FljA-Mediated Posttranscriptional Control of Phase 1 Flagellin Expression in Flagellar Phase Variation of Salmonella enterica Serovar Typhim
http://www.100md.com
《细菌学杂志》
Graduate School of Natural Science and Technology,Department of Biology, Faculty of Science, Okayama University, Tsushima-Naka 3-1-1, Okayama 700-8530, Japan
ABSTRACT
Flagellar phase variation of Salmonella is a phenomenon where two flagellin genes, fliC (phase 1) and fljB (phase 2), are expressed alternately. This is controlled by the inversion of a DNA segment containing the promoter for the fljB gene. The fljB gene constitutes an operon with the fljA gene, which encodes a negative regulator for fliC expression. Previous biochemical analysis suggested that phase variation might depend on alternative synthesis of phase-specific flagellin mRNA (H. Suzuki and T. Iino, J. Mol. Biol. 81:57-70, 1973). However, recently reported results suggested that FljA-dependent inhibition might be mediated by a posttranscriptional control mechanism (H. R. Bonifield and K. T. Hughes, J. Bacteriol. 185:3567-3574, 2003). In this study, we reexamined the mechanism of FljA-mediated inhibition of fliC expression more carefully. Northern blotting analysis revealed that no fliC mRNA was detected in phase 2 cells. However, only a moderate decrease in -galactosidase activity was observed from the fliC-lacZ transcriptional fusion gene in phase 2 cells compared with that in phase 1 cells. In contrast, the expression of the fliC-lacZ translational fusion gene was severely impaired in phase 2 cells. The half-life of fliC mRNA was shown to be much shorter in phase 2 cells than in phase 1 cells. Purified His-tagged FljA protein was shown to bind specifically to fliC mRNA and inhibit the translation from fliC mRNA in vitro. On the basis of these results, we propose that in phase 2 cells, FljA binds to fliC mRNA and inhibits its translation, which in turn facilitates its degradation.
INTRODUCTION
Salmonella enterica serovar Typhimurium has 5 to 10 flagella per cell. The individual flagellum is composed of three substructures, a basal body, a hook, and a filament (33). The filament extends into the extracellular space and is up to 10 μm in length. Although the filament is the largest substructure in the flagellum, it is composed of a single species of protein, flagellin. Serovar Typhimurium has two nonallelic flagellin genes, fliC and fljB, which encode antigenically distinct proteins. Individual cells express only one of these two flagellin genes and alternate the expression between the two at a rate of 10–3 to 10–5 per cell generation. This phenomenon is known as flagellar phase variation, and the FliC-expressing cells are called phase 1, while FljB-expressing cells are called phase 2 (19).
Flagellar phase variation is controlled by the reversible inversion of a DNA segment, called the H segment, containing the promoter for the fljB gene (55, 56). The H segment is flanked by inverted repetitious sequences, hixL and hixR, between which site-specific recombination occurs, leading to H inversion (25). The cognate recombinase, called DNA invertase, is encoded by the hin gene, which is located within the H segment (28, 29, 47). The fljB gene constitutes an operon together with the fljA gene, which encodes a negative regulator for fliC expression (12, 13, 40, 48). Therefore, when the H segment is in the "on" orientation, both the fljB and fljA genes are transcribed, resulting in phase 2 flagellin being synthesized and the fliC gene being repressed. On the other hand, when the H segment turns to the "off" orientation, neither fljB nor fljA is expressed, resulting in phase 1 flagellin being synthesized.
A pioneering study to understand the molecular mechanism of FljA-mediated inhibition of the fliC expression in phase 2 cells was performed previously by Suzuki and Iino (50). They isolated mRNAs from phase 1 and phase 2 cells separately and used them as templates in an in vitro protein-synthesizing system. They found that the flagellin molecules synthesized were of the same phase as the cells from which the mRNAs were derived. This indicated that phase 2 cells did not contain fliC mRNA, suggesting that fliC repression should be accomplished at a transcriptional level. Inoue et al. (21) isolated mutants whose fliC expression became insensitive to FljA. Their mutation sites were mapped around the Shine-Dalgarno (SD) sequence of the fliC gene, which locates far downstream of the fliC promoter. This situation is unusual because the classical operator sequences are usually located close to or within the promoter regions (9). This suggested the possibility that FljA-mediated inhibition of fliC expression might occur at a translational level. Recently, Bonifield and Hughes (2) showed that FljA reduced -galactosidase activity 200-fold from the fliC-lacZ translational fusion gene, while FljA reduced the enzyme activity from the fliC-lacZ transcriptional fusion gene by only fivefold and the steady-state level of fliC mRNA by only threefold. Based on these results, they proposed that FljA might act at both transcriptional and posttranscriptional levels.
This study was aimed at understanding the mechanism of fliC repression by FljA at a molecular level. To address this issue, we reexamined the mechanism of fliC repression by FljA more carefully. Northern blotting analysis revealed that no fliC mRNA was detected in phase 2 cells. However, only a marginal or moderate decrease in -galactosidase activity was observed from the fliC-lacZ transcriptional fusion gene in phase 2 cells. In contrast, the expression of the fliC-lacZ translational fusion gene was severely impaired in phase 2 cells. We further showed that the half-life of fliC mRNA was much shorter in phase 2 cells than in phase 1 cells. Using purified His-tagged FljA protein, FljA was shown to bind specifically to fliC mRNA and inhibit the translation in vitro from fliC mRNA. On the basis of these results, we propose the mechanism of posttranscriptional control of fliC expression by FljA in phase 2 cells.
MATERIALS AND METHODS
Bacterial strains and plasmids. Salmonella strains used in this study are listed in Table 1. LT2 has another DNA invertase gene, fin, in addition to hin (32). KK211 and KK212 are phase-locked mutants of LT2 whose H segments are fixed in the "on" and "off" orientations, respectively, due to null mutations in both hin and fin genes (32). Therefore, KK211 is fixed in phase 1 and does not express FljA, whereas KK212 is fixed in phase 2 and expresses FljA constitutively. When mRNAs transcribed from plasmid-borne fliC genes were analyzed, KK211CT and KK212CT were used as host strains. These strains carry an fliC::Tn10 mutation introduced from KK2604 by P22-mediated transduction. KK211CL and KK212CL carry a chromosomal fliC-lacZ transcriptional fusion gene introduced from KK1110 by P22-mediated transduction. The Escherichia coli strain used was JM109 (53). Plasmids used are listed in Table 2. Procedures for plasmid construction are described below.
-Galactosidase enzyme assay. The activity of -galactosidase was assayed as described previously (30, 31, 37) using cells grown to an optical density at 600 nm (OD600) of 0.4 in LB at 37°C. If necessary, tetracycline or ampicillin was added to the medium at a final concentration of 5 or 100 μg/ml, respectively. Each sample was assayed in triplicate.
Protein analysis. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and Western blotting analysis of proteins were performed according to methods described previously (31).
DNA procedures. DNA manipulation and transformation were performed as described previously (31). Unless otherwise specified, all the chemicals used were purchased from Nacalai Tesque (Kyoto, Japan). Restriction enzymes and T4 DNA ligase were purchased from Toyobo (Osaka, Japan) or Takara (Kyoto, Japan). PCR amplification was carried out with an iCycler (Bio-Rad, CA) as specified by the manufacturer by using Thermococcus kodakaraensis DNA polymerase (KOD Dash) purchased from Toyobo. Customized DNA primers were purchased from Hokkaido System Science (Sapporo, Japan). Their sequences are summarized in Table 3.
Construction of plasmids carrying fliC-lacZ translational and transcriptional fusions. Plasmid pKK1012 contains the entire fliC operon of serovar Typhimurium. Using this plasmid as a template, a DNA sequence encompassing the promoter through the last codon of the fliC gene was amplified by PCR with primers (fliC)SmaIS and (fliC)SmaIE. The amplified product was digested with SmaI and inserted into the corresponding site of pMC1871 to obtain pMCIC0. In this plasmid, the lacZ gene was translationally fused to the entire fliC open reading frame (ORF) and expressed under the transcriptional and translational control of the fliC gene. Using the same template, a DNA sequence encompassing the promoter through the termination codon of the fliC gene was PCR amplified with primers (fliC)S1(Sph)124 and (fliC)Eco3E. The amplified product was digested with SphI and EcoRI and inserted into the corresponding site of pRL124 to obtain pRLIC0. In this plasmid, the lacZ gene was transcriptionally fused to the fliC gene.
Construction of plasmids expressing fliC mRNAs of various structures. Plasmids expressing various versions of fliC mRNAs were constructed as follows. Structures of the hybrid fliC genes on the constructed plasmids are summarized in Fig. 1.
The nucleotide sequence between positions –35 and +102 (with reference to the transcription start site) of the fliC gene was PCR amplified using pKK1012 as a template and primers ICf1Bg and ICr1E. The amplified product was digested with BglII and EcoRI and inserted into the corresponding site of pTrc97A to obtain pICP1. In this plasmid, the first 102 nucleotides of fliC mRNA fused to a 313-nucleotide RNA derived from the vector sequence are expressed from the fliC promoter, and transcription terminates at the rrnB terminator.
The entire fliC coding sequence (positions +63 to +1550) was PCR amplified from pKK1012 with primers FLIC1 and FLIC9. The amplified product was digested with NcoI and BamHI and inserted into the corresponding site of pTrc99A to yield pIC495-1. In this plasmid, mRNA consisting of the 39-nucleotide 5' untranslated region (5'-UTR) of trc fused to the entire fliC ORF is expressed from the trc promoter and terminated at the rrnB terminator.
A DNA fragment (positions +1 to +1550) containing the entire 5'-UTR and ORF of the fliC gene was PCR amplified from pKK1012 with primers CifBiSa and CirBi2Sp. The amplified product was digested with SalI and SphI and inserted into the corresponding site of pBAD33 to yield pBAD1549. A 1.5-kb SacI-HindIII fragment was excised from this plasmid and inserted into the corresponding site of pTrc99A to obtain pIC495-2. In this plasmid, mRNA consisting of the 39-nucleotide trc 5'-UTR, the 62-nucleotide fliC 5'-UTR, and the entire fliC ORF is expressed from the trc promoter and terminated at the rrnB terminator.
