001) Fruiting body formation by Bacillus subtilis. Proc Natl Acad Sci U S A 98: 11621–11626. pmid:11572999
- 38. Rasmussen S, Nielsen HB, Jarmer H (2009) The transcriptionally active regions in the genome of Bacillus subtilis. Mol Microbiol 73: 1043–1057. pmid:19682248
- 39. Irnov I, Sharma CM, Vogel J, Winkler WC (2010) Identification of regulatory RNAs in Bacillus subtilis. Nucleic Acids Res 38: 6637–6651. pmid:20525796
- 40. Bohn C, Rigoulay C, Chabelskaya S, Sharma CM, Marchais A, Skorski P, et al. (2010) Experimental discovery of small RNAs in Staphylococcus aureus reveals a riboregulator of central metabolism. Nucleic Acids Res 38: 6620–6636. pmid:20511587
- 41. Fujita M, Gonzalez-Pastor JE, Losick R (2005) High- and low-threshold genes in the Spo0A regulon of Bacillus subtilis. J Bacteriol 187: 1357–1368. pmid:15687200
- 42. Schmalisch M, Maiques E, Nikolov L, Camp AH, Chevreux B, Muffler A, et al. (2010) Small genes under sporulation control in the Bacillus subtilis genome. J Bacteriol 192: 5402–5412. pmid:20709900
- 43. Gonzalez-Pastor JE (2011) Cannibalism: a social behavior in sporulating Bacillus subtilis. FEMS Microbiol Rev 35: 415–424. pmid:20955377
- 44. Figueroa-Bossi N, Valentini M, Malleret L, Fiorini F, Bossi L (2009) Caught at its own game: regulatory small RNA inactivated by an inducible transcript mimicking its target. Genes Dev 23: 2004–2015. pmid:19638370
- 45. Sharma CM, Papenfort K, Pernitzsch SR, Mollenkopf HJ, Hinton JC, Vogel J. (2011) Pervasive post-transcriptional control of genes involved in amino acid metabolism by the Hfq-dependent GcvB small RNA. Mol Microbiol 81: 1144–1165. pmid:21696468
- 46. Douchin V, Bohn C, Bouloc P (2006) Down-regulation of porins by a small RNA bypasses the essentiality of the regulated intramembrane proteolysis protease RseP in Escherichia coli. J Biol Chem 281: 12253–12259. pmid:16513633
- 47. Figueroa-Bossi N, Lemire S, Maloriol D, Balbontin R, Casadesus J, Bossi L (2006) Loss of Hfq activates the sigmaE-dependent envelope stress response in Salmonella enterica. Mol Microbiol 62: 838–852. pmid:16999834
- 48. Guisbert E, Rhodius VA, Ahuja N, Witkin E, Gross CA (2007) Hfq modulates the sigmaE-mediated envelope stress response and the sigma32-mediated cytoplasmic stress response in Escherichia coli. J Bacteriol 189: 1963–1973. pmid:17158661
- 49. Papenfort K, Pfeiffer V, Mika F, Lucchini S, Hinton JC, Vogel J (2006) SigmaE-dependent small RNAs of Salmonella respond to membrane stress by accelerating global omp mRNA decay. Mol Microbiol 62: 1674–1688. pmid:17427289
- 50. Boudry P, Gracia C, Monot M, Caillet J, Saujet L, Hajnsdorf E, et al. (2014) Pleiotropic role of the RNA chaperone protein Hfq in the human pathogen Clostridium difficile. J Bacteriol 196: 3234–3248. pmid:24982306
- 51. Caillet J, Gracia C, Fontaine F, Hajnsdorf E (2014) Clostridium difficile Hfq can replace Escherichia coli Hfq for most of its function. RNA 20: 1567–1578. pmid:25147238
- 52. Vagner V, Dervyn E, Ehrlich SD (1998) A vector for systematic gene inactivation in Bacillus subtilis. Microbiology 144 (Pt 11): 3097–3104. pmid:9846745
- 53. Chedin F, Dervyn E, Dervyn R, Ehrlich SD, Noirot P (1994) Frequency of deletion formation decreases exponentially with distance between short direct repeats. Mol Microbiol 12: 561–569. pmid:7934879
- 54. Steinmetz M, Richter R (1994) Plasmids designed to alter the antibiotic resistance expressed by insertion mutations in Bacillus subtilis, through in vivo recombination. Gene 142: 79–83. pmid:8181761
- 55. Westers H, Dorenbos R, van Dijl JM, Kabel J, Flanagan T, Devine KM, et al. (2003) Genome engineering reveals large dispensable regions in Bacillus subtilis. Mol Biol Evol 20: 2076–2090. pmid:12949151
- 56. Tanaka K, Henry CS, Zinner JF, Jolivet E, Cohoon MP, Xia F, et al. (2013) Building the repertoire of dispensable chromosome regions in Bacillus subtilis entails major refinement of cognate large-scale metabolic model. Nucleic Acids Res 41: 687–699. pmid:23109554
- 57. Rochat T, Nicolas P, Delumeau O, Rabatinova A, Korelusova J, Leduc A, et al. (2012) Genome-wide identification of genes directly regulated by the pleiotropic transcription factor Spx in Bacillus subtilis. Nucleic Acids Res.
- 58. Zeghouf M, Li J, Butland G, Borkowska A, Canadien V, Richards D, et al. (2004) Sequential Peptide Affinity (SPA) system for the identification of mammalian and bacterial protein complexes. J Proteome Res 3: 463–468. pmid:15253427
- 59. Lecointe F, Serena C, Velten M, Costes A, McGovern S, Meile JC, et al. (2007) Anticipating chromosomal replication fork arrest: SSB targets repair DNA helicases to active forks. EMBO J 26: 4239–4251. pmid:17853894
- 60. Figueroa-Bossi N, Coissac E, Netter P, Bossi L (1997) Unsuspected prophage-like elements in Salmonella typhimurium. Mol Microbiol 25: 161–173. pmid:11902718
- 61. Karlinsey JE (2007) λ-Red genetic engineering in Salmonella enterica serovar Typhimurium. Methods Enzymol: 199–209. pmid:17998056
- 62. Uzzau S, Figueroa-Bossi N, Rubino S, Bossi L (2001) Epitope tagging of chromosomal genes in Salmonella. Proc Natl Acad Sci U S A 98: 15264–15269. pmid:11742086
- 63. Bossi L, Figueroa-Bossi N (2007) A small RNA downregulates LamB maltoporin in Salmonella. Mol Microbiol 65: 799–810. pmid:17608792
- 64. Marincola G, Schafer T, Behler J, Bernhardt J, Ohlsen K, Goerke C, et al. (2012) RNase Y of Staphylococcus aureus and its role in the activation of virulence genes. Mol Microbiol 85: 817–832. pmid:22780584
- 65. Branda SS, Gonzalez-Pastor JE, Dervyn E, Ehrlich SD, Losick R, Kolter R (2004) Genes involved in formation of structured multicellular communities by Bacillus subtilis. J Bacteriol 186: 3970–3979. pmid:15175311
- 66. Smyth GK (2004) Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol 3: Article3. pmid:16646809
- 67. Edgar R, Domrachev M, Lash AE (2002) Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res 30: 207–210. pmid:11752295
- 68. Miller JH (1972) Experiments in molecular genetics. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory. xvi, 466 p. p.