Hfq Is a Regulator of F-Plasmid TraJ and TraM Synthesis in Escherichia coli
http://www.100md.com
细菌学杂志 2006年第1期
Department of Biological Sciences, University of Alberta, Edmonton, Alberta, T6G 2E9, Canada
ABSTRACT
The F plasmid of Escherichia coli allows horizontal DNA transfer between an F+ donor cell and an F– recipient. Expression of the transfer genes is tightly controlled by a number of factors, including the following plasmid-encoded regulatory proteins: TraJ, the primary activator of the 33-kb tra operon, and the autoregulators TraM and TraY. Here, we demonstrate that the host RNA binding protein, Hfq, represses TraJ and TraM synthesis by destabilizing their respective mRNAs. Mating assays and immunoblot analyses for TraM and TraJ showed that transfer efficiency and protein levels increased in host cells containing a disruption in hfq compared to wild-type cells in stationary phase. The stability of transcripts containing a putative Hfq binding site located in the intergenic untranslated region between traM and traJ was increased in hfq mutant donor cells, suggesting that Hfq destabilizes these transcripts. Electrophoretic mobility shift assays demonstrated that Hfq specifically binds this region but not the antisense RNA, FinP, encoded on the opposite strand. Together, these findings indicate that Hfq regulates traM and traJ transcript stability by a mechanism separate from FinOP-mediated repression.
INTRODUCTION
The F plasmid of Escherichia coli is considered a model for bacterial conjugation and has been studied extensively (15, 18). F-plasmid transfer is facilitated by a large multicomponent protein complex which spans the donor cell membrane to transport plasmid DNA into a recipient cell. This complex is encoded by a single 33-kb polycistronic transfer (tra) operon (11). Recent studies have demonstrated that the regulatory circuit controlling the expression of this tra operon is complex. However, the focal point of control appears to be TraJ. TraJ is the primary activator of tra operon expression, encoded on a monocistronic operon immediately upstream of the tra operon and downstream from traM (11, 38) (Fig. 1). It appears to act by opposing H-NS-mediated repression of the tra operon promoter PY (36a). The resulting derepression of PY allows the synthesis of the secondary regulator, TraY, encoded by the tra operon, which further activates PY, as well as traM (27). TraM is required to transmit the signal for the initiation of DNA transfer between the transfer complex and the relaxosome, which nicks and unwinds the plasmid DNA during transfer (7, 22). TraM also has a regulatory role as an autorepressor that represses expression from the two tandem traM promoters, referred to here collectively as PM (27).
In most F-like plasmids, TraJ synthesis is subject to posttranscriptional control via the FinOP antisense RNA system, whereby finP encodes a small antisense RNA complementary to the 5' untranslated region (UTR) of the traJ transcript (33). FinP is a short, 79-base RNA consisting of two stem-loop sequences, SL-I and SL-II (34) (Fig. 2). These are complementary to stem-loops present in the 5' UTR of traJ, referred to as SL-Ic and SL-IIc, respectively. A third stem-loop, SL-III, is also present in the 5' UTR of the traJ mRNA, but not in FinP, and is separated from SL-IIc by a 24-base AU-rich spacer region. These structural elements appear to be well conserved and are present in most F-like traJ transcripts (11). FinO, a plasmid-encoded RNA chaperone, binds FinP, protecting it from degradation by RNase E and promoting duplex formation between the complementary stem-loop structures in both FinP and traJ (4, 17). The resulting duplex sequesters the ribosome binding site (RBS) and prevents the translation of traJ, resulting in the repression of plasmid transfer (14, 19). However, the F plasmid is naturally derepressed, since its copy of finO has been disrupted by the insertion of an IS3 element (6).
In this study, we considered the possibility that host proteins might also affect traJ mRNA synthesis using a posttranscriptional mechanism similar to the FinOP system in F-like plasmids, with Hfq, the host RNA chaperone, being the most likely candidate. Originally characterized as host factor I (HF-I), a host-encoded protein necessary for the in vitro replication of the Q RNA bacteriophage (10), Hfq has emerged as a potent regulator of many aspects of RNA biology, influencing stability, translation, and RNA bacteriophage replication, often via small RNAs (13, 32). Hfq is an 11.2-kDa protein, forming hexamers that preferentially bind sequences of AU-rich RNA, often flanked by structured regions (25, 39). It is relatively abundant and is present at intracellular levels of approximately 10,000 hexamers per cell (2). However, the expression profile throughout the growth cycle remains uncertain, as some studies have indicated that Hfq levels peak and then decrease rapidly during lag phase, dropping to approximately one third of the intracellular maximum throughout the rest of the growth cycle (2). Other studies have found that Hfq levels increase as cultures enter stationary phase or during periods of slow growth (30, 35).
Given that TraJ synthesis is regulated by both an antisense RNA system, a common target of Hfq, and H-NS, which has been shown to overlap with Hfq in many regulatory circuits, including rpoS, hns, and bgl regulation (8, 16, 20, 28, 29), we considered the possibility that Hfq might target and regulate TraJ mRNA in some manner. In this study, we demonstrate that Hfq binds to the intergenic UTR, 3' to traM and 5' to traJ, and decreases the stability of transcripts containing this region. Hfq does not appear to be involved in fertility inhibition and has no role in FinOP-mediated repression. Instead, Hfq appears to act as a repressor of TraJ and TraM synthesis, as well as F-plasmid transfer in general, by destabilizing the corresponding transcripts.
MATERIALS AND METHODS
Experimental strains and growth conditions. All bacterial strains and plasmids used in this study are described in Table 1. Unless otherwise stated, all liquid cultures were grown in Luria-Bertani (LB) broth at 37°C with shaking. Solid medium cultures were grown on LB agar incubated at 37°C overnight. Antibiotics were added as indicated to selective media at the following concentrations: ampicillin, 25 μg/ml; kanamycin, 25 μg/ml; spectinomycin, 100 μg/ml; streptomycin, 200 μg/ml; and tetracycline, 10 μg/ml.
Mating assays. Standing overnight cultures of AM111 and AM112 containing the F-plasmid derivative pOX38-Tc were diluted 200-fold into fresh LB. These donor cultures were then incubated at 37°C with shaking. At the indicated time points, the optical density at 600 nm (OD600) was taken and donor culture samples were removed to assay for mating ability. Aliquots of both donor cells and ED24 recipient cells, each equivalent to an OD600 of 0.1, were mixed in a total volume of 1 ml of spent LB broth. Spent LB broth refers to broth which is removed at the same time point as the donor cells and centrifuged to remove any bacteria present. These mating reactions were incubated, standing, at 37°C for 1 h. Mating reactions were then vigorously vortexed for 10 to 20 seconds and placed on ice for approximately 5 min to halt any further plasmid transfer. This was followed by serially diluting the mating reactions in SSC buffer (150 mM NaCl, 15 mM sodium citrate, pH 7.0) and plating the diluted cells on LB agar medium containing antibiotics selective for either donors (streptomycin and tetracycline) or transconjugants (spectinomycin and tetracycline). Growth was scored, and mating efficiency was measured as transconjugants per donor. All mating data represent the average of quadruplicate experiments.
Immunoblot analysis. For growth phase-dependent immunoblot analysis, standing overnight cultures of AM112/pOX38-Tc and AM111/pOX38-Tc were grown in LB broth containing tetracycline and kanamycin at 37°C, then diluted 1:200 in fresh LB containing tetracycline, and grown with shaking. At the indicated time points, culture aliquots equivalent to an OD600 of 0.1 were removed, and the cells were pelleted and frozen. Samples were resuspended in sodium dodecyl sulfate-polyacrylamide gel electrophoresis sample loading buffer and then boiled for 5 min before electrophoresis in a 15% sodium dodecyl sulfate-polyacrylamide gel. Following electrophoresis, gels were transferred onto Immobilon-P membrane (Millipore) and blocked overnight at 4°C in blocking solution consisting of 10% skim milk dissolved in TBST (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, and 0.1% Tween 20). Blots were probed for 1 h at room temperature in fresh blocking solution containing the appropriately diluted primary antisera (anti-TraJ, 1:40,000; anti-TraM, 1:10,000; and anti-TraY, 1:2,000) and then washed in TBST four times for approximately 15 min each time. Blots were subsequently probed with horseradish peroxidase-conjugated donkey anti-rabbit immunoglobulin G (Amersham Biosciences) diluted 1:20,000 in fresh blocking solution for 1 h at room temperature. Washing was then performed as described above, and the blots were developed using Western Lightning Chemiluminescence Reagent Plus (Perkin-Elmer Life Sciences) and exposed to X-OMAT AR film (Kodak).