A DNA fragment (positions –35 to +662) containing the promoter, the entire fliC 5'-UTR, and the N-terminal region of the fliC ORF was PCR amplified from pKK1012 with primers ICf1Bg and ICE2E. The amplified product was digested with BglII and EcoRI and inserted into the corresponding site of pTrc97A to obtain pICP2. A plasmid, pICP2M, in which adenine of the initiation codon of the fliC gene on pICP2 was replaced with cytosine was constructed as follows. Using two primer pairs, ICf1Bg and ICmt1 and ICmt2 and ICE2E, two DNA fragments were PCR amplified from pKK1012. These two fragments were mixed and used as templates for a second-round PCR with primers ICf1Bg and ICE2E. The amplified product was inserted into pTrc97A as described above to obtain pICP2M.
Construction of plasmids for in vitro synthesis of fliC mRNA. A 1.5-kb SacI-HindIII fragment was excised from pBAD1549 and inserted into the corresponding site of pSPT18 to yield pSPT18fliCUTR-ORF. In this plasmid, the fliC mRNA containing its own 5'-UTR and ORF is transcribed from the T7 promoter.
A 1.5-kb DNA fragment was PCR amplified from pIC495-1 with primers TRCf1Sa and CirBi2Sp. The amplified product was digested with SalI and SphI and inserted into the corresponding site of pBAD33 to yield pBAD33trcUTR-fliCORF. From this plasmid, a 1.5-kb SacI-HindIII fragment was excised and inserted into pSPT18 to obtain pSPT18trcUTR-fliCORF. This plasmid directs the synthesis of mRNA consisting of the 39-nucleotide trc 5'-UTR and the entire fliC ORF from the T7 promoter.
Purification of His-tagged FljA protein. The fljA gene was PCR amplified with primers JAf2B and JAr1S using genomic DNA of LT2 as a template. After digestion with BamHI and SalI, the amplified product was inserted into the corresponding site of pQE80L to obtain pQEfljA. In this plasmid, His6-tagged FljA (His-FljA) was expressed from the tac promoter. Strain JM109 harboring this plasmid was grown at 37°C with shaking in 40 ml of LB containing 100 μg of ampicillin/ml. When the cell growth reached an OD600 of 0.5, IPTG (isopropyl--D-thiogalactopyranoside) was added to a final concentration of 1 mM. After further cultivation for 5 h, cells were harvested by centrifugation and resuspended in 1 ml of NA buffer (50 mM NaH2PO4, 300 mM NaCl, 20 mM imidazole, pH 8.0) containing 1 mg of lysozyme (Seikagaku Kogyo, Tokyo, Japan). After incubation on ice for 30 min, the cells were disrupted by sonication. The sonicated sample was centrifuged, and the resulting supernatant was mixed with 50 μl of Ni+-nitrilotriacetic acid agarose (QIAGEN, Hilden, Germany). After the mixture was shaken gently at 4°C for 30 min, the Ni+-nitrilotriacetic acid agarose was collected by centrifugation, washed three times with 1 ml of NB buffer (50 mM NaH2PO4, 300 mM NaCl, 40 mM imidazole, pH 8.0), and resuspended in 50 μl of NC buffer (50 mM NaH2PO4, 300 mM NaCl, 250 mM imidazole, pH 8.0). After centrifugation, the supernatant fractions containing His-FljA were pooled and analyzed by SDS-PAGE. The purified His-FljA exhibited a single band at a position corresponding to approximately 21 kDa (data not shown).
Preparation of DIG-labeled RNA probes. The RNA probe complementary to the 5' portion of fliC mRNA (probe 1) was prepared as follows. A DNA fragment corresponding to the first 300 nucleotides of fliC mRNA was PCR amplified from pKK1012 with primers ICS3E and ICE4B. After digestion with EcoRI and BamHI, the amplified product was inserted into the corresponding site of pSPT18 to obtain pIC18UTR. This plasmid DNA was linearized by digestion with EcoRI and used as a template for in vitro RNA synthesis from SP6 RNA polymerase with a digoxigenin (DIG) RNA labeling kit (Roche, Basel, Switzerland). The DIG-labeled RNA was precipitated by ethanol and dissolved in diethylpyrocarbonate-treated H2O.
The plasmid for preparation of the RNA probe complementary to the region from positions +590 to +1109 of fliC mRNA (probe 2) was constructed as follows. The DNA fragment corresponding to this region was PCR amplified from pKK1012 with primers ICiPr1Bm and ICiPr2Ec. After digestion with BamHI and EcoRI, the amplified product was inserted into the corresponding site of pSPT19 to obtain pSPT19fliC. After this plasmid DNA was linearized by digestion with BamHI, DIG-labeled probe 2 was prepared as described above.
The plasmid for preparation of the RNA probe complementary to the region from positions +586 to +1138 of fljB mRNA (probe 3) was constructed as follows. The DNA fragment corresponding to this region was PCR amplified from pKK1001 with primers JBPr1Bm and JBPr2Ec. After digestion with BamHI and EcoRI, the amplified product was inserted into the corresponding site of pSPT19 to obtain pSPT19fljB. Using this plasmid, DIG-labeled probe 3 was prepared by the same procedure as that for probe 2.
Northern blotting analysis. Cells were grown at 37°C to an OD600 of 0.4 in 1.5 ml of LB. If necessary, ampicillin was added to the medium at a final concentration of 100 μg/ml. RNAs were extracted from the cells with a TRI reagent (Sigma Chemical, MO) according to the manufacturer's instructions. A mixture containing 0.5 μg of the RNA sample, 5.25% formaldehyde, and 50% formamide in MOPS buffer [20 mM 3-(N-morpholino)propanesulfonic acid, 1 mM EDTA, 5 mM sodium acetate, pH 7.0] was heated at 85°C for 5 min and cooled quickly. The RNAs were then separated electrophoretically on a 1.5% agarose gel containing 2% formamide in MOPS buffer. After electrophoresis, RNAs were transferred to a Hybond-N+ nylon membrane (Amersham Biosciences, NJ) by capillary blotting in 20x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) buffer (3 M NaCl, 0.3 M trisodium citrate, pH 7.0). After being baked at 120°C for 30 min, the membrane was treated for 2 h at 65°C with a hybridization buffer (0.2% Ficoll, 0.2% polyvinylpyrrolidone, 0.2% bovine serum albumin, 0.1 mM Na2H2PO4 [pH 6.5], 0.5% SDS, 50% formamide, 0.1 mg of salmon sperm DNA/ml in 5x SSC). The membrane was then incubated for 12 h at 65°C in the hybridization buffer containing the DIG-labeled RNA probe. After hybridization, the membrane was first washed under low-stringency conditions in 2x SSC containing 0.1% SDS for 10 min at room temperature and was then washed under high-stringency conditions in 0.2x SSC containing 0.1% SDS for 30 min at 65°C. DIG-labeled bands were visualized using the DIG luminescence detection kit (Roche) according to the manufacturer's instructions.
In vitro transcription and translation. A 611-bp DNA fragment encompassing the trc promoter, the fliC 5'-UTR, and the first 79 codons of the fliC gene was PCR amplified from pIC495-2 with primers PTRCP7 and ICE4B and used as a template for the in vitro transcription experiment. The transcription reaction mixture (50 μl) contained 16 nM template DNA, 1 unit of E. coli RNA polymerase holoenzyme (Epicenter, WI), 40 mM Tris-HCl (pH 7.5), 150 mM KCl, 10 mM MgCl2, 1 mM dithiothreitol, 50 units of RNase inhibitor (Takara), and various concentrations (0 to 1,600 nM) of His-FljA. After preincubation for 30 min at 37°C, the reaction was initiated by the addition of four nucleotides containing [-32P]UTP (Amersham Biosciences) at a final concentration of 0.2 mM. After incubation for 20 min at 37°C, the reaction was terminated by the addition of 83 μl of a stop solution (0.6 M sodium acetate [pH 5.5], 20 mM EDTA, 200 μg tRNA/ml). RNAs were precipitated by ethanol and separated on a 6% polyacrylamide gel containing 6 M urea. The labeled transcripts were detected by autoradiography.
The DNA-directed coupled transcription-translation and RNA-directed translation experiments were performed using an E. coli S30 extract (Promega, WI) according to the manufacturer's instruction as described previously (31). For the coupled transcription-translation system, the reaction mixture (70 μl) contained 11 nM DNA template linearized by HindIII digestion and 0 or 1,100 nM His-FljA. After incubation for 4 h at 37°C, proteins were precipitated by acetone and vacuum dried. Synthesized FliC proteins were detected by Western blotting using an anti-FliC antibody.
The structure of the RNA template used in the RNA-directed translation experiment is shown in Fig. 2. It was prepared as follows. DNA of pSPT18fliCUTR-ORF was linearized by digestion with HindIII and used as a template for in vitro transcription by T7 RNA polymerase (Roche). The RNA transcript was purified as described above. The translation mixture (70 μl), containing 21 nM RNA template, various concentrations (0, 210, and 2,100 nM) of His-FljA, 21 MBq [35S]methionine/ml (Amersham Biosciences), and 50 units of RNase inhibitor (Takara), was incubated for 4 h at 37°C and then mixed with 220 μl of 10 mM Tris-HCl (pH 7.5) containing 5 μg of trypsin inhibitor and 5 μl of anti-FliC antiserum. After incubation for 12 h at 4°C with gentle shaking, 5 mg of protein A-Sepharose CL-4B (Amersham Biosciences) was added to the mixture, and gentle shaking was continued for another 1 h at 4°C. The ternary complex consisting of protein A-Sepharose CL-4B, FliC, and the antibody was collected by centrifugation; washed twice with 300 μl of a wash buffer containing 50 mM Tris-HCl (pH 8.0), 0.15 M NaCl, 0.1 mM EDTA, and 0.1% lubrol; and then subjected to electrophoresis on an SDS-12% polyacrylamide gel. 35S-labeled proteins were detected by fluorography using an Amplify fluorographic reagent (Amersham Biosciences).
RNA stability analysis. Growing cells in LB were treated with rifampin at a final concentration of 200 μg/ml, and aliquots were sampled at 0, 1, 5, 10, and 20 min after rifampin addition. RNAs extracted from the cells were analyzed by Northern blotting as described above.
RNA binding assay. The structure of the RNA fragments used for the gel mobility shift assay is shown in Fig. 2. They were prepared as follows. DNAs of pSPT18fliCUTR-ORF and pSPT18trcUTR-fliCORF were linearized by digestion with HincII and used as templates for runoff transcription by T7 RNA polymerase to obtain 121- and 98-nucleotide transcripts, respectively. The former (named RNA C) contains a 5' portion (positions +1 to +100) of native fliC mRNA, and the latter (named RNA T) contains the trc 5'-UTR (positions +1 to +39) and the 5' portion of the fliC ORF (positions +63 to +100). These RNAs were gel purified, dephosphorylated with calf intestinal alkaline phosphatase (Toyobo), and then labeled at the 5' end with [-32P]ATP (Amersham Biosciences) by T4 polynucleotide kinase (Toyobo). The binding reaction mixture (20 μl) contained 0.5 nM labeled RNA, 10 mM Tris-acetate (pH 7.5), 10 mM MgCl2, 50 mM NaCl, 50 mM KCl, 10 mM dithiothreitol, 5% glycerol, and various concentrations of His-FljA. In the competition assay, unlabeled RNA competitors were added to the reaction mixture at a final concentration of 50 nM. The reaction mixture was incubated for 30 min at 37°C and subjected to electrophoresis at 4°C on native 6% polyacrylamide gels containing 5% glycerol in 0.5x TBE buffer (46 mM Tris base, 46 mM boric acid, 1 mM EDTA). Labeled RNAs were detected by autoradiography.
RESULTS
Absence of fliC mRNA in phase 2 cells. In order to resolve the confusion regarding two previously reported conflicting results on the fliC mRNA level in phase 2 cells (2, 50), we performed Northern blotting analysis of flagellin mRNAs using fliC- and fljB-specific RNA probes. The probes used were designed to be complementary to the central regions of the respective flagellin mRNAs. Since the central region corresponds to the sequence-variable domain of flagellin (26), the respective probe was expected to hybridize specifically with either fliC or fljB transcript. In hybridization with the fliC-specific probe (probe 2), an approximately 1.6-kb band was detected in the mRNA sample prepared from phase 1 cells (KK211), whereas no band was observed in the mRNA sample prepared from phase 2 cells (KK212) (Fig. 3A). Since the size of this band is equivalent to the expected size of fliC mRNA, we conclude that this band corresponds to fliC mRNA. Conversely, in the hybridization with the fljB-specific probe (probe 3), an approximately 1.6-kb band was detected in the mRNA sample prepared from phase 2 cells, whereas no band was detected in the mRNA sample prepared from phase 1 cells (Fig. 3B). The full-length size of mRNA for the fljBA operon is estimated to be approximately 2.2 kb, which is larger than the size of the band detected in Fig. 3B. It was known that a -independent terminator exists between the fljB and fljA genes and that 99% of mRNAs initiated from the fljB promoter stop transcription at this terminator, yielding a 1.6-kb RNA transcript (16). Therefore, we conclude that the 1.6-kb band hybridized with the fljB probe corresponds to the fljB transcript. These results clearly indicate that fliC mRNA is absent from phase 2 cells, which is consistent with results previously reported by Suzuki and Iino (50).
Bonifield and Hughes (2) used a T2 RNase protection assay to show that fliC mRNA was produced even in phase 2 cells. The RNA probe that they used is complementary to the first 200 nucleotides of fliC mRNA. This region is highly homologous (92.1% identity) to the corresponding region of fljB mRNA, suggesting that this probe could hybridize with not only fliC mRNA but also fljB mRNA. Therefore, we suppose that the amount of fliC mRNA in phase 2 cells might have been overestimated in their T2 protection assay. This supposition is supported by our Northern blotting analysis with probe 1, which covers the first 300 nucleotides of fliC mRNA. As expected, this probe hybridized with the 1.6-kb RNA extracted from phase 2 cells as well as that from phase 1 cells (data not shown).
Gene fusion analysis of fliC expression in the presence of FljA. The results described above led us to reexamine the effect of FljA on the expression of fliC-lacZ transcriptional and translational fusion genes. Two types of transcriptional fusions were used in this study. One was a chromosomal fliC-lacZ fusion (30), in which the Mud1(amp lac) phage DNA had been inserted into the fliC ORF after nucleotide position +941 from the transcriptional start site, and the other was a fliC-lacZ fusion gene on a plasmid, pRLIC0, in which a DNA segment (positions –35 through +1550) containing the promoter-operator region and the entire fliC ORF with its stop codon was fused transcriptionally to the lacZ gene on a promoter-probe vector, pRL124. In the chromosomal fusion, -galactosidase activity was reduced 2.5-fold in phase 2 cells compared with that in phase 1 cells (Table 4). With the fusion gene on the plasmid, a small reduction in the enzyme activity was also observed in phase 2 cells.
The fliC-lacZ translational fusion gene used was on a plasmid, pMCIC0, in which the DNA segment (positions –35 through +1547) containing the promoter-operator region and the entire ORF of fliC without its stop codon was fused translationally to the lacZ gene on a vector, pMC1871. In phase 2 cells, -galactosidase activity was reduced 10-fold compared with that in phase 1 cells (Table 4). Therefore, in the analysis with fliC-lacZ fusion genes, FljA is able to repress fliC expression at both transcriptional and translational levels, and the translational repression is more profound. These results coincide with those from the chromosomal gene fusion study reported previously by Bonifield and Hughes (2).
Effect of FljA on stability of fliC mRNA. The above-mentioned results mean that fliC mRNA is absent from phase 2 cells, although a low but considerable level of transcription of the fliC gene occurs in phase 2 cells. This raised the possibility that the fliC gene might be transcribed efficiently in phase 2 cells but that the synthesized fliC mRNA might be degraded rapidly in the presence of FljA. In order to test this possibility, we examined the effect of FljA on the stability of fliC mRNA.
Total RNA was isolated from phase 1 cells at various time points after rifampin addition and analyzed by Northern blotting with probe 2 (Fig. 4A). The half-life of the fliC mRNA calculated from the band density was approximately 5 min. In order to make it possible to detect fliC mRNA in phase 2 cells, we used cells harboring pICP1, a multicopy plasmid carrying a DNA segment between positions –35 and +102 of the fliC gene followed by the rrnB terminator, producing a 415-nucleotide mRNA transcribed from the fliC promoter. As shown in Fig. 3C, the steady-state level of the fliC mRNA encoded by pICP1 was much reduced in phase 2 cells compared with that in phase 1 cells. This indicates that the fliC DNA sequence downstream of position +103 is dispensable for regulation by FljA. However, a low but detectable amount of fliC mRNA was produced from pICP1 even in the phase 2 cells (Fig. 3C), probably because of a high level of expression of fliC mRNA from a multicopy plasmid. This enabled us to measure the half-life of fliC mRNA in the presence of FljA. In phase 1 cells, the half-life of fliC mRNA from pICP1 was approximately 5 min (Fig. 4B), which is equivalent to that of the full-length fliC mRNA transcribed from a chromosomal fliC gene. In contrast, the half-life of fliC mRNA in phase 2 cells was less than 1 min (Fig. 4C), which is much shorter than that in phase 1 cells. This result suggested that the absence of fliC mRNA in phase 2 cells might result from its loss of stability in the presence of FljA.
Effect of FljA on fliC mRNA expression from a foreign promoter. The operator mutations that rendered fliC expression insensitive to FljA were all mapped within the 5'-UTR of the fliC gene, but their positions are far downstream (positions +46 to +59) of the fliC promoter (21). This suggested that the DNA sequence around the fliC promoter is not important for regulation by FljA. To address this issue, we constructed a hybrid plasmid, pIC495-2, in which a full-length fliC mRNA including its native 5'-UTR is transcribed from the trc promoter and transcription terminates at the rrnB terminator, yielding a 1,868-nucleotide transcript (Fig. 1). Total RNA isolated from cells harboring this plasmid was analyzed by Northern blotting with probe 2. The amount of fliC mRNA was much reduced in phase 2 cells compared with that in phase 1 cells (Fig. 3D), indicating that FljA can repress the fliC gene even under the control of the trc promoter.
Next, we analyzed fliC expression using cells harboring pIC495-1 in which not only the promoter but also the entire 5'-UTR of the fliC gene had been replaced with those of trc (Fig. 1). From this plasmid, approximately equal amounts of the hybrid fliC mRNA were produced in phase 1 and phase 2 cells (Fig. 3E). Therefore, we conclude that the 5'-UTR of the fliC gene or transcript that includes the putative operator sequence plays an important role in FljA-dependent regulation.
Translational repression of the fliC gene by FljA. His6-tagged FljA (His-FljA) was purified as described in Materials and Methods and used for in vitro analysis of fliC expression. In the in vitro transcription-translation-coupled system with an E. coli S30 extract, two plasmids, pIC495-1 and pIC495-2, were used as templates. The FliC proteins synthesized were analyzed by Western blotting with anti-FliC antibody (Fig. 5A). In the absence of His-FljA, FliC was synthesized from both templates. However, when His-FljA was added at a 100-fold molar excess over the template DNA, FliC was not synthesized from pIC495-2. In contrast, FliC synthesis from pIC495-1 was not affected significantly in this condition.
In order to know whether transcription or translation of the fliC gene was inhibited by FljA, we performed an in vitro transcription experiment using E. coli RNA polymerase holoenzyme. It was found that the addition of His-FljA at up to a 100-fold molar excess over the DNA template did not affect the production of fliC transcript (Fig. 5B). Next, we performed an in vitro translation experiment with an E. coli S30 extract using in vitro-synthesized fliC mRNA as a template. FliC was synthesized in the absence of His-FljA, whereas FliC synthesis was completely inhibited in the presence of His-FljA at a 10-fold molar excess over the template RNA (Fig. 5C). In this experiment, stability of fliC mRNA was also monitored by Northern blotting. It was shown that the presence of His-FljA did not affect the stability of fliC mRNA (data not shown). These results indicate that FljA inhibits translation, and not transcription, of the fliC gene in vitro.