Northern analysis. AM111/pOX38-Tc and AM112/pOX38-Tc donor cultures in LB with kanamycin and tetracycline were diluted 200-fold from standing overnight cultures grown at 37°C into fresh LB broth and then grown with shaking. For transcript half-life studies, rifampin was added to the cultures to a final concentration of 200 μg/ml at 0 min. At the indicated time points, culture samples equivalent to an OD600 of approximately 1.0 were rapidly pelleted, snap-frozen in a dry ice-ethanol bath, and stored at –80°C. Total cellular RNA was extracted using the hot phenol method previously described (17). From each sample, 20-μg RNA aliquots were loaded and electrophoresed in a 1.5% agarose gel containing 5% formaldehyde. All subsequent steps, including transfer, hybridization, and washing, were performed as previously described (37). traJ mRNA was detected using universally labeled in vitro-transcribed FinP RNA, while traM mRNA was detected using 32P-end-labeled SPE5-ext. Loading controls were performed using 23SR3, an oligonucleotide which hybridizes specifically to the 3' end of the 23S rRNA, using the same hybridization protocol as for traM. Blots were then exposed on a Phosphor Storage screen (GE Healthcare) and visualized using a PhosphorImager (GE Healthcare), and images were analyzed using ImageQuant software (GE Healthcare).
In vitro transcription. To generate FinP RNA for use as a probe for traJ mRNA and for Hfq binding studies, RNA was transcribed in vitro as previously described using pLJ5-13, which carries finP behind a T7 promoter, as a template (17). Briefly, pLJ5-13 was linearized by digesting with BamHI, which cuts the plasmid immediately downstream of finP, ensuring all transcripts are of a uniform length. FinP has an additional G residue at the 5' terminus, which is necessary to facilitate transcription by the T7 RNA polymerase, and the sequence 5'-GGGGAUC-3' at the 3' terminus due to the presence of the BamHI restriction site. Transcription reactions were performed for 3 hours at 37°C with T7 RNA polymerase in the presence of [-32P]UTP and then treated with DNase I to remove any remaining template. This was followed by electrophoresis in a denaturing 8% polyacrylamide gel containing 8 M urea in 1x TBE (90 mM Tris-HCl, 90 mM H3BO3, 2 mM EDTA [pH 8.0]) buffer. The product was cut out of the gel and eluted in diethyl pyrocarbonate-treated elution buffer (0.5 M ammonium acetate, 1 mM EDTA) overnight at 37°C. The eluted RNA was then phenol extracted, ethanol precipitated, and then dissolved in diethyl pyrocarbonate-treated H2O. traJ RNA was similarly synthesized from a PCR-generated template using the primer TvB15, which contains a functional T7 promoter, and the 3' primer TvB14.
Overexpression and purification of Hfq. Hfq was overexpressed and purified as a six-His fusion protein from pTE607 (kindly provided by T. Elliott) in BL21(DE3) as described by Folichon et al. (9) with the following modification. Purified protein was dialyzed against storage buffer containing 50 mM Tris-HCl (pH 7.5), 1 mM EDTA, 50 mM NH4Cl, 20% glycerol (vol/vol), and 0.1% Triton X-100 (vol/vol). Pure Hfq was quantified using a standard curve of bovine serum albumin via the Bradford protein assay.
Electrophoretic mobility shift assays. RNA binding assays were performed in 50 μl of reaction buffer containing 10 mM Tris-HCl (pH 8.0), 1 mM EDTA, 80 mM NaCl, 10% glycerol (vol/vol), and 0.01% dodecyl maltoside (vol/vol). Five femtomoles of in vitro-transcribed 32P-labeled RNA was incubated in the reaction with increasing concentrations of Hfq for 20 min at 37°C. Competitive binding assays were performed in the presence of 100 ng/μl E. coli tRNA. Following the incubation, reactions were loaded onto native 6% acrylamide gels run in 1x TBE at 4°C. Gels were then dried, exposed to a Phosphor Storage screen (GE Healthcare) overnight, developed on a PhosphorImager (GE Healthcare), and analyzed using ImageQuant software (GE Healthcare).
RESULTS
Hfq represses F-plasmid transfer as donor cell cultures enter stationary phase. To examine the effect of Hfq on F-plasmid transfer, two mutant host strains were used, since hfq is located in a polycistronic operon and disruption of hfq has polar effects on downstream genes, which might affect the F plasmid indirectly. The first mutant strain, AM111, contains an hfq-1:: mutation, which is located near the 5' terminus of the hfq gene, disrupting it and resulting in an hfq phenotype (31). To assay for possible downstream polar effects, a second strain, AM112, containing the hfq-2:: mutation located near the 3' terminus of the hfq gene, which allows for functional Hfq synthesis while still disrupting transcription of the downstream genes, was used (31). In our analyses of the F plasmid, AM112 behaved identically to the wild-type isogenic strain, MC4100, and was used as such (data not shown).
AM111 and AM112 containing pOX38-Tc, an F-plasmid derivative containing the entire transfer region, were grown to stationary phase, and at the indicated time points, samples were taken and assayed for mating efficiency (Fig. 3). The results indicated that Hfq has a growth phase-dependent, negative effect on plasmid transfer. Mating efficiency in AM112 decreased rapidly as the culture approached stationary phase, as was observed in the wild-type strain MC4100 (37). However, the hfq donor strain, AM111, exhibited a slower and smaller decrease in mating efficiency than AM112, with the difference between the two strains becoming more pronounced as the cultures entered stationary phase. There was a slight difference in growth rate between the two strains, but this was discounted as a factor, since there was a significant difference in mating efficiency after 24 h of growth. Mating efficiency was also assayed in donor strains containing pSnO104, which provides functional FinO in trans, restoring FinOP-mediated repression of TraJ synthesis. Equivalent levels of mating were observed in AM111 and AM112, suggesting that Hfq was not necessary for FinOP-mediated repression (data not shown).
Stationary-phase TraJ and TraM levels are increased in an hfq host. Since Hfq could be influencing plasmid transfer by altering the expression of the three regulators, TraM, TraJ, and TraY, their respective transcripts were examined for putative Hfq binding sites. Only one site, which was located in the traM-traJ intergenic region (Fig. 2), was identified. The site consisted of the 24-base AU-rich spacer flanked by stem-loops SL-III and SL-IIc in the 5' traJ mRNA UTR, which was similar to many other documented Hfq binding sites. However, this sequence is absent from the complementary antisense RNA, FinP. Whereas this UTR is present at the 5' terminus of the traJ mRNA, it is also present at the 3' terminus of the traM transcript, potentially acting as a rho-independent transcriptional terminator. Thus, we reasoned that the traM and traJ transcripts were likely targets of Hfq.
To determine if this was the case, immunoblot analysis was performed on AM111/pOX38-Tc and AM112/pOX38-Tc donor cell cultures in exponential-phase growth (at an OD600 of 0.5, after approximately 2.5 h of growth), early stationary phase (8 h), and late stationary phase (24 h) to determine the intracellular levels of the three plasmid regulators (Fig. 4). In exponential phase, only TraJ levels varied between the two strains, appearing to be slightly lower in AM111/pOX38-Tc. However, after both 8 and 24 h of growth, the intracellular levels of both TraJ and TraM were higher in AM111/pOX38-Tc than AM112/pOX38-Tc. TraY levels did not differ between the two strains at any time point. This suggests that the increase in TraM levels was not due to an increase in TraY because of an indirect effect of TraJ. Rather, it implies that Hfq acted directly on both traM and traJ mRNA. TraM and TraJ levels were also assayed in both donor strains containing pSnO104, providing FinO in trans. The mutation of hfq did not have any effect on FinOP-mediated repression, further suggesting that Hfq acts independently of FinOP (data not shown).
Hfq binds the traM-traJ intergenic UTR. To determine whether Hfq binds the traM-traJ intergenic UTR, electrophoretic mobility shift assays were performed with pure Hfq protein and in vitro-synthesized RNA templates (Fig. 2). As the complementary FinP lacks most of the complementary spacer sequence, as well as any sequence complementary to SL-III, Hfq was predicted not to bind to FinP or to do so with a much lower affinity. Binding assays were performed with traJ184 RNA, which consists of the first 184 bases of the traJ transcript, including the predicted binding site, and FinP RNA. In the absence of competitor, Hfq bound both fragments, although its affinity for traJ184 was significantly higher than that for FinP (data not shown). The binding assays were then repeated in the presence of an excess of tRNA, acting as a competitor, to demonstrate specific binding. Hfq bound traJ184 in the presence of the competitor (Fig. 5A) with a dissociation constant (Kd) of 83 nM. However, binding to FinP was almost completely inhibited by the presence of tRNA (Fig. 5B), suggesting that Hfq binds the predicted site in the traJ UTR specifically.