Stability of fliC mRNA in the absence of its translation. According to the results described above, FljA facilitates the degradation of fliC mRNA in vivo, whereas FljA inhibits the translation of fliC mRNA in vitro. This raised the possibility that FljA-enhanced degradation of fliC mRNA in vivo may result from the inhibition of its translation by FljA. In order to test this possibility, two plasmids, pICP2 and pICP2M, were constructed (Fig. 1). pICP2 encodes a C-terminally truncated FliC. In pICP2M, the initiation codon, ATG, of the fliC gene on pICP2 was replaced with CTG, which was expected to inhibit the translation of the fliC gene. Cells harboring either one of these plasmids were examined for production of fliC mRNA and FliC protein by Northern blotting with probe 1 and Western blotting with anti-FliC antibody, respectively. As expected, FliC protein was not produced from pICP2M even in phase 1 cells (Fig. 6B). It was shown that the steady-state level of fliC mRNA from pICP2M was much reduced even in phase 1 cells (Fig. 6B). This suggests that fliC mRNA is destabilized in the absence of its translation. In order to confirm this, we compared the half-lives of fliC mRNAs from pICP2 and pICP2M in phase 1 cells (Fig. 6C and D). As expected, the half-life of fliC mRNAs from pICP2M (less than 1 min) was much shorter than that from pICP2 (approximately 5 min).
Binding of FljA to fliC mRNA. DNA binding activity was not observed in the purified His-FljA protein in a gel mobility shift assay using 32P-labeled fliC DNA as a probe (data not shown). The fact that FljA inhibits the translation of fliC mRNA suggested that FljA binds to fliC mRNA. In order to test this, a gel mobility shift assay was performed using 32P-labeled fliC mRNA (RNA C) synthesized in vitro. This RNA was composed of the first 100 nucleotides of fliC mRNA containing the putative operator sequence. A shifted band was detected in the presence of 5 nM His-FljA. When 32P-labeled RNA lacking the putative operator sequence (RNA T) was used, no shifted band was observed even in the presence of 500 nM His-FljA (Fig. 7). Binding specificity was examined by a competition experiment using unlabeled RNAs as competitors. As a result, the unlabeled RNA C behaved as an effective competitor, whereas the unlabeled RNA T did not, indicating a high specificity in binding of FljA to fliC mRNA containing the operator sequence.
DISCUSSION
The dual controlling system governs flagellar phase variation of Salmonella; one part of the system is Hin-mediated inversion of the H segment containing the promoter for the fljBA operon, and the other is FljA-mediated inhibition of the fliC expression. Although the mechanism of the former system has already been well documented at a molecular level (18, 24, 25), there have been very few studies on the mechanism of the latter system. Suzuki and Iino (50) showed that phase 2 cells did not contain mRNA with FliC-synthesizing activity. Since then, FljA-mediated repression of the fliC gene has long been believed to be achieved at a transcriptional level. However, Bonifield and Hughes (2) showed evidence suggesting that this repression might occur mainly at a posttranscriptional level. This study was carried out to clarify this discrepancy. We showed that fliC mRNA was not detected in the presence of FljA, which coincides with the observation previously reported by Suzuki and Iino (50). However, in the fliC-lacZ gene fusion analysis, FljA-dependent repression was not so strong at the transcriptional level but was severe at the translational level, which conforms to the observation reported previously by Bonifield and Hughes (2). In order to explain these conflicting results, we examined the effect of FljA on the stability of fliC mRNA and found that it was much shorter in the presence of FljA than in the absence of FljA. Therefore, we can explain the above-mentioned conflicting results as follows: in phase 2 cells, transcription of the fliC gene occurs, but synthesized fliC mRNA is degraded rapidly in the presence of FljA.
Using purified His-tagged FljA protein, we showed that FljA binds specifically to fliC mRNA and inhibits the translation in vitro from fliC mRNA. Consistent with this, FljA belongs to a superfamily of Lsm ribonucleoproteins, which includes important modulators of RNA biosynthesis and function (17, 27). The binding affinity of FljA to fliC mRNA seems to be very low, because a complete shift could not be attained in a gel mobility shift assay in the presence of FljA, even at a 1,000-fold molar excess over fliC mRNA (Fig. 7). This suggests a possibility that another protein(s) may be required for maximal interaction between FljA and fliC mRNA.
Previously isolated operator mutations that rendered fliC expression insensitive to FljA defined the putative operator site around the SD sequence of the fliC gene (21). It is reasonable to postulate that this putative operator sequence acts as the FljA-binding site. This site locates within the 5'-UTR of the fliC gene and thus is transcribed into mRNA. Therefore, we propose that FljA binds to fliC mRNA at this site. Consistent with this, the fliC mRNA lacking this sequence showed no activity to bind FljA in the gel mobility shift assay (Fig. 6). Because the SD sequence should act as a ribosome-binding site (46), binding of FljA to this site may prevent ribosomes from binding to this site, which leads to translation inhibition of the fliC gene. Therefore, like some ribosomal proteins such as S4 and S15 of E. coli (10, 11, 41) and Reg protein of bacteriophage T4 (52), FljA must act as a translational repressor.
Next, we consider a possible mechanism whereby fliC mRNA is destabilized in phase 2 cells. In several genes, translating ribosomes have been known to interfere with the degradation of mRNAs (3, 6, 22, 23, 42, 54). For example, ribosomes inhibit an RNase E cleavage of rpsO mRNA in E. coli (3). By analogy with this, we propose that in phase 2 cells, FljA binds to fliC mRNA and inhibits its translation, which in turn facilitates its degradation. This is supported by our observation that an engineered untranslatable fliC mRNA was degraded rapidly even in the absence of FljA (Fig. 6). It is unlikely that FljA itself may act as an RNase responsible for the degradation of fliC mRNA, because FljA did not facilitate the degradation of fliC mRNA in vitro (data not shown). In our preliminary experiment, fliC mRNA was shown to be stabilized in the mutant defective in the RNase E gene even in the presence of FljA (our unpublished results), suggesting that the degradosome including RNase E (4) may be implicated in this regulation.
In the fliC-lacZ transcriptional fusion gene, a weak decrease in -galactosidase activity was observed in the presence of FljA (Table 4). This suggests that transcriptional repression may also be involved in FljA-mediated inhibition of fliC expression. The fljA gene is known to be transcribed inefficiently, owing to the -independent terminator between the fljB and fljA genes (16). This is also supported by our observation that a full-length fljBA mRNA was hardly observed in phase 2 cells (Fig. 3). Therefore, the cellular level of FljA should be significantly low. However, in this study, the chromosomal fljA gene was found to be able to inhibit the expression of the fliC gene on multicopy plasmids (Fig. 3). This indicates that multiple DNA copies of the operator sequence cannot efficiently titrate the intracellular FljA proteins. Consistent with this, the DNA fragment containing the operator sequence showed no activity to bind FljA in a gel mobility shift assay (data not shown). Therefore, it is unlikely that FljA may repress transcription of the fliC gene by binding to the operator site on DNA. At present, we cannot think of a special mechanism for transcriptional repression by FljA. However, there is a possibility that the fliC-lacZ hybrid mRNA transcribed from the fliC-lacZ transcriptional fusion gene may be susceptible to RNase attack due to the presence of an FljA-binding site, which may reduce -galactosidase activity to some extent in the presence of FljA.
We also showed that the chromosomal fljA gene is sufficient to inhibit fliC expression from the trc promoter, which is known to be one of the strongest promoters in bacteria (38). This suggests that overproduced fliC mRNAs cannot titrate intracellular FljA proteins either. This is not a surprising result, because fliC mRNA is degraded rapidly in the presence of FljA (Fig. 4). However, the possibility that the binding affinity of FljA to fliC mRNA is too low to titrate out the cellular FljA proteins cannot be ruled out.
In Fig. 8, we summarize the present understanding of the mechanism controlling flagellar phase variation of Salmonella. H inversion is an on-off switch of transcription of the fljBA operon, whereas FljA is a posttranscriptional regulator of the fliC operon. It is believed that posttranscriptional control enables bacteria to quickly adjust their gene expression to environmental changes (8, 14, 43, 44, 49). Flagellar filaments or flagellin subunits act as potent antigens in host organisms, and there have been several reports suggesting that flagellar phase variation might influence the virulence phenotype of Salmonella (7, 20, 45, 51). Therefore, it is possible that the mode of its genetic control may have some effect on the growth or survival of Salmonella cells in hosts. However, at present, the biological significance of Salmonella employing posttranscriptional control for FljA-mediated inhibition of fliC expression remains unknown.
Two FljA paralogues are known to exist: one is HCM2.0081, encoded by a plasmid, pHCM2, found in S. enterica serovar Typhi CT18 (39), and the other is Y1019, encoded by a plasmid, pMT1, found in Yersinia pestis KIM (35). More than 30% of their amino acids are identical to those of FljA (data not shown). Therefore, although their function has not yet been characterized, it is reasonable to postulate that they may regulate gene expression through a mechanism similar to FljA.
ACKNOWLEDGMENTS
We thank Naoya Tanaka for his assistance in construction of some plasmids used in this study. Radioisotope facilities were provided by the Advanced Science Research Center Tsushima Division, Okayama University.
REFERENCES
Amann, E., B. Ochs, and K.-J. Abel. 1988. Tightly regulated tac promoter vectors useful for the expression of unfused and fused proteins in Escherichia coli. Gene 69:301-315.
Bonifield, H. R., and K. T. Hughes. 2003. Flagellar phase variation in Salmonella enterica is mediated by a posttranscriptional control mechanism. J. Bacteriol. 185:3567-3574.
Braun, F., J. Le Derout, and P. Regnier. 1998. Ribosomes inhibit an RNase E cleavage which induces the decay of the rpsO mRNA of Escherichia coli. EMBO J. 17:4790-4797.
Carpousis, A. J. 2002. The Escherichia coli RNA degradosome: structure, function and relationship to other ribonucleolytic multienzyme complexes. Biochem. Soc. Trans. 30:150-155.
Casadaban, M., A. Martinez-Arias, S. Shapiro, and J. Chou. 1983. -Galactosidase gene fusions for analyzing gene expression in E. coli and yeast. Methods Enzymol. 100:293-308.
Cho, K.-O., and C. Yanofsky. 1986. Sequence changes preceding a Shine-Dalgarno region influence trpE mRNA translation and decay. J. Mol. Biol. 204:51-60.
Ciacci-Woolwine, F., I. C. Blomfield, S. H. Richardson, and S. B. Mizel. 1998. Salmonella flagellin induces tumor necrosis factor alpha in a human promonocytic cell line. Infect. Immun. 66:1127-1134.