Plasmid transcriptional profiles are altered in an hfq mutant host. In wild-type donor strains, transcription of traM and traJ is highest early in exponential phase and then decreases rapidly as cells progress towards stationary phase (37). If Hfq influences transcript levels, likely by targeting the traJ and traM mRNA for degradation, transcript levels would be predicted to increase as the donor culture approaches stationary phase in AM111 compared to AM112. To determine if Hfq directly alters transcript levels, samples were collected from AM111/pOX38-Tc and AM112/pOX38-Tc donor cultures at regular intervals throughout the growth curve, and total cellular RNA was extracted and examined by Northern blot analysis. Blots were probed with in vitro-synthesized 32P-labeled FinP RNA to detect traJ mRNA (Fig. 6A). AM112/pOX38-Tc displayed a wild-type transcriptional profile, with traJ transcripts decreasing as the donor cultures progressed towards stationary phase, becoming nearly undetectable after 8 h of growth. AM111/pOX38-Tc had a very different profile. Transcript levels were lower than wild-type levels at the earliest time points but slowly rose and peaked after 6 h of growth. The initial lag in transcript levels can be explained by secondary effects via increased H-NS levels in AM111 (29) and, hence, increased H-NS-mediated repression of traJ transcription (see the Discussion). Despite the increase in H-NS-mediated repression, transcripts accumulated and remained detectable in stationary phase, likely due to increased stability. Transcripts larger than the predicted traJ transcript (which is approximately 900 bases) were attributed to transcriptional read-through, both from the upstream traM promoter, PM, and from PJ, driving transcription downstream into traY. These larger transcripts include traMJY, a 2.4-kb transcript which is thought to be transcribed from PM, and traJY, a 1.8-kb transcript which is thought to be transcribed from PJ (37). Both of these transcripts should carry the Hfq binding site: traMJY should carry the site in the middle of the transcript, between traM and traJ, whereas traJY should carry the site at its 5' terminus. The transcription profile for these read-through products was also altered, peaking after 6 h of growth. Blots were stripped and reprobed with 32P-end-labeled SPE5-ext, which detects traM-containing transcripts (Fig. 6B). As with traJ mRNA, transcript levels appeared to peak after approximately 6 h of growth and were still readily detectable after 8 h. The increased transcript levels present at later time points suggest that despite increased transcriptional repression via H-NS, Hfq negatively controls traJ and traM transcript levels, possibly by altering transcript stability.
Transcript stability is increased in hfq mutant host cultures. To determine whether Hfq influences traJ and traM transcript stability, possibly via the binding site in the traM-traJ intergenic region, we examined the half-life of RNA species detected by probes specific for both traJ and traM in the hfq-1:: host strain, AM111. AM111/pOX38-Tc and AM112/pOX38-Tc donor cell cultures were grown in LB broth, transcription was then inhibited by the addition of rifampin, and total cellular RNA was collected at the indicated time points. RNA was then used in Northern blot analysis and probed with in vitro-synthesized 32P-labeled FinP RNA specific to the 5' UTR of the traJ transcript (Fig. 7A) and 32P-end-labeled SPE5-ext, which is specific to traM (Fig. 7B). The half-life of the traMJY read-through transcript increased from approximately 4 min in AM112 to 8 min in AM111. This suggests that Hfq destabilizes the traMJY transcript, thereby promoting its degradation. The second read-through transcript, traJY, also increased in stability in AM111. The half-life of traJY was 4.8 min in AM112, which increased to approximately 10 min in AM111. The stability of the smaller monocistronic traJ transcript was also affected by Hfq (Fig. 7C), but its signal was obscured by what appeared to be degradation products that migrated to the same position, preventing an accurate determination of half-life. The smaller fragments migrating below traJ are thought to be degradation products. Their stability did not appear to be affected by Hfq. Accurate half-life determination of the monocistronic traM transcripts was not possible because the appropriate probe detected two closely migrating bands, the smaller of which was a degradation product that cannot be resolved from the traM mRNA, under these experimental conditions (37). Transcript stabilities do not appear to be altered in a growth phase-dependent manner, as similar half-lives were observed after both 3 and 6 h of growth (data not shown). Nonetheless, Hfq appears to decrease the stability of traM and traJ transcripts containing the Hfq binding site.
DISCUSSION
The data presented in this study indicate that the host Sm-like protein, Hfq, regulates F-plasmid transfer by repressing the synthesis of the plasmid regulators TraJ and TraM by binding to an AU-rich UTR encoded between traM and traJ and destabilizing the corresponding transcripts. Hfq is a global RNA chaperone that can regulate genes posttranscriptionally by modulating both transcript stability and translational initiation, either positively or negatively (13). In the case of the F plasmid, Hfq appears to destabilize multiple transcripts that share a single Hfq binding site located in the traM-traJ intergenic UTR (Fig. 2). However, despite the fact that Hfq often regulates other systems via small RNAs (23, 39), it does not appear to be involved in the FinOP antisense RNA system, which represses traJ translation in F-like plasmids. This suggests that Hfq-mediated regulation of transfer gene expression occurs by a distinct mechanism.
Analysis of mating efficiency in both wild-type (AM112) and hfq-1:: (AM111) donor cultures demonstrated increased mating ability in hfq-1:: cultures as they entered stationary phase. Similarly, the intracellular levels of two of the three plasmid transfer regulatory proteins, TraJ and TraM, were increased in hfq-1:: donor cultures as they entered stationary phase. However, there was no difference in intracellular TraY levels in wild-type and hfq-1:: donor cultures at any point during the growth curve. This suggested that Hfq influenced F-plasmid transfer by affecting traM and traJ mRNA. Furthermore, these results indicated that Hfq-mediated repression of TraM was direct, since TraY, an essential activator of traM transcription, was unaffected by Hfq (27).
Hfq typically binds stretches of AU-rich RNA, often flanked by regions with significant secondary structure (25, 39). Therefore, the transcript sequences for the three F-plasmid regulatory genes traJ, traM, and traY were examined for putative Hfq binding sites. One site, located in the traM-traJ intergenic UTR (Fig. 2), was identified. This sequence is located at the 5' terminus of transcripts originating from PJ, the traJ promoter, as well as at the 3' terminus of traM transcripts, and in the middle of read-through transcripts from PM. This sequence is also critical for duplex formation with FinP, the complementary antisense RNA, which represses translation of traJ (14, 19). Electrophoretic mobility shift analysis demonstrated that Hfq bound traJ RNA containing the putative binding site with specificity. However, FinP, which lacks most of the AU-rich spacer, was not bound by Hfq in a physiologically relevant manner.
Transcriptional analysis of F-plasmid tra gene expression in the hfq-1:: (AM111) and wild-type (AM112) strains via Northern blotting presented an intriguing profile. Whereas transcription of tra genes peaked early in exponential growth in a wild-type host and then rapidly decreased as the culture progressed towards stationary phase, reaching undetectable levels (37; W. R. Will and L. S. Frost, unpublished data), transfer gene transcript levels in hfq-1:: mutant hosts were lower in early exponential phase and peaked in late exponential phase. traJ and traM transcript levels were still detectable in hfq-1:: donor cultures in early stationary phase, albeit at lower than maximal levels. This delayed transcript peak is thought to be due to increased H-NS levels, as H-NS has been shown to be a potent repressor of tra gene expression (37; W. R. Will and L. S. Frost, unpublished data). Hfq is critical in facilitating repression of H-NS synthesis by the small RNA DsrA, originally described as an H-NS antisilencer (28, 29). Disruption of hfq results in partial activity of DsrA, leading to increased H-NS and H-NS-mediated repression of the tra genes and other H-NS targets. This could explain the lower levels of traJ and traM transcription observed in exponential phase in AM111. However, tra gene expression is still sufficient in the hfq-1:: (AM111) host to support normal transfer levels. Analysis of transcript stability indicated that traJ- and traM-containing transcripts were stabilized in an hfq-1:: host. Transcript half-lives, particularly for the larger transcripts, increased approximately twofold in the hfq-1:: donor strain. An increase in transcript stability would allow for transcript accumulation, resulting in increased transcript levels at later time points. These results suggest that Hfq promotes the degradation of traJ and traM transcripts by an unidentified nuclease. The destabilization of multiple transcripts does not require multiple Hfq binding sites but appears to utilize a single binding region located in the traJ-traM intergenic UTR, independent of its position within the transcript. Thus, Hfq binding could alter secondary structure so as to promote cleavage by endoribonucleases, in particular RNase E, which is involved in the degradation of FinP (17). RNase E attacks AU-rich sequences, similar to Hfq binding sites, and is commonly involved in Hfq-mediated control of transcript stability (23, 24). Alternatively, Hfq might inhibit ribosome binding, as described in the case of the ompA transcript, which then allows for more efficient degradation by RNase E (36).
Despite its well-documented role in many antisense control systems (13), Hfq does not appear to be involved in FinOP control of TraJ synthesis. It seemed plausible that Hfq, given its role in other systems as an RNA chaperone (12), might serve to promote duplex formation between FinP and the traJ 5' UTR and that FinOP-mediated repression of F might also require Hfq as a cofactor. Although our findings do not indicate a role for Hfq in the repression of TraJ synthesis by the antisense RNA FinP, it is possible that Hfq-mediated destabilization of traM and traJ mRNA requires the presence of a small RNA encoded by the F plasmid or the host. However, the traJ 5' UTR, including the AU-rich spacer, is well conserved in F-like plasmids (11), suggesting that Hfq-mediated control of transfer gene expression is common among these plasmids. We conclude that Hfq is important in destabilizing traM and traJ transcripts to promote a rate of turnover which is in balance with the transfer potential of F donor cells, allowing for greater sensitivity to transcriptional cues from the environment.
ACKNOWLEDGMENTS
This work was supported by Canadian Institutes for Health Research grant MT11249.
We are grateful to Regine Hengge (University of Konstanz) and Karin Schnetz (University of Cologne) for kindly providing strains, Tom Elliott (West Virginia University) for providing pTE607, and James Sandercock and W. J. Page (University of Alberta) for technical assistance with Northern analysis.
REFERENCES
Achtman, M., N. Willetts, and A. J. Clark. 1971. Beginning a genetic analysis of conjugational transfer determined by the F factor in Escherichia coli by isolation and characterization of transfer-deficient mutants. J. Bacteriol. 106:529-538.
Ali Azam, T., A. Iwata, A. Nishimura, S. Ueda, and A. Ishihama. 1999. Growth phase-dependent variation in protein composition of the Escherichia coli nucleoid. J. Bacteriol. 181:6361-6370.
Anthony, K. G., C. Sherburne, R. Sherburne, and L. S. Frost. 1994. The role of the pilus in recipient cell recognition during bacterial conjugation mediated by F-like plasmids. Mol. Microbiol. 13:939-953.
Arthur, D. C., A. F. Ghetu, M. J. Gubbins, R. A. Edwards, L. S. Frost, and J. N. Glover. 2003. FinO is an RNA chaperone that facilitates sense-antisense RNA interactions. EMBO J. 22:6346-6355.
Casadaban, M. J. 1976. Transposition and fusion of the lac genes to selected promoters in Escherichia coli using bacteriophage lambda and Mu. J. Mol. Biol. 104:541-555.
Cheah, K. C., and R. Skurray. 1986. The F plasmid carries an IS3 insertion within finO. J. Gen. Microbiol. 132:3269-3275.
Disque-Kochem, C., and B. Dreiseikelmann. 1997. The cytoplasmic DNA-binding protein TraM binds to the inner membrane protein TraD in vitro. J. Bacteriol. 179:6133-6137.
Dole, S., Y. Klingen, V. Nagarajavel, and K. Schnetz. 2004. The protease Lon and the RNA-binding protein Hfq reduce silencing of the Escherichia coli bgl operon by H-NS. J. Bacteriol. 186:2708-2716.
Folichon, M., V. Arluison, O. Pellegrini, E. Huntzinger, P. Regnier, and E. Hajnsdorf. 2003. The poly(A) binding protein Hfq protects RNA from RNase E and exoribonucleolytic degradation. Nucleic Acids Res. 31:7302-7310.
Franze de Fernandez, M. T., L. Eoyang, and J. T. August. 1968. Factor fraction required for the synthesis of bacteriophage Q RNA. Nature 219:588-590.
Frost, L. S., K. Ippen-Ihler, and R. A. Skurray. 1994. Analysis of the sequence and gene products of the transfer region of the F sex factor. Microbiol. Rev. 58:162-210.
Geissmann, T. A., and D. Touati. 2004. Hfq, a new chaperoning role: binding to messenger RNA determines access for small RNA regulator. EMBO J. 23:396-405.
Gottesman, S. 2004. The small RNA regulators of Escherichia coli: roles and mechanisms. Annu. Rev. Microbiol. 58:303-328.
Gubbins, M. J., D. C. Arthur, A. F. Ghetu, J. N. Glover, and L. S. Frost. 2003. Characterizing the structural features of RNA/RNA interactions of the F-plasmid FinOP fertility inhibition system. J. Biol. Chem. 278:27663-27671.
Gubbins, M. J., W. R. Will, and L. S. Frost. 2005. The F plasmid, a paradigm for bacterial conjugation, p. 151-206. In P. Mullany (ed.), The dynamic bacterial genome. Cambridge University Press, Cambridge, United Kingdom.
Hengge-Aronis, R. 1996. Back to log phase: sigma S as a global regulator in the osmotic control of gene expression in Escherichia coli. Mol. Microbiol. 21:887-893.
Jerome, L. J., T. van Biesen, and L. S. Frost. 1999. Degradation of FinP antisense RNA from F-like plasmids: the RNA-binding protein, FinO, protects FinP from ribonuclease E. J. Mol. Biol. 285:1457-1473.
Komarova, A. V., L. S. Tchufistova, M. Dreyfus, and I. V. Boni. 2005. AU-rich sequences within 5' untranslated leaders enhance translation and stabilize mRNA in Escherichia coli. J. Bacteriol. 187:1344-1349.
Koraimann, G., K. Teferle, G. Markolin, W. Woger, and G. Hogenauer. 1996. The FinOP repressor system of plasmid R1: analysis of the antisense RNA control of traJ expression and conjugative DNA transfer. Mol. Microbiol. 21:811-821.
Lease, R. A., M. E. Cusick, and M. Belfort. 1998. Riboregulation in Escherichia coli: DsrA RNA acts by RNA:RNA interactions at multiple loci. Proc. Natl. Acad. Sci. USA 95:12456-12461.
Lee, S. H., L. S. Frost, and W. Paranchych. 1992. FinOP repression of the F plasmid involves extension of the half-life of FinP antisense RNA by FinO. Mol. Gen. Genet. 235:131-139.
Lu, J., and L. S. Frost. 2005. Mutations in the C-terminal region of TraM provide evidence for in vivo TraM-TraD interactions during F-plasmid conjugation. J. Bacteriol. 187:4767-4773.
Masse, E., F. E. Escorcia, and S. Gottesman. 2003. Coupled degradation of a small regulatory RNA and its mRNA targets in Escherichia coli. Genes Dev. 17:2374-2383.
Moll, I., T. Afonyushkin, O. Vytvytska, V. R. Kaberdin, and U. Blasi. 2003. Coincident Hfq binding and RNase E cleavage sites on mRNA and small regulatory RNAs. RNA 9:1308-1314.
Mller, T., T. Franch, P. Hjrup, D. R. Keene, H. P. Bachinger, R. G. Brennan, and P. Valentin-Hansen. 2002. Hfq: a bacterial Sm-like protein that mediates RNA-RNA interaction. Mol. Cell 9:23-30.
Muffler, A., D. Fischer, and R. Hengge-Aronis. 1996. The RNA-binding protein HF-I, known as a host factor for phage Q RNA replication, is essential for rpoS translation in Escherichia coli. Genes Dev. 10:1143-1151.
Penfold, S. S., J. Simon, and L. S. Frost. 1996. Regulation of the expression of the traM gene of the F sex factor of Escherichia coli. Mol. Microbiol. 20:549-558.
Sledjeski, D., and S. Gottesman. 1995. A small RNA acts as an antisilencer of the H-NS-silenced rcsA gene of Escherichia coli. Proc. Natl. Acad. Sci. USA 92:2003-2007.
Sledjeski, D. D., C. Whitman, and A. Zhang. 2001. Hfq is necessary for regulation by the untranslated RNA DsrA. J. Bacteriol. 183:1997-2005.
Tsui, H. C., G. Feng, and M. E. Winkler. 1997. Negative regulation of mutS and mutH repair gene expression by the Hfq and RpoS global regulators of Escherichia coli K-12. J. Bacteriol. 179:7476-7487.
Tsui, H. C., H. C. Leung, and M. E. Winkler. 1994. Characterization of broadly pleiotropic phenotypes caused by an hfq insertion mutation in Escherichia coli K-12. Mol. Microbiol. 13:35-49.
Valentin-Hansen, P., M. Eriksen, and C. Udesen. 2004. The bacterial Sm-like protein Hfq: a key player in RNA transactions. Mol. Microbiol. 51:1525-1533.
van Biesen, T., and L. S. Frost. 1994. The FinO protein of IncF plasmids binds FinP antisense RNA and its target, traJ mRNA, and promotes duplex formation. Mol. Microbiol. 14:427-436.
van Biesen, T., F. Soderbom, E. G. Wagner, and L. S. Frost. 1993. Structural and functional analyses of the FinP antisense RNA regulatory system of the F conjugative plasmid. Mol. Microbiol. 10:35-43.
Vytvytska, O., J. S. Jakobsen, G. Balcunaite, J. S. Andersen, M. Baccarini, and A. von Gabain. 1998. Host factor I, Hfq, binds to Escherichia coli ompA mRNA in a growth rate-dependent fashion and regulates its stability. Proc. Natl. Acad. Sci. USA 95:14118-14123.
Vytvytska, O., I. Moll, V. R. Kaberdin, A. von Gabain, and U. Blasi. 2000. Hfq (HF1) stimulates ompA mRNA decay by interfering with ribosome binding. Genes Dev. 14:1109-1118.
Will, W. R., and L. S. Frost. Characterization of the opposing roles of H-NS and TraJ in transcriptional regulation of the F-plasmid tra operon. J. Bacteriol., in press.
Will, W. R., J. Lu, and L. S. Frost. 2004. The role of H-NS in silencing F transfer gene expression during entry into stationary phase. Mol. Microbiol. 54:769-782.
Willetts, N. 1977. The transcriptional control of fertility in F-like plasmids. J. Mol. Biol. 112:141-148.
Zhang, A., K. M. Wassarman, J. Ortega, A. C. Steven, and G. Storz. 2002. The Sm-like Hfq protein increases OxyS RNA interaction with target mRNAs. Mol. Cell 9:11-22.(William R. Will and Laura)
ABSTRACT
The F plasmid of Escherichia coli allows horizontal DNA transfer between an F+ donor cell and an F– recipient. Expression of the transfer genes is tightly controlled by a number of factors, including the following plasmid-encoded regulatory proteins: TraJ, the primary activator of the 33-kb tra operon, and the autoregulators TraM and TraY. Here, we demonstrate that the host RNA binding protein, Hfq, represses TraJ and TraM synthesis by destabilizing their respective mRNAs. Mating assays and immunoblot analyses for TraM and TraJ showed that transfer efficiency and protein levels increased in host cells containing a disruption in hfq compared to wild-type cells in stationary phase. The stability of transcripts containing a putative Hfq binding site located in the intergenic untranslated region between traM and traJ was increased in hfq mutant donor cells, suggesting that Hfq destabilizes these transcripts. Electrophoretic mobility shift assays demonstrated that Hfq specifically binds this region but not the antisense RNA, FinP, encoded on the opposite strand. Together, these findings indicate that Hfq regulates traM and traJ transcript stability by a mechanism separate from FinOP-mediated repression.
INTRODUCTION
The F plasmid of Escherichia coli is considered a model for bacterial conjugation and has been studied extensively (15, 18). F-plasmid transfer is facilitated by a large multicomponent protein complex which spans the donor cell membrane to transport plasmid DNA into a recipient cell. This complex is encoded by a single 33-kb polycistronic transfer (tra) operon (11). Recent studies have demonstrated that the regulatory circuit controlling the expression of this tra operon is complex. However, the focal point of control appears to be TraJ. TraJ is the primary activator of tra operon expression, encoded on a monocistronic operon immediately upstream of the tra operon and downstream from traM (11, 38) (Fig. 1). It appears to act by opposing H-NS-mediated repression of the tra operon promoter PY (36a). The resulting derepression of PY allows the synthesis of the secondary regulator, TraY, encoded by the tra operon, which further activates PY, as well as traM (27). TraM is required to transmit the signal for the initiation of DNA transfer between the transfer complex and the relaxosome, which nicks and unwinds the plasmid DNA during transfer (7, 22). TraM also has a regulatory role as an autorepressor that represses expression from the two tandem traM promoters, referred to here collectively as PM (27).
In most F-like plasmids, TraJ synthesis is subject to posttranscriptional control via the FinOP antisense RNA system, whereby finP encodes a small antisense RNA complementary to the 5' untranslated region (UTR) of the traJ transcript (33). FinP is a short, 79-base RNA consisting of two stem-loop sequences, SL-I and SL-II (34) (Fig. 2). These are complementary to stem-loops present in the 5' UTR of traJ, referred to as SL-Ic and SL-IIc, respectively. A third stem-loop, SL-III, is also present in the 5' UTR of the traJ mRNA, but not in FinP, and is separated from SL-IIc by a 24-base AU-rich spacer region. These structural elements appear to be well conserved and are present in most F-like traJ transcripts (11). FinO, a plasmid-encoded RNA chaperone, binds FinP, protecting it from degradation by RNase E and promoting duplex formation between the complementary stem-loop structures in both FinP and traJ (4, 17). The resulting duplex sequesters the ribosome binding site (RBS) and prevents the translation of traJ, resulting in the repression of plasmid transfer (14, 19). However, the F plasmid is naturally derepressed, since its copy of finO has been disrupted by the insertion of an IS3 element (6).
In this study, we considered the possibility that host proteins might also affect traJ mRNA synthesis using a posttranscriptional mechanism similar to the FinOP system in F-like plasmids, with Hfq, the host RNA chaperone, being the most likely candidate. Originally characterized as host factor I (HF-I), a host-encoded protein necessary for the in vitro replication of the Q RNA bacteriophage (10), Hfq has emerged as a potent regulator of many aspects of RNA biology, influencing stability, translation, and RNA bacteriophage replication, often via small RNAs (13, 32). Hfq is an 11.2-kDa protein, forming hexamers that preferentially bind sequences of AU-rich RNA, often flanked by structured regions (25, 39). It is relatively abundant and is present at intracellular levels of approximately 10,000 hexamers per cell (2). However, the expression profile throughout the growth cycle remains uncertain, as some studies have indicated that Hfq levels peak and then decrease rapidly during lag phase, dropping to approximately one third of the intracellular maximum throughout the rest of the growth cycle (2). Other studies have found that Hfq levels increase as cultures enter stationary phase or during periods of slow growth (30, 35).
Given that TraJ synthesis is regulated by both an antisense RNA system, a common target of Hfq, and H-NS, which has been shown to overlap with Hfq in many regulatory circuits, including rpoS, hns, and bgl regulation (8, 16, 20, 28, 29), we considered the possibility that Hfq might target and regulate TraJ mRNA in some manner. In this study, we demonstrate that Hfq binds to the intergenic UTR, 3' to traM and 5' to traJ, and decreases the stability of transcripts containing this region. Hfq does not appear to be involved in fertility inhibition and has no role in FinOP-mediated repression. Instead, Hfq appears to act as a repressor of TraJ and TraM synthesis, as well as F-plasmid transfer in general, by destabilizing the corresponding transcripts.
MATERIALS AND METHODS
Experimental strains and growth conditions. All bacterial strains and plasmids used in this study are described in Table 1. Unless otherwise stated, all liquid cultures were grown in Luria-Bertani (LB) broth at 37°C with shaking. Solid medium cultures were grown on LB agar incubated at 37°C overnight. Antibiotics were added as indicated to selective media at the following concentrations: ampicillin, 25 μg/ml; kanamycin, 25 μg/ml; spectinomycin, 100 μg/ml; streptomycin, 200 μg/ml; and tetracycline, 10 μg/ml.
Mating assays. Standing overnight cultures of AM111 and AM112 containing the F-plasmid derivative pOX38-Tc were diluted 200-fold into fresh LB. These donor cultures were then incubated at 37°C with shaking. At the indicated time points, the optical density at 600 nm (OD600) was taken and donor culture samples were removed to assay for mating ability. Aliquots of both donor cells and ED24 recipient cells, each equivalent to an OD600 of 0.1, were mixed in a total volume of 1 ml of spent LB broth. Spent LB broth refers to broth which is removed at the same time point as the donor cells and centrifuged to remove any bacteria present. These mating reactions were incubated, standing, at 37°C for 1 h. Mating reactions were then vigorously vortexed for 10 to 20 seconds and placed on ice for approximately 5 min to halt any further plasmid transfer. This was followed by serially diluting the mating reactions in SSC buffer (150 mM NaCl, 15 mM sodium citrate, pH 7.0) and plating the diluted cells on LB agar medium containing antibiotics selective for either donors (streptomycin and tetracycline) or transconjugants (spectinomycin and tetracycline). Growth was scored, and mating efficiency was measured as transconjugants per donor. All mating data represent the average of quadruplicate experiments.
Immunoblot analysis. For growth phase-dependent immunoblot analysis, standing overnight cultures of AM112/pOX38-Tc and AM111/pOX38-Tc were grown in LB broth containing tetracycline and kanamycin at 37°C, then diluted 1:200 in fresh LB containing tetracycline, and grown with shaking. At the indicated time points, culture aliquots equivalent to an OD600 of 0.1 were removed, and the cells were pelleted and frozen. Samples were resuspended in sodium dodecyl sulfate-polyacrylamide gel electrophoresis sample loading buffer and then boiled for 5 min before electrophoresis in a 15% sodium dodecyl sulfate-polyacrylamide gel. Following electrophoresis, gels were transferred onto Immobilon-P membrane (Millipore) and blocked overnight at 4°C in blocking solution consisting of 10% skim milk dissolved in TBST (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, and 0.1% Tween 20). Blots were probed for 1 h at room temperature in fresh blocking solution containing the appropriately diluted primary antisera (anti-TraJ, 1:40,000; anti-TraM, 1:10,000; and anti-TraY, 1:2,000) and then washed in TBST four times for approximately 15 min each time. Blots were subsequently probed with horseradish peroxidase-conjugated donkey anti-rabbit immunoglobulin G (Amersham Biosciences) diluted 1:20,000 in fresh blocking solution for 1 h at room temperature. Washing was then performed as described above, and the blots were developed using Western Lightning Chemiluminescence Reagent Plus (Perkin-Elmer Life Sciences) and exposed to X-OMAT AR film (Kodak).
Northern analysis. AM111/pOX38-Tc and AM112/pOX38-Tc donor cultures in LB with kanamycin and tetracycline were diluted 200-fold from standing overnight cultures grown at 37°C into fresh LB broth and then grown with shaking. For transcript half-life studies, rifampin was added to the cultures to a final concentration of 200 μg/ml at 0 min. At the indicated time points, culture samples equivalent to an OD600 of approximately 1.0 were rapidly pelleted, snap-frozen in a dry ice-ethanol bath, and stored at –80°C. Total cellular RNA was extracted using the hot phenol method previously described (17). From each sample, 20-μg RNA aliquots were loaded and electrophoresed in a 1.5% agarose gel containing 5% formaldehyde. All subsequent steps, including transfer, hybridization, and washing, were performed as previously described (37). traJ mRNA was detected using universally labeled in vitro-transcribed FinP RNA, while traM mRNA was detected using 32P-end-labeled SPE5-ext. Loading controls were performed using 23SR3, an oligonucleotide which hybridizes specifically to the 3' end of the 23S rRNA, using the same hybridization protocol as for traM. Blots were then exposed on a Phosphor Storage screen (GE Healthcare) and visualized using a PhosphorImager (GE Healthcare), and images were analyzed using ImageQuant software (GE Healthcare).
In vitro transcription. To generate FinP RNA for use as a probe for traJ mRNA and for Hfq binding studies, RNA was transcribed in vitro as previously described using pLJ5-13, which carries finP behind a T7 promoter, as a template (17). Briefly, pLJ5-13 was linearized by digesting with BamHI, which cuts the plasmid immediately downstream of finP, ensuring all transcripts are of a uniform length. FinP has an additional G residue at the 5' terminus, which is necessary to facilitate transcription by the T7 RNA polymerase, and the sequence 5'-GGGGAUC-3' at the 3' terminus due to the presence of the BamHI restriction site. Transcription reactions were performed for 3 hours at 37°C with T7 RNA polymerase in the presence of [-32P]UTP and then treated with DNase I to remove any remaining template. This was followed by electrophoresis in a denaturing 8% polyacrylamide gel containing 8 M urea in 1x TBE (90 mM Tris-HCl, 90 mM H3BO3, 2 mM EDTA [pH 8.0]) buffer. The product was cut out of the gel and eluted in diethyl pyrocarbonate-treated elution buffer (0.5 M ammonium acetate, 1 mM EDTA) overnight at 37°C. The eluted RNA was then phenol extracted, ethanol precipitated, and then dissolved in diethyl pyrocarbonate-treated H2O. traJ RNA was similarly synthesized from a PCR-generated template using the primer TvB15, which contains a functional T7 promoter, and the 3' primer TvB14.
Overexpression and purification of Hfq. Hfq was overexpressed and purified as a six-His fusion protein from pTE607 (kindly provided by T. Elliott) in BL21(DE3) as described by Folichon et al. (9) with the following modification. Purified protein was dialyzed against storage buffer containing 50 mM Tris-HCl (pH 7.5), 1 mM EDTA, 50 mM NH4Cl, 20% glycerol (vol/vol), and 0.1% Triton X-100 (vol/vol). Pure Hfq was quantified using a standard curve of bovine serum albumin via the Bradford protein assay.
Electrophoretic mobility shift assays. RNA binding assays were performed in 50 μl of reaction buffer containing 10 mM Tris-HCl (pH 8.0), 1 mM EDTA, 80 mM NaCl, 10% glycerol (vol/vol), and 0.01% dodecyl maltoside (vol/vol). Five femtomoles of in vitro-transcribed 32P-labeled RNA was incubated in the reaction with increasing concentrations of Hfq for 20 min at 37°C. Competitive binding assays were performed in the presence of 100 ng/μl E. coli tRNA. Following the incubation, reactions were loaded onto native 6% acrylamide gels run in 1x TBE at 4°C. Gels were then dried, exposed to a Phosphor Storage screen (GE Healthcare) overnight, developed on a PhosphorImager (GE Healthcare), and analyzed using ImageQuant software (GE Healthcare).
RESULTS
Hfq represses F-plasmid transfer as donor cell cultures enter stationary phase. To examine the effect of Hfq on F-plasmid transfer, two mutant host strains were used, since hfq is located in a polycistronic operon and disruption of hfq has polar effects on downstream genes, which might affect the F plasmid indirectly. The first mutant strain, AM111, contains an hfq-1:: mutation, which is located near the 5' terminus of the hfq gene, disrupting it and resulting in an hfq phenotype (31). To assay for possible downstream polar effects, a second strain, AM112, containing the hfq-2:: mutation located near the 3' terminus of the hfq gene, which allows for functional Hfq synthesis while still disrupting transcription of the downstream genes, was used (31). In our analyses of the F plasmid, AM112 behaved identically to the wild-type isogenic strain, MC4100, and was used as such (data not shown).
AM111 and AM112 containing pOX38-Tc, an F-plasmid derivative containing the entire transfer region, were grown to stationary phase, and at the indicated time points, samples were taken and assayed for mating efficiency (Fig. 3). The results indicated that Hfq has a growth phase-dependent, negative effect on plasmid transfer. Mating efficiency in AM112 decreased rapidly as the culture approached stationary phase, as was observed in the wild-type strain MC4100 (37). However, the hfq donor strain, AM111, exhibited a slower and smaller decrease in mating efficiency than AM112, with the difference between the two strains becoming more pronounced as the cultures entered stationary phase. There was a slight difference in growth rate between the two strains, but this was discounted as a factor, since there was a significant difference in mating efficiency after 24 h of growth. Mating efficiency was also assayed in donor strains containing pSnO104, which provides functional FinO in trans, restoring FinOP-mediated repression of TraJ synthesis. Equivalent levels of mating were observed in AM111 and AM112, suggesting that Hfq was not necessary for FinOP-mediated repression (data not shown).
Stationary-phase TraJ and TraM levels are increased in an hfq host. Since Hfq could be influencing plasmid transfer by altering the expression of the three regulators, TraM, TraJ, and TraY, their respective transcripts were examined for putative Hfq binding sites. Only one site, which was located in the traM-traJ intergenic region (Fig. 2), was identified. The site consisted of the 24-base AU-rich spacer flanked by stem-loops SL-III and SL-IIc in the 5' traJ mRNA UTR, which was similar to many other documented Hfq binding sites. However, this sequence is absent from the complementary antisense RNA, FinP. Whereas this UTR is present at the 5' terminus of the traJ mRNA, it is also present at the 3' terminus of the traM transcript, potentially acting as a rho-independent transcriptional terminator. Thus, we reasoned that the traM and traJ transcripts were likely targets of Hfq.
To determine if this was the case, immunoblot analysis was performed on AM111/pOX38-Tc and AM112/pOX38-Tc donor cell cultures in exponential-phase growth (at an OD600 of 0.5, after approximately 2.5 h of growth), early stationary phase (8 h), and late stationary phase (24 h) to determine the intracellular levels of the three plasmid regulators (Fig. 4). In exponential phase, only TraJ levels varied between the two strains, appearing to be slightly lower in AM111/pOX38-Tc. However, after both 8 and 24 h of growth, the intracellular levels of both TraJ and TraM were higher in AM111/pOX38-Tc than AM112/pOX38-Tc. TraY levels did not differ between the two strains at any time point. This suggests that the increase in TraM levels was not due to an increase in TraY because of an indirect effect of TraJ. Rather, it implies that Hfq acted directly on both traM and traJ mRNA. TraM and TraJ levels were also assayed in both donor strains containing pSnO104, providing FinO in trans. The mutation of hfq did not have any effect on FinOP-mediated repression, further suggesting that Hfq acts independently of FinOP (data not shown).
Hfq binds the traM-traJ intergenic UTR. To determine whether Hfq binds the traM-traJ intergenic UTR, electrophoretic mobility shift assays were performed with pure Hfq protein and in vitro-synthesized RNA templates (Fig. 2). As the complementary FinP lacks most of the complementary spacer sequence, as well as any sequence complementary to SL-III, Hfq was predicted not to bind to FinP or to do so with a much lower affinity. Binding assays were performed with traJ184 RNA, which consists of the first 184 bases of the traJ transcript, including the predicted binding site, and FinP RNA. In the absence of competitor, Hfq bound both fragments, although its affinity for traJ184 was significantly higher than that for FinP (data not shown). The binding assays were then repeated in the presence of an excess of tRNA, acting as a competitor, to demonstrate specific binding. Hfq bound traJ184 in the presence of the competitor (Fig. 5A) with a dissociation constant (Kd) of 83 nM. However, binding to FinP was almost completely inhibited by the presence of tRNA (Fig. 5B), suggesting that Hfq binds the predicted site in the traJ UTR specifically.
Plasmid transcriptional profiles are altered in an hfq mutant host. In wild-type donor strains, transcription of traM and traJ is highest early in exponential phase and then decreases rapidly as cells progress towards stationary phase (37). If Hfq influences transcript levels, likely by targeting the traJ and traM mRNA for degradation, transcript levels would be predicted to increase as the donor culture approaches stationary phase in AM111 compared to AM112. To determine if Hfq directly alters transcript levels, samples were collected from AM111/pOX38-Tc and AM112/pOX38-Tc donor cultures at regular intervals throughout the growth curve, and total cellular RNA was extracted and examined by Northern blot analysis. Blots were probed with in vitro-synthesized 32P-labeled FinP RNA to detect traJ mRNA (Fig. 6A). AM112/pOX38-Tc displayed a wild-type transcriptional profile, with traJ transcripts decreasing as the donor cultures progressed towards stationary phase, becoming nearly undetectable after 8 h of growth. AM111/pOX38-Tc had a very different profile. Transcript levels were lower than wild-type levels at the earliest time points but slowly rose and peaked after 6 h of growth. The initial lag in transcript levels can be explained by secondary effects via increased H-NS levels in AM111 (29) and, hence, increased H-NS-mediated repression of traJ transcription (see the Discussion). Despite the increase in H-NS-mediated repression, transcripts accumulated and remained detectable in stationary phase, likely due to increased stability. Transcripts larger than the predicted traJ transcript (which is approximately 900 bases) were attributed to transcriptional read-through, both from the upstream traM promoter, PM, and from PJ, driving transcription downstream into traY. These larger transcripts include traMJY, a 2.4-kb transcript which is thought to be transcribed from PM, and traJY, a 1.8-kb transcript which is thought to be transcribed from PJ (37). Both of these transcripts should carry the Hfq binding site: traMJY should carry the site in the middle of the transcript, between traM and traJ, whereas traJY should carry the site at its 5' terminus. The transcription profile for these read-through products was also altered, peaking after 6 h of growth. Blots were stripped and reprobed with 32P-end-labeled SPE5-ext, which detects traM-containing transcripts (Fig. 6B). As with traJ mRNA, transcript levels appeared to peak after approximately 6 h of growth and were still readily detectable after 8 h. The increased transcript levels present at later time points suggest that despite increased transcriptional repression via H-NS, Hfq negatively controls traJ and traM transcript levels, possibly by altering transcript stability.
Transcript stability is increased in hfq mutant host cultures. To determine whether Hfq influences traJ and traM transcript stability, possibly via the binding site in the traM-traJ intergenic region, we examined the half-life of RNA species detected by probes specific for both traJ and traM in the hfq-1:: host strain, AM111. AM111/pOX38-Tc and AM112/pOX38-Tc donor cell cultures were grown in LB broth, transcription was then inhibited by the addition of rifampin, and total cellular RNA was collected at the indicated time points. RNA was then used in Northern blot analysis and probed with in vitro-synthesized 32P-labeled FinP RNA specific to the 5' UTR of the traJ transcript (Fig. 7A) and 32P-end-labeled SPE5-ext, which is specific to traM (Fig. 7B). The half-life of the traMJY read-through transcript increased from approximately 4 min in AM112 to 8 min in AM111. This suggests that Hfq destabilizes the traMJY transcript, thereby promoting its degradation. The second read-through transcript, traJY, also increased in stability in AM111. The half-life of traJY was 4.8 min in AM112, which increased to approximately 10 min in AM111. The stability of the smaller monocistronic traJ transcript was also affected by Hfq (Fig. 7C), but its signal was obscured by what appeared to be degradation products that migrated to the same position, preventing an accurate determination of half-life. The smaller fragments migrating below traJ are thought to be degradation products. Their stability did not appear to be affected by Hfq. Accurate half-life determination of the monocistronic traM transcripts was not possible because the appropriate probe detected two closely migrating bands, the smaller of which was a degradation product that cannot be resolved from the traM mRNA, under these experimental conditions (37). Transcript stabilities do not appear to be altered in a growth phase-dependent manner, as similar half-lives were observed after both 3 and 6 h of growth (data not shown). Nonetheless, Hfq appears to decrease the stability of traM and traJ transcripts containing the Hfq binding site.
DISCUSSION
The data presented in this study indicate that the host Sm-like protein, Hfq, regulates F-plasmid transfer by repressing the synthesis of the plasmid regulators TraJ and TraM by binding to an AU-rich UTR encoded between traM and traJ and destabilizing the corresponding transcripts. Hfq is a global RNA chaperone that can regulate genes posttranscriptionally by modulating both transcript stability and translational initiation, either positively or negatively (13). In the case of the F plasmid, Hfq appears to destabilize multiple transcripts that share a single Hfq binding site located in the traM-traJ intergenic UTR (Fig. 2). However, despite the fact that Hfq often regulates other systems via small RNAs (23, 39), it does not appear to be involved in the FinOP antisense RNA system, which represses traJ translation in F-like plasmids. This suggests that Hfq-mediated regulation of transfer gene expression occurs by a distinct mechanism.
Analysis of mating efficiency in both wild-type (AM112) and hfq-1:: (AM111) donor cultures demonstrated increased mating ability in hfq-1:: cultures as they entered stationary phase. Similarly, the intracellular levels of two of the three plasmid transfer regulatory proteins, TraJ and TraM, were increased in hfq-1:: donor cultures as they entered stationary phase. However, there was no difference in intracellular TraY levels in wild-type and hfq-1:: donor cultures at any point during the growth curve. This suggested that Hfq influenced F-plasmid transfer by affecting traM and traJ mRNA. Furthermore, these results indicated that Hfq-mediated repression of TraM was direct, since TraY, an essential activator of traM transcription, was unaffected by Hfq (27).
Hfq typically binds stretches of AU-rich RNA, often flanked by regions with significant secondary structure (25, 39). Therefore, the transcript sequences for the three F-plasmid regulatory genes traJ, traM, and traY were examined for putative Hfq binding sites. One site, located in the traM-traJ intergenic UTR (Fig. 2), was identified. This sequence is located at the 5' terminus of transcripts originating from PJ, the traJ promoter, as well as at the 3' terminus of traM transcripts, and in the middle of read-through transcripts from PM. This sequence is also critical for duplex formation with FinP, the complementary antisense RNA, which represses translation of traJ (14, 19). Electrophoretic mobility shift analysis demonstrated that Hfq bound traJ RNA containing the putative binding site with specificity. However, FinP, which lacks most of the AU-rich spacer, was not bound by Hfq in a physiologically relevant manner.
Transcriptional analysis of F-plasmid tra gene expression in the hfq-1:: (AM111) and wild-type (AM112) strains via Northern blotting presented an intriguing profile. Whereas transcription of tra genes peaked early in exponential growth in a wild-type host and then rapidly decreased as the culture progressed towards stationary phase, reaching undetectable levels (37; W. R. Will and L. S. Frost, unpublished data), transfer gene transcript levels in hfq-1:: mutant hosts were lower in early exponential phase and peaked in late exponential phase. traJ and traM transcript levels were still detectable in hfq-1:: donor cultures in early stationary phase, albeit at lower than maximal levels. This delayed transcript peak is thought to be due to increased H-NS levels, as H-NS has been shown to be a potent repressor of tra gene expression (37; W. R. Will and L. S. Frost, unpublished data). Hfq is critical in facilitating repression of H-NS synthesis by the small RNA DsrA, originally described as an H-NS antisilencer (28, 29). Disruption of hfq results in partial activity of DsrA, leading to increased H-NS and H-NS-mediated repression of the tra genes and other H-NS targets. This could explain the lower levels of traJ and traM transcription observed in exponential phase in AM111. However, tra gene expression is still sufficient in the hfq-1:: (AM111) host to support normal transfer levels. Analysis of transcript stability indicated that traJ- and traM-containing transcripts were stabilized in an hfq-1:: host. Transcript half-lives, particularly for the larger transcripts, increased approximately twofold in the hfq-1:: donor strain. An increase in transcript stability would allow for transcript accumulation, resulting in increased transcript levels at later time points. These results suggest that Hfq promotes the degradation of traJ and traM transcripts by an unidentified nuclease. The destabilization of multiple transcripts does not require multiple Hfq binding sites but appears to utilize a single binding region located in the traJ-traM intergenic UTR, independent of its position within the transcript. Thus, Hfq binding could alter secondary structure so as to promote cleavage by endoribonucleases, in particular RNase E, which is involved in the degradation of FinP (17). RNase E attacks AU-rich sequences, similar to Hfq binding sites, and is commonly involved in Hfq-mediated control of transcript stability (23, 24). Alternatively, Hfq might inhibit ribosome binding, as described in the case of the ompA transcript, which then allows for more efficient degradation by RNase E (36).
Despite its well-documented role in many antisense control systems (13), Hfq does not appear to be involved in FinOP control of TraJ synthesis. It seemed plausible that Hfq, given its role in other systems as an RNA chaperone (12), might serve to promote duplex formation between FinP and the traJ 5' UTR and that FinOP-mediated repression of F might also require Hfq as a cofactor. Although our findings do not indicate a role for Hfq in the repression of TraJ synthesis by the antisense RNA FinP, it is possible that Hfq-mediated destabilization of traM and traJ mRNA requires the presence of a small RNA encoded by the F plasmid or the host. However, the traJ 5' UTR, including the AU-rich spacer, is well conserved in F-like plasmids (11), suggesting that Hfq-mediated control of transfer gene expression is common among these plasmids. We conclude that Hfq is important in destabilizing traM and traJ transcripts to promote a rate of turnover which is in balance with the transfer potential of F donor cells, allowing for greater sensitivity to transcriptional cues from the environment.
ACKNOWLEDGMENTS
This work was supported by Canadian Institutes for Health Research grant MT11249.