Coburn, G. A., and G. A. Mackie. 1999. Degradation of mRNA in Escherichia coli: an old problem with some new twists. Prog. Nucleic Acid Res. Mol. Biol. 62:55-108.
Collado-Vides, J., B. Magasanik, and J. D. Gralla. 1991. Control site location and transcriptional regulation in Escherichia coli. Microbiol. Rev. 55:371-394.
Deckman, I. C., and D. E. Draper. 1987. S4-alpha mRNA translation regulation complex. II. Secondary structures of the RNA regulatory site in the presence and absence of S4. J. Mol. Biol. 196:323-332.
Deckman, I. C., D. E. Draper, and M. S. Thomas. 1987. S4-alpha mRNA translation regulation complex. I. Thermodynamics of formation. J. Mol. Biol. 196:313-322.
Fujita, H., S. Yamaguchi, and T. Iino. 1973. Studies on H-O variants in Salmonella in relation to phase variation. J. Gen. Microbiol. 76:127-134.
Fujita, H., S. Yamaguchi, T. Taira, and T. Iino. 1987. Isolation and genetic analysis of operator-constitutive mutants of the H1 operon in Salmonella typhimurium. J. Gen. Microbiol. 133:3071-3080.
Grunberg-Manago, M. 1999. Messenger RNA stability and its role in control of gene expression in bacteria and phages. Annu. Rev. Genet. 33:193-227.
Guzman, L. M., D. Belin, M. J. Carson, and J. Beckwith. 1995. Tight regulation, modulation, and High-level expression by vectors containing the arabinose PBAD promoter. J. Bacteriol. 177:4121-4130.
Hanafusa, T., K. Saito, A. Tominaga, and M. Enomoto. 1993. Nucleotide sequence and regulated expression of the Salmonella fljA gene encoding a repressor of the phase 1 flagellin gene. Mol. Gen. Genet. 236:260-266.
He, W., and R. Parker. 2000. Functions of Lsm proteins in mRNA degradation and splicing. Curr. Opin. Cell Biol. 12:346-350.
Heichman, K. A., and R. C. Johnson. 1990. The Hin invertasome: protein-mediated joining of distant recombination sites at the enhancer. Science 249:511-517.
Iino, T. 1969. Genetics and chemistry of bacterial flagella. Bacteriol. Rev. 33:454-475.
Ikeda, J. S., C. K. Schmitt, S. C. Darnell, P. R. Watson, J. Bispham, T. S. Wallis, D. L. Weinstein, E. S. Metcalf, P. Adams, C. D. O'Connor, and A. D. O'Brien. 2001. Flagellar phase variation of Salmonella enterica serovar Typhimurium contributes to virulence in the murine typhoid infection model but does not influence Salmonella-induced enteropathogenesis. Infect. Immun. 69:3021-3030.
Inoue, Y. H., K. Kutsukake, T. Iino, and S. Yamaguchi. 1989. Sequence analysis of operator mutants of the phase-1 flagellin-encoding gene, fliC, in Salmonella typhimurium. Gene 85:221-226.
Iost, I., and M. Dreyfus. 1995. The stability of Escherichia coli lacZ mRNA depends upon the simultaneity of it synthesis and translation. EMBO J. 14:3252-3261.
Jain, C., and N. Kleckner. 1993. IS10 mRNA stability and steady-state levels in Escherichia coli: indirect effects of translation and role of rne function. Mol. Microbiol. 9:233-247.
Johnson, R. C., A. C. Glasgow, and M. I. Simon. 1987. Spatial relationship of the Fis binding sites for Hin recombinational enhancer activity. Nature 329:462-465.
Johnson, R. C., and M. I. Simon. 1985. Hin-mediated site specific recombination requires two 26 bp recombination sites and a 60 bp recombination enhancer. Cell 41:781-791.
Joys, T. M. 1985. The covalent structure of the phase-1 flagellar filament protein of Salmonella typhimurium and its comparison with other flagellins. J. Biol. Chem. 260:15758-15761.
Kufel, J., C. Allmang, E. Petfalski, J. Beggs, and D. Tollervey. 2002. Lsm proteins are required for normal processing and stability of ribosomal RNAs. J. Biol. Chem. 278:2147-2156.
Kutsukake, K., and T. Iino. 1980. A trans-acting factor mediates inversion of a specific DNA segment in flagellar phase variation of Salmonella. Nature 284:479-481.
Kutsukake, K., and T. Iino. 1980. Inversions of specific DNA segments in flagellar phase variation of Salmonella and inversion systems of bacteriophages P1 and Mu. Proc. Natl. Acad. Sci. USA 77:7338-7341.
Kutsukake, K., and T. Iino. 1994. Role of the FliA-FlgM regulatory system on the transcriptional control of the flagellar regulon and flagellar formation in Salmonella typhimurium. J. Bacteriol. 176:3598-3605.
Kutsukake, K., S. Iyoda, K. Ohnishi, and T. Iino. 1994. Genetic and molecular analyses of the interaction between the flagellum-specific sigma and anti-sigma factors in Salmonella typhimurium. EMBO J. 13:4568-4576.
Kutsukake, K., H. Nakashima, A. Tominaga, and T. Abo. 2006. Two DNA invertases contribute to flagellar phase variation in Salmonella enterica serovar Typhimurium strain LT2. J. Bacteriol. 188:950-957.
Kutsukake, K., and T. Nambu. 2000. Bacterial flagellum: a paradigm for biogenesis of transenvelope supramolecular structures. Recent Res. Dev. Microbiol. 4:607-615.
Kutsukake, K., Y. Ohya, S. Yamaguchi, and T. Iino. 1988. Operon structure of flagellar genes in Salmonella typhimurium. Mol. Gen. Genet. 214:11-15.
Lindler, L. E., G. V. Plano, V. Burland, G. F. Mayhew, and F. R. Blattner. 1998. Complete DNA sequence and detailed analysis of the Yersinia pestis KIM5 plasmid encoding murine toxin and capsular antigen. Infect. Immun. 66:5731-5742.
Malo, M. S., and R. L. Loughlin. 1988. Promoter-detection vectors for Escherichia coli with multiple useful features. Gene 64:207-215.
Miller, J. 1972. Experiments in molecular genetics, p. 352-355. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Mulligan, M. E., J. Brosius, and W. R. McClure. 1985. Characterization in vitro of the effect of spacer length on the activity of Escherichia coli RNA polymerase at the TAC promoter. J. Biol. Chem. 260:3529-3538.
Parkhill, J., G. Dougan, K. D. James, N. R. Thomson, D. Pickard, J. Wain, C. Churcher, K. L. Mungall, S. D. Bentley, M. T. G. Holden, M. Sebaihia, S. Baker, D. Basham, K. Brooks, T. Chillingworth, P. Connerton, A. Cronin, P. Davis, R. M. Davies, L. Dowd, N. White, J. Farrar, T. Feltwell, N. Hamlin, A. Haque, T. T. Hien, S. Holroyd, K. Jagels, A. Krogh, T. S. Larsen, S. Leather, S. Moule, P. o'Gaora, C. Parry, M. Quail, K. Rutherford, M. Simmonds, J. Skelton, K. Stevens, S. Whitehead, and B. G. Barrell. 2001. Complete genome sequence of a multiple drug resistant Salmonella enterica serovar Typhi CT18. Nature 413:848-852.
Pearce, U. B., and B. A. D. Stocker. 1967. Phase variation of flagellar antigens in Salmonella: abortive transduction studies. J. Gen. Microbiol. 49:335-349.
Philippe, C., F. Eyermann, L. Benard, C. Portier, B. Ehresmann, and C. Ehresmann. 1993. Ribosomal protein S15 from Escherichia coli modulates its own translation by trapping the ribosome on the mRNA initiation loading site. Proc. Natl. Acad. Sci. USA 90:4394-4398.
Rapaport, L. R., and G. A. Mackie. 1994. Influence of translation efficiency on the stability of the mRNA for ribosomal protein S20 in Escherichia coli. J. Bacteriol. 176:992-998.
Rauhut, R., and G. Klug. 1999. mRNA degradation in bacteria. FEMS Microbiol. Rev. 23:353-370.
Regnier, P., and C. M. Arraiano. 2000. Degradation of mRNA in bacteria: emergence of ubiquitous features. Bioessays 22:235-244.
Schmitt, C. K., S. C. Darnell, and A. D. O'Brien. 1996. The attenuated phenotype of a Salmonella typhimurium flgM mutant is related to expression of FliC flagellin. J. Bacteriol. 178:2911-2915.
Shine, J., and L. Dalgarno. 1974. The 3'-terminal sequence of Escherichia coli 16S ribosomal RNA: complementarity to nonsense triplets and ribosome binding sites. Proc. Natl. Acad. Sci. USA 71:1342-1346.
Silverman, M., and M. Simon. 1980. Phase variation: genetic analysis of switching mutants. Cell 19:845-854.
Silverman, M., J. Zieg, and M. Simon. 1979. Flagellar phase variation: isolation of the rh1 gene. J. Bacteriol. 137:517-523.
Steege, D. A. 2000. Emerging features of mRNA decay in bacteria. RNA 6:1079-1090.
Suzuki, H., and T. Iino. 1973. In vitro synthesis of phase-specific flagellin of Salmonella. J. Mol. Biol. 81:57-70.
Van der Woude, M. W., and A. J. Baumler. 2004. Phase and antigenic variation in bacteria. Clin. Microbiol. Rev. 17:581-611.
Winter, R. B., L. Morrisey, P. Gauss, L. Gold, T. Hsu, and J. Karam. 1987. Bacteriophage RegA protein binds to mRNAs and prevents translation initiation. Proc. Natl. Acad. Sci. USA 84:7822-7826.
Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13 mp18 and pUC19 vectors. Gene 33:103-119.
Yarchuk, O., I. Iost, and M. Dreyfus. 1991. The relation between translation and mRNA degradation in the lacZ gene. Biochimie 73:1533-1541.
Zieg, J., M. Silverman, M. Hilmen, and M. Simon. 1977. Recombinational switch for gene expression. Science 196:170-172.
Zieg, J., and M. Simon. 1980. Analysis of the nucleotide sequence of an invertible controlling element. Proc. Natl. Acad. Sci. USA 77:4196-4200.(Shouji Yamamoto1 and Kazu)
ABSTRACT
Flagellar phase variation of Salmonella is a phenomenon where two flagellin genes, fliC (phase 1) and fljB (phase 2), are expressed alternately. This is controlled by the inversion of a DNA segment containing the promoter for the fljB gene. The fljB gene constitutes an operon with the fljA gene, which encodes a negative regulator for fliC expression. Previous biochemical analysis suggested that phase variation might depend on alternative synthesis of phase-specific flagellin mRNA (H. Suzuki and T. Iino, J. Mol. Biol. 81:57-70, 1973). However, recently reported results suggested that FljA-dependent inhibition might be mediated by a posttranscriptional control mechanism (H. R. Bonifield and K. T. Hughes, J. Bacteriol. 185:3567-3574, 2003). In this study, we reexamined the mechanism of FljA-mediated inhibition of fliC expression more carefully. Northern blotting analysis revealed that no fliC mRNA was detected in phase 2 cells. However, only a moderate decrease in -galactosidase activity was observed from the fliC-lacZ transcriptional fusion gene in phase 2 cells compared with that in phase 1 cells. In contrast, the expression of the fliC-lacZ translational fusion gene was severely impaired in phase 2 cells. The half-life of fliC mRNA was shown to be much shorter in phase 2 cells than in phase 1 cells. Purified His-tagged FljA protein was shown to bind specifically to fliC mRNA and inhibit the translation from fliC mRNA in vitro. On the basis of these results, we propose that in phase 2 cells, FljA binds to fliC mRNA and inhibits its translation, which in turn facilitates its degradation.
INTRODUCTION
Salmonella enterica serovar Typhimurium has 5 to 10 flagella per cell. The individual flagellum is composed of three substructures, a basal body, a hook, and a filament (33). The filament extends into the extracellular space and is up to 10 μm in length. Although the filament is the largest substructure in the flagellum, it is composed of a single species of protein, flagellin. Serovar Typhimurium has two nonallelic flagellin genes, fliC and fljB, which encode antigenically distinct proteins. Individual cells express only one of these two flagellin genes and alternate the expression between the two at a rate of 10–3 to 10–5 per cell generation. This phenomenon is known as flagellar phase variation, and the FliC-expressing cells are called phase 1, while FljB-expressing cells are called phase 2 (19).
Flagellar phase variation is controlled by the reversible inversion of a DNA segment, called the H segment, containing the promoter for the fljB gene (55, 56). The H segment is flanked by inverted repetitious sequences, hixL and hixR, between which site-specific recombination occurs, leading to H inversion (25). The cognate recombinase, called DNA invertase, is encoded by the hin gene, which is located within the H segment (28, 29, 47). The fljB gene constitutes an operon together with the fljA gene, which encodes a negative regulator for fliC expression (12, 13, 40, 48). Therefore, when the H segment is in the "on" orientation, both the fljB and fljA genes are transcribed, resulting in phase 2 flagellin being synthesized and the fliC gene being repressed. On the other hand, when the H segment turns to the "off" orientation, neither fljB nor fljA is expressed, resulting in phase 1 flagellin being synthesized.
A pioneering study to understand the molecular mechanism of FljA-mediated inhibition of the fliC expression in phase 2 cells was performed previously by Suzuki and Iino (50). They isolated mRNAs from phase 1 and phase 2 cells separately and used them as templates in an in vitro protein-synthesizing system. They found that the flagellin molecules synthesized were of the same phase as the cells from which the mRNAs were derived. This indicated that phase 2 cells did not contain fliC mRNA, suggesting that fliC repression should be accomplished at a transcriptional level. Inoue et al. (21) isolated mutants whose fliC expression became insensitive to FljA. Their mutation sites were mapped around the Shine-Dalgarno (SD) sequence of the fliC gene, which locates far downstream of the fliC promoter. This situation is unusual because the classical operator sequences are usually located close to or within the promoter regions (9). This suggested the possibility that FljA-mediated inhibition of fliC expression might occur at a translational level. Recently, Bonifield and Hughes (2) showed that FljA reduced -galactosidase activity 200-fold from the fliC-lacZ translational fusion gene, while FljA reduced the enzyme activity from the fliC-lacZ transcriptional fusion gene by only fivefold and the steady-state level of fliC mRNA by only threefold. Based on these results, they proposed that FljA might act at both transcriptional and posttranscriptional levels.
This study was aimed at understanding the mechanism of fliC repression by FljA at a molecular level. To address this issue, we reexamined the mechanism of fliC repression by FljA more carefully. Northern blotting analysis revealed that no fliC mRNA was detected in phase 2 cells. However, only a marginal or moderate decrease in -galactosidase activity was observed from the fliC-lacZ transcriptional fusion gene in phase 2 cells. In contrast, the expression of the fliC-lacZ translational fusion gene was severely impaired in phase 2 cells. We further showed that the half-life of fliC mRNA was much shorter in phase 2 cells than in phase 1 cells. Using purified His-tagged FljA protein, FljA was shown to bind specifically to fliC mRNA and inhibit the translation in vitro from fliC mRNA. On the basis of these results, we propose the mechanism of posttranscriptional control of fliC expression by FljA in phase 2 cells.
MATERIALS AND METHODS
Bacterial strains and plasmids. Salmonella strains used in this study are listed in Table 1. LT2 has another DNA invertase gene, fin, in addition to hin (32). KK211 and KK212 are phase-locked mutants of LT2 whose H segments are fixed in the "on" and "off" orientations, respectively, due to null mutations in both hin and fin genes (32). Therefore, KK211 is fixed in phase 1 and does not express FljA, whereas KK212 is fixed in phase 2 and expresses FljA constitutively. When mRNAs transcribed from plasmid-borne fliC genes were analyzed, KK211CT and KK212CT were used as host strains. These strains carry an fliC::Tn10 mutation introduced from KK2604 by P22-mediated transduction. KK211CL and KK212CL carry a chromosomal fliC-lacZ transcriptional fusion gene introduced from KK1110 by P22-mediated transduction. The Escherichia coli strain used was JM109 (53). Plasmids used are listed in Table 2. Procedures for plasmid construction are described below.
-Galactosidase enzyme assay. The activity of -galactosidase was assayed as described previously (30, 31, 37) using cells grown to an optical density at 600 nm (OD600) of 0.4 in LB at 37°C. If necessary, tetracycline or ampicillin was added to the medium at a final concentration of 5 or 100 μg/ml, respectively. Each sample was assayed in triplicate.
Protein analysis. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and Western blotting analysis of proteins were performed according to methods described previously (31).
DNA procedures. DNA manipulation and transformation were performed as described previously (31). Unless otherwise specified, all the chemicals used were purchased from Nacalai Tesque (Kyoto, Japan). Restriction enzymes and T4 DNA ligase were purchased from Toyobo (Osaka, Japan) or Takara (Kyoto, Japan). PCR amplification was carried out with an iCycler (Bio-Rad, CA) as specified by the manufacturer by using Thermococcus kodakaraensis DNA polymerase (KOD Dash) purchased from Toyobo. Customized DNA primers were purchased from Hokkaido System Science (Sapporo, Japan). Their sequences are summarized in Table 3.
Construction of plasmids carrying fliC-lacZ translational and transcriptional fusions. Plasmid pKK1012 contains the entire fliC operon of serovar Typhimurium. Using this plasmid as a template, a DNA sequence encompassing the promoter through the last codon of the fliC gene was amplified by PCR with primers (fliC)SmaIS and (fliC)SmaIE. The amplified product was digested with SmaI and inserted into the corresponding site of pMC1871 to obtain pMCIC0. In this plasmid, the lacZ gene was translationally fused to the entire fliC open reading frame (ORF) and expressed under the transcriptional and translational control of the fliC gene. Using the same template, a DNA sequence encompassing the promoter through the termination codon of the fliC gene was PCR amplified with primers (fliC)S1(Sph)124 and (fliC)Eco3E. The amplified product was digested with SphI and EcoRI and inserted into the corresponding site of pRL124 to obtain pRLIC0. In this plasmid, the lacZ gene was transcriptionally fused to the fliC gene.
Construction of plasmids expressing fliC mRNAs of various structures. Plasmids expressing various versions of fliC mRNAs were constructed as follows. Structures of the hybrid fliC genes on the constructed plasmids are summarized in Fig. 1.
The nucleotide sequence between positions –35 and +102 (with reference to the transcription start site) of the fliC gene was PCR amplified using pKK1012 as a template and primers ICf1Bg and ICr1E. The amplified product was digested with BglII and EcoRI and inserted into the corresponding site of pTrc97A to obtain pICP1. In this plasmid, the first 102 nucleotides of fliC mRNA fused to a 313-nucleotide RNA derived from the vector sequence are expressed from the fliC promoter, and transcription terminates at the rrnB terminator.
The entire fliC coding sequence (positions +63 to +1550) was PCR amplified from pKK1012 with primers FLIC1 and FLIC9. The amplified product was digested with NcoI and BamHI and inserted into the corresponding site of pTrc99A to yield pIC495-1. In this plasmid, mRNA consisting of the 39-nucleotide 5' untranslated region (5'-UTR) of trc fused to the entire fliC ORF is expressed from the trc promoter and terminated at the rrnB terminator.
A DNA fragment (positions +1 to +1550) containing the entire 5'-UTR and ORF of the fliC gene was PCR amplified from pKK1012 with primers CifBiSa and CirBi2Sp. The amplified product was digested with SalI and SphI and inserted into the corresponding site of pBAD33 to yield pBAD1549. A 1.5-kb SacI-HindIII fragment was excised from this plasmid and inserted into the corresponding site of pTrc99A to obtain pIC495-2. In this plasmid, mRNA consisting of the 39-nucleotide trc 5'-UTR, the 62-nucleotide fliC 5'-UTR, and the entire fliC ORF is expressed from the trc promoter and terminated at the rrnB terminator.