We are grateful to Regine Hengge (University of Konstanz) and Karin Schnetz (University of Cologne) for kindly providing strains, Tom Elliott (West Virginia University) for providing pTE607, and James Sandercock and W. J. Page (University of Alberta) for technical assistance with Northern analysis.
REFERENCES
Achtman, M., N. Willetts, and A. J. Clark. 1971. Beginning a genetic analysis of conjugational transfer determined by the F factor in Escherichia coli by isolation and characterization of transfer-deficient mutants. J. Bacteriol. 106:529-538.
Ali Azam, T., A. Iwata, A. Nishimura, S. Ueda, and A. Ishihama. 1999. Growth phase-dependent variation in protein composition of the Escherichia coli nucleoid. J. Bacteriol. 181:6361-6370.
Anthony, K. G., C. Sherburne, R. Sherburne, and L. S. Frost. 1994. The role of the pilus in recipient cell recognition during bacterial conjugation mediated by F-like plasmids. Mol. Microbiol. 13:939-953.
Arthur, D. C., A. F. Ghetu, M. J. Gubbins, R. A. Edwards, L. S. Frost, and J. N. Glover. 2003. FinO is an RNA chaperone that facilitates sense-antisense RNA interactions. EMBO J. 22:6346-6355.
Casadaban, M. J. 1976. Transposition and fusion of the lac genes to selected promoters in Escherichia coli using bacteriophage lambda and Mu. J. Mol. Biol. 104:541-555.
Cheah, K. C., and R. Skurray. 1986. The F plasmid carries an IS3 insertion within finO. J. Gen. Microbiol. 132:3269-3275.
Disque-Kochem, C., and B. Dreiseikelmann. 1997. The cytoplasmic DNA-binding protein TraM binds to the inner membrane protein TraD in vitro. J. Bacteriol. 179:6133-6137.
Dole, S., Y. Klingen, V. Nagarajavel, and K. Schnetz. 2004. The protease Lon and the RNA-binding protein Hfq reduce silencing of the Escherichia coli bgl operon by H-NS. J. Bacteriol. 186:2708-2716.
Folichon, M., V. Arluison, O. Pellegrini, E. Huntzinger, P. Regnier, and E. Hajnsdorf. 2003. The poly(A) binding protein Hfq protects RNA from RNase E and exoribonucleolytic degradation. Nucleic Acids Res. 31:7302-7310.
Franze de Fernandez, M. T., L. Eoyang, and J. T. August. 1968. Factor fraction required for the synthesis of bacteriophage Q RNA. Nature 219:588-590.
Frost, L. S., K. Ippen-Ihler, and R. A. Skurray. 1994. Analysis of the sequence and gene products of the transfer region of the F sex factor. Microbiol. Rev. 58:162-210.
Geissmann, T. A., and D. Touati. 2004. Hfq, a new chaperoning role: binding to messenger RNA determines access for small RNA regulator. EMBO J. 23:396-405.
Gottesman, S. 2004. The small RNA regulators of Escherichia coli: roles and mechanisms. Annu. Rev. Microbiol. 58:303-328.
Gubbins, M. J., D. C. Arthur, A. F. Ghetu, J. N. Glover, and L. S. Frost. 2003. Characterizing the structural features of RNA/RNA interactions of the F-plasmid FinOP fertility inhibition system. J. Biol. Chem. 278:27663-27671.
Gubbins, M. J., W. R. Will, and L. S. Frost. 2005. The F plasmid, a paradigm for bacterial conjugation, p. 151-206. In P. Mullany (ed.), The dynamic bacterial genome. Cambridge University Press, Cambridge, United Kingdom.
Hengge-Aronis, R. 1996. Back to log phase: sigma S as a global regulator in the osmotic control of gene expression in Escherichia coli. Mol. Microbiol. 21:887-893.
Jerome, L. J., T. van Biesen, and L. S. Frost. 1999. Degradation of FinP antisense RNA from F-like plasmids: the RNA-binding protein, FinO, protects FinP from ribonuclease E. J. Mol. Biol. 285:1457-1473.
Komarova, A. V., L. S. Tchufistova, M. Dreyfus, and I. V. Boni. 2005. AU-rich sequences within 5' untranslated leaders enhance translation and stabilize mRNA in Escherichia coli. J. Bacteriol. 187:1344-1349.
Koraimann, G., K. Teferle, G. Markolin, W. Woger, and G. Hogenauer. 1996. The FinOP repressor system of plasmid R1: analysis of the antisense RNA control of traJ expression and conjugative DNA transfer. Mol. Microbiol. 21:811-821.
Lease, R. A., M. E. Cusick, and M. Belfort. 1998. Riboregulation in Escherichia coli: DsrA RNA acts by RNA:RNA interactions at multiple loci. Proc. Natl. Acad. Sci. USA 95:12456-12461.
Lee, S. H., L. S. Frost, and W. Paranchych. 1992. FinOP repression of the F plasmid involves extension of the half-life of FinP antisense RNA by FinO. Mol. Gen. Genet. 235:131-139.
Lu, J., and L. S. Frost. 2005. Mutations in the C-terminal region of TraM provide evidence for in vivo TraM-TraD interactions during F-plasmid conjugation. J. Bacteriol. 187:4767-4773.
Masse, E., F. E. Escorcia, and S. Gottesman. 2003. Coupled degradation of a small regulatory RNA and its mRNA targets in Escherichia coli. Genes Dev. 17:2374-2383.
Moll, I., T. Afonyushkin, O. Vytvytska, V. R. Kaberdin, and U. Blasi. 2003. Coincident Hfq binding and RNase E cleavage sites on mRNA and small regulatory RNAs. RNA 9:1308-1314.
Mller, T., T. Franch, P. Hjrup, D. R. Keene, H. P. Bachinger, R. G. Brennan, and P. Valentin-Hansen. 2002. Hfq: a bacterial Sm-like protein that mediates RNA-RNA interaction. Mol. Cell 9:23-30.
Muffler, A., D. Fischer, and R. Hengge-Aronis. 1996. The RNA-binding protein HF-I, known as a host factor for phage Q RNA replication, is essential for rpoS translation in Escherichia coli. Genes Dev. 10:1143-1151.
Penfold, S. S., J. Simon, and L. S. Frost. 1996. Regulation of the expression of the traM gene of the F sex factor of Escherichia coli. Mol. Microbiol. 20:549-558.
Sledjeski, D., and S. Gottesman. 1995. A small RNA acts as an antisilencer of the H-NS-silenced rcsA gene of Escherichia coli. Proc. Natl. Acad. Sci. USA 92:2003-2007.
Sledjeski, D. D., C. Whitman, and A. Zhang. 2001. Hfq is necessary for regulation by the untranslated RNA DsrA. J. Bacteriol. 183:1997-2005.
Tsui, H. C., G. Feng, and M. E. Winkler. 1997. Negative regulation of mutS and mutH repair gene expression by the Hfq and RpoS global regulators of Escherichia coli K-12. J. Bacteriol. 179:7476-7487.
Tsui, H. C., H. C. Leung, and M. E. Winkler. 1994. Characterization of broadly pleiotropic phenotypes caused by an hfq insertion mutation in Escherichia coli K-12. Mol. Microbiol. 13:35-49.
Valentin-Hansen, P., M. Eriksen, and C. Udesen. 2004. The bacterial Sm-like protein Hfq: a key player in RNA transactions. Mol. Microbiol. 51:1525-1533.
van Biesen, T., and L. S. Frost. 1994. The FinO protein of IncF plasmids binds FinP antisense RNA and its target, traJ mRNA, and promotes duplex formation. Mol. Microbiol. 14:427-436.
van Biesen, T., F. Soderbom, E. G. Wagner, and L. S. Frost. 1993. Structural and functional analyses of the FinP antisense RNA regulatory system of the F conjugative plasmid. Mol. Microbiol. 10:35-43.
Vytvytska, O., J. S. Jakobsen, G. Balcunaite, J. S. Andersen, M. Baccarini, and A. von Gabain. 1998. Host factor I, Hfq, binds to Escherichia coli ompA mRNA in a growth rate-dependent fashion and regulates its stability. Proc. Natl. Acad. Sci. USA 95:14118-14123.
Vytvytska, O., I. Moll, V. R. Kaberdin, A. von Gabain, and U. Blasi. 2000. Hfq (HF1) stimulates ompA mRNA decay by interfering with ribosome binding. Genes Dev. 14:1109-1118.
Will, W. R., and L. S. Frost. Characterization of the opposing roles of H-NS and TraJ in transcriptional regulation of the F-plasmid tra operon. J. Bacteriol., in press.
Will, W. R., J. Lu, and L. S. Frost. 2004. The role of H-NS in silencing F transfer gene expression during entry into stationary phase. Mol. Microbiol. 54:769-782.
Willetts, N. 1977. The transcriptional control of fertility in F-like plasmids. J. Mol. Biol. 112:141-148.
Zhang, A., K. M. Wassarman, J. Ortega, A. C. Steven, and G. Storz. 2002. The Sm-like Hfq protein increases OxyS RNA interaction with target mRNAs. Mol. Cell 9:11-22.(William R. Will and Laura)