A DNA fragment (positions –35 to +662) containing the promoter, the entire fliC 5'-UTR, and the N-terminal region of the fliC ORF was PCR amplified from pKK1012 with primers ICf1Bg and ICE2E. The amplified product was digested with BglII and EcoRI and inserted into the corresponding site of pTrc97A to obtain pICP2. A plasmid, pICP2M, in which adenine of the initiation codon of the fliC gene on pICP2 was replaced with cytosine was constructed as follows. Using two primer pairs, ICf1Bg and ICmt1 and ICmt2 and ICE2E, two DNA fragments were PCR amplified from pKK1012. These two fragments were mixed and used as templates for a second-round PCR with primers ICf1Bg and ICE2E. The amplified product was inserted into pTrc97A as described above to obtain pICP2M.
Construction of plasmids for in vitro synthesis of fliC mRNA. A 1.5-kb SacI-HindIII fragment was excised from pBAD1549 and inserted into the corresponding site of pSPT18 to yield pSPT18fliCUTR-ORF. In this plasmid, the fliC mRNA containing its own 5'-UTR and ORF is transcribed from the T7 promoter.
A 1.5-kb DNA fragment was PCR amplified from pIC495-1 with primers TRCf1Sa and CirBi2Sp. The amplified product was digested with SalI and SphI and inserted into the corresponding site of pBAD33 to yield pBAD33trcUTR-fliCORF. From this plasmid, a 1.5-kb SacI-HindIII fragment was excised and inserted into pSPT18 to obtain pSPT18trcUTR-fliCORF. This plasmid directs the synthesis of mRNA consisting of the 39-nucleotide trc 5'-UTR and the entire fliC ORF from the T7 promoter.
Purification of His-tagged FljA protein. The fljA gene was PCR amplified with primers JAf2B and JAr1S using genomic DNA of LT2 as a template. After digestion with BamHI and SalI, the amplified product was inserted into the corresponding site of pQE80L to obtain pQEfljA. In this plasmid, His6-tagged FljA (His-FljA) was expressed from the tac promoter. Strain JM109 harboring this plasmid was grown at 37°C with shaking in 40 ml of LB containing 100 μg of ampicillin/ml. When the cell growth reached an OD600 of 0.5, IPTG (isopropyl--D-thiogalactopyranoside) was added to a final concentration of 1 mM. After further cultivation for 5 h, cells were harvested by centrifugation and resuspended in 1 ml of NA buffer (50 mM NaH2PO4, 300 mM NaCl, 20 mM imidazole, pH 8.0) containing 1 mg of lysozyme (Seikagaku Kogyo, Tokyo, Japan). After incubation on ice for 30 min, the cells were disrupted by sonication. The sonicated sample was centrifuged, and the resulting supernatant was mixed with 50 μl of Ni+-nitrilotriacetic acid agarose (QIAGEN, Hilden, Germany). After the mixture was shaken gently at 4°C for 30 min, the Ni+-nitrilotriacetic acid agarose was collected by centrifugation, washed three times with 1 ml of NB buffer (50 mM NaH2PO4, 300 mM NaCl, 40 mM imidazole, pH 8.0), and resuspended in 50 μl of NC buffer (50 mM NaH2PO4, 300 mM NaCl, 250 mM imidazole, pH 8.0). After centrifugation, the supernatant fractions containing His-FljA were pooled and analyzed by SDS-PAGE. The purified His-FljA exhibited a single band at a position corresponding to approximately 21 kDa (data not shown).
Preparation of DIG-labeled RNA probes. The RNA probe complementary to the 5' portion of fliC mRNA (probe 1) was prepared as follows. A DNA fragment corresponding to the first 300 nucleotides of fliC mRNA was PCR amplified from pKK1012 with primers ICS3E and ICE4B. After digestion with EcoRI and BamHI, the amplified product was inserted into the corresponding site of pSPT18 to obtain pIC18UTR. This plasmid DNA was linearized by digestion with EcoRI and used as a template for in vitro RNA synthesis from SP6 RNA polymerase with a digoxigenin (DIG) RNA labeling kit (Roche, Basel, Switzerland). The DIG-labeled RNA was precipitated by ethanol and dissolved in diethylpyrocarbonate-treated H2O.
The plasmid for preparation of the RNA probe complementary to the region from positions +590 to +1109 of fliC mRNA (probe 2) was constructed as follows. The DNA fragment corresponding to this region was PCR amplified from pKK1012 with primers ICiPr1Bm and ICiPr2Ec. After digestion with BamHI and EcoRI, the amplified product was inserted into the corresponding site of pSPT19 to obtain pSPT19fliC. After this plasmid DNA was linearized by digestion with BamHI, DIG-labeled probe 2 was prepared as described above.
The plasmid for preparation of the RNA probe complementary to the region from positions +586 to +1138 of fljB mRNA (probe 3) was constructed as follows. The DNA fragment corresponding to this region was PCR amplified from pKK1001 with primers JBPr1Bm and JBPr2Ec. After digestion with BamHI and EcoRI, the amplified product was inserted into the corresponding site of pSPT19 to obtain pSPT19fljB. Using this plasmid, DIG-labeled probe 3 was prepared by the same procedure as that for probe 2.
Northern blotting analysis. Cells were grown at 37°C to an OD600 of 0.4 in 1.5 ml of LB. If necessary, ampicillin was added to the medium at a final concentration of 100 μg/ml. RNAs were extracted from the cells with a TRI reagent (Sigma Chemical, MO) according to the manufacturer's instructions. A mixture containing 0.5 μg of the RNA sample, 5.25% formaldehyde, and 50% formamide in MOPS buffer [20 mM 3-(N-morpholino)propanesulfonic acid, 1 mM EDTA, 5 mM sodium acetate, pH 7.0] was heated at 85°C for 5 min and cooled quickly. The RNAs were then separated electrophoretically on a 1.5% agarose gel containing 2% formamide in MOPS buffer. After electrophoresis, RNAs were transferred to a Hybond-N+ nylon membrane (Amersham Biosciences, NJ) by capillary blotting in 20x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) buffer (3 M NaCl, 0.3 M trisodium citrate, pH 7.0). After being baked at 120°C for 30 min, the membrane was treated for 2 h at 65°C with a hybridization buffer (0.2% Ficoll, 0.2% polyvinylpyrrolidone, 0.2% bovine serum albumin, 0.1 mM Na2H2PO4 [pH 6.5], 0.5% SDS, 50% formamide, 0.1 mg of salmon sperm DNA/ml in 5x SSC). The membrane was then incubated for 12 h at 65°C in the hybridization buffer containing the DIG-labeled RNA probe. After hybridization, the membrane was first washed under low-stringency conditions in 2x SSC containing 0.1% SDS for 10 min at room temperature and was then washed under high-stringency conditions in 0.2x SSC containing 0.1% SDS for 30 min at 65°C. DIG-labeled bands were visualized using the DIG luminescence detection kit (Roche) according to the manufacturer's instructions.
In vitro transcription and translation. A 611-bp DNA fragment encompassing the trc promoter, the fliC 5'-UTR, and the first 79 codons of the fliC gene was PCR amplified from pIC495-2 with primers PTRCP7 and ICE4B and used as a template for the in vitro transcription experiment. The transcription reaction mixture (50 μl) contained 16 nM template DNA, 1 unit of E. coli RNA polymerase holoenzyme (Epicenter, WI), 40 mM Tris-HCl (pH 7.5), 150 mM KCl, 10 mM MgCl2, 1 mM dithiothreitol, 50 units of RNase inhibitor (Takara), and various concentrations (0 to 1,600 nM) of His-FljA. After preincubation for 30 min at 37°C, the reaction was initiated by the addition of four nucleotides containing [-32P]UTP (Amersham Biosciences) at a final concentration of 0.2 mM. After incubation for 20 min at 37°C, the reaction was terminated by the addition of 83 μl of a stop solution (0.6 M sodium acetate [pH 5.5], 20 mM EDTA, 200 μg tRNA/ml). RNAs were precipitated by ethanol and separated on a 6% polyacrylamide gel containing 6 M urea. The labeled transcripts were detected by autoradiography.
The DNA-directed coupled transcription-translation and RNA-directed translation experiments were performed using an E. coli S30 extract (Promega, WI) according to the manufacturer's instruction as described previously (31). For the coupled transcription-translation system, the reaction mixture (70 μl) contained 11 nM DNA template linearized by HindIII digestion and 0 or 1,100 nM His-FljA. After incubation for 4 h at 37°C, proteins were precipitated by acetone and vacuum dried. Synthesized FliC proteins were detected by Western blotting using an anti-FliC antibody.
The structure of the RNA template used in the RNA-directed translation experiment is shown in Fig. 2. It was prepared as follows. DNA of pSPT18fliCUTR-ORF was linearized by digestion with HindIII and used as a template for in vitro transcription by T7 RNA polymerase (Roche). The RNA transcript was purified as described above. The translation mixture (70 μl), containing 21 nM RNA template, various concentrations (0, 210, and 2,100 nM) of His-FljA, 21 MBq [35S]methionine/ml (Amersham Biosciences), and 50 units of RNase inhibitor (Takara), was incubated for 4 h at 37°C and then mixed with 220 μl of 10 mM Tris-HCl (pH 7.5) containing 5 μg of trypsin inhibitor and 5 μl of anti-FliC antiserum. After incubation for 12 h at 4°C with gentle shaking, 5 mg of protein A-Sepharose CL-4B (Amersham Biosciences) was added to the mixture, and gentle shaking was continued for another 1 h at 4°C. The ternary complex consisting of protein A-Sepharose CL-4B, FliC, and the antibody was collected by centrifugation; washed twice with 300 μl of a wash buffer containing 50 mM Tris-HCl (pH 8.0), 0.15 M NaCl, 0.1 mM EDTA, and 0.1% lubrol; and then subjected to electrophoresis on an SDS-12% polyacrylamide gel. 35S-labeled proteins were detected by fluorography using an Amplify fluorographic reagent (Amersham Biosciences).
RNA stability analysis. Growing cells in LB were treated with rifampin at a final concentration of 200 μg/ml, and aliquots were sampled at 0, 1, 5, 10, and 20 min after rifampin addition. RNAs extracted from the cells were analyzed by Northern blotting as described above.
RNA binding assay. The structure of the RNA fragments used for the gel mobility shift assay is shown in Fig. 2. They were prepared as follows. DNAs of pSPT18fliCUTR-ORF and pSPT18trcUTR-fliCORF were linearized by digestion with HincII and used as templates for runoff transcription by T7 RNA polymerase to obtain 121- and 98-nucleotide transcripts, respectively. The former (named RNA C) contains a 5' portion (positions +1 to +100) of native fliC mRNA, and the latter (named RNA T) contains the trc 5'-UTR (positions +1 to +39) and the 5' portion of the fliC ORF (positions +63 to +100). These RNAs were gel purified, dephosphorylated with calf intestinal alkaline phosphatase (Toyobo), and then labeled at the 5' end with [-32P]ATP (Amersham Biosciences) by T4 polynucleotide kinase (Toyobo). The binding reaction mixture (20 μl) contained 0.5 nM labeled RNA, 10 mM Tris-acetate (pH 7.5), 10 mM MgCl2, 50 mM NaCl, 50 mM KCl, 10 mM dithiothreitol, 5% glycerol, and various concentrations of His-FljA. In the competition assay, unlabeled RNA competitors were added to the reaction mixture at a final concentration of 50 nM. The reaction mixture was incubated for 30 min at 37°C and subjected to electrophoresis at 4°C on native 6% polyacrylamide gels containing 5% glycerol in 0.5x TBE buffer (46 mM Tris base, 46 mM boric acid, 1 mM EDTA). Labeled RNAs were detected by autoradiography.
RESULTS
Absence of fliC mRNA in phase 2 cells. In order to resolve the confusion regarding two previously reported conflicting results on the fliC mRNA level in phase 2 cells (2, 50), we performed Northern blotting analysis of flagellin mRNAs using fliC- and fljB-specific RNA probes. The probes used were designed to be complementary to the central regions of the respective flagellin mRNAs. Since the central region corresponds to the sequence-variable domain of flagellin (26), the respective probe was expected to hybridize specifically with either fliC or fljB transcript. In hybridization with the fliC-specific probe (probe 2), an approximately 1.6-kb band was detected in the mRNA sample prepared from phase 1 cells (KK211), whereas no band was observed in the mRNA sample prepared from phase 2 cells (KK212) (Fig. 3A). Since the size of this band is equivalent to the expected size of fliC mRNA, we conclude that this band corresponds to fliC mRNA. Conversely, in the hybridization with the fljB-specific probe (probe 3), an approximately 1.6-kb band was detected in the mRNA sample prepared from phase 2 cells, whereas no band was detected in the mRNA sample prepared from phase 1 cells (Fig. 3B). The full-length size of mRNA for the fljBA operon is estimated to be approximately 2.2 kb, which is larger than the size of the band detected in Fig. 3B. It was known that a -independent terminator exists between the fljB and fljA genes and that 99% of mRNAs initiated from the fljB promoter stop transcription at this terminator, yielding a 1.6-kb RNA transcript (16). Therefore, we conclude that the 1.6-kb band hybridized with the fljB probe corresponds to the fljB transcript. These results clearly indicate that fliC mRNA is absent from phase 2 cells, which is consistent with results previously reported by Suzuki and Iino (50).
Bonifield and Hughes (2) used a T2 RNase protection assay to show that fliC mRNA was produced even in phase 2 cells. The RNA probe that they used is complementary to the first 200 nucleotides of fliC mRNA. This region is highly homologous (92.1% identity) to the corresponding region of fljB mRNA, suggesting that this probe could hybridize with not only fliC mRNA but also fljB mRNA. Therefore, we suppose that the amount of fliC mRNA in phase 2 cells might have been overestimated in their T2 protection assay. This supposition is supported by our Northern blotting analysis with probe 1, which covers the first 300 nucleotides of fliC mRNA. As expected, this probe hybridized with the 1.6-kb RNA extracted from phase 2 cells as well as that from phase 1 cells (data not shown).
Gene fusion analysis of fliC expression in the presence of FljA. The results described above led us to reexamine the effect of FljA on the expression of fliC-lacZ transcriptional and translational fusion genes. Two types of transcriptional fusions were used in this study. One was a chromosomal fliC-lacZ fusion (30), in which the Mud1(amp lac) phage DNA had been inserted into the fliC ORF after nucleotide position +941 from the transcriptional start site, and the other was a fliC-lacZ fusion gene on a plasmid, pRLIC0, in which a DNA segment (positions –35 through +1550) containing the promoter-operator region and the entire fliC ORF with its stop codon was fused transcriptionally to the lacZ gene on a promoter-probe vector, pRL124. In the chromosomal fusion, -galactosidase activity was reduced 2.5-fold in phase 2 cells compared with that in phase 1 cells (Table 4). With the fusion gene on the plasmid, a small reduction in the enzyme activity was also observed in phase 2 cells.
The fliC-lacZ translational fusion gene used was on a plasmid, pMCIC0, in which the DNA segment (positions –35 through +1547) containing the promoter-operator region and the entire ORF of fliC without its stop codon was fused translationally to the lacZ gene on a vector, pMC1871. In phase 2 cells, -galactosidase activity was reduced 10-fold compared with that in phase 1 cells (Table 4). Therefore, in the analysis with fliC-lacZ fusion genes, FljA is able to repress fliC expression at both transcriptional and translational levels, and the translational repression is more profound. These results coincide with those from the chromosomal gene fusion study reported previously by Bonifield and Hughes (2).
Effect of FljA on stability of fliC mRNA. The above-mentioned results mean that fliC mRNA is absent from phase 2 cells, although a low but considerable level of transcription of the fliC gene occurs in phase 2 cells. This raised the possibility that the fliC gene might be transcribed efficiently in phase 2 cells but that the synthesized fliC mRNA might be degraded rapidly in the presence of FljA. In order to test this possibility, we examined the effect of FljA on the stability of fliC mRNA.
Total RNA was isolated from phase 1 cells at various time points after rifampin addition and analyzed by Northern blotting with probe 2 (Fig. 4A). The half-life of the fliC mRNA calculated from the band density was approximately 5 min. In order to make it possible to detect fliC mRNA in phase 2 cells, we used cells harboring pICP1, a multicopy plasmid carrying a DNA segment between positions –35 and +102 of the fliC gene followed by the rrnB terminator, producing a 415-nucleotide mRNA transcribed from the fliC promoter. As shown in Fig. 3C, the steady-state level of the fliC mRNA encoded by pICP1 was much reduced in phase 2 cells compared with that in phase 1 cells. This indicates that the fliC DNA sequence downstream of position +103 is dispensable for regulation by FljA. However, a low but detectable amount of fliC mRNA was produced from pICP1 even in the phase 2 cells (Fig. 3C), probably because of a high level of expression of fliC mRNA from a multicopy plasmid. This enabled us to measure the half-life of fliC mRNA in the presence of FljA. In phase 1 cells, the half-life of fliC mRNA from pICP1 was approximately 5 min (Fig. 4B), which is equivalent to that of the full-length fliC mRNA transcribed from a chromosomal fliC gene. In contrast, the half-life of fliC mRNA in phase 2 cells was less than 1 min (Fig. 4C), which is much shorter than that in phase 1 cells. This result suggested that the absence of fliC mRNA in phase 2 cells might result from its loss of stability in the presence of FljA.
Effect of FljA on fliC mRNA expression from a foreign promoter. The operator mutations that rendered fliC expression insensitive to FljA were all mapped within the 5'-UTR of the fliC gene, but their positions are far downstream (positions +46 to +59) of the fliC promoter (21). This suggested that the DNA sequence around the fliC promoter is not important for regulation by FljA. To address this issue, we constructed a hybrid plasmid, pIC495-2, in which a full-length fliC mRNA including its native 5'-UTR is transcribed from the trc promoter and transcription terminates at the rrnB terminator, yielding a 1,868-nucleotide transcript (Fig. 1). Total RNA isolated from cells harboring this plasmid was analyzed by Northern blotting with probe 2. The amount of fliC mRNA was much reduced in phase 2 cells compared with that in phase 1 cells (Fig. 3D), indicating that FljA can repress the fliC gene even under the control of the trc promoter.
Next, we analyzed fliC expression using cells harboring pIC495-1 in which not only the promoter but also the entire 5'-UTR of the fliC gene had been replaced with those of trc (Fig. 1). From this plasmid, approximately equal amounts of the hybrid fliC mRNA were produced in phase 1 and phase 2 cells (Fig. 3E). Therefore, we conclude that the 5'-UTR of the fliC gene or transcript that includes the putative operator sequence plays an important role in FljA-dependent regulation.
Translational repression of the fliC gene by FljA. His6-tagged FljA (His-FljA) was purified as described in Materials and Methods and used for in vitro analysis of fliC expression. In the in vitro transcription-translation-coupled system with an E. coli S30 extract, two plasmids, pIC495-1 and pIC495-2, were used as templates. The FliC proteins synthesized were analyzed by Western blotting with anti-FliC antibody (Fig. 5A). In the absence of His-FljA, FliC was synthesized from both templates. However, when His-FljA was added at a 100-fold molar excess over the template DNA, FliC was not synthesized from pIC495-2. In contrast, FliC synthesis from pIC495-1 was not affected significantly in this condition.
In order to know whether transcription or translation of the fliC gene was inhibited by FljA, we performed an in vitro transcription experiment using E. coli RNA polymerase holoenzyme. It was found that the addition of His-FljA at up to a 100-fold molar excess over the DNA template did not affect the production of fliC transcript (Fig. 5B). Next, we performed an in vitro translation experiment with an E. coli S30 extract using in vitro-synthesized fliC mRNA as a template. FliC was synthesized in the absence of His-FljA, whereas FliC synthesis was completely inhibited in the presence of His-FljA at a 10-fold molar excess over the template RNA (Fig. 5C). In this experiment, stability of fliC mRNA was also monitored by Northern blotting. It was shown that the presence of His-FljA did not affect the stability of fliC mRNA (data not shown). These results indicate that FljA inhibits translation, and not transcription, of the fliC gene in vitro.
Stability of fliC mRNA in the absence of its translation. According to the results described above, FljA facilitates the degradation of fliC mRNA in vivo, whereas FljA inhibits the translation of fliC mRNA in vitro. This raised the possibility that FljA-enhanced degradation of fliC mRNA in vivo may result from the inhibition of its translation by FljA. In order to test this possibility, two plasmids, pICP2 and pICP2M, were constructed (Fig. 1). pICP2 encodes a C-terminally truncated FliC. In pICP2M, the initiation codon, ATG, of the fliC gene on pICP2 was replaced with CTG, which was expected to inhibit the translation of the fliC gene. Cells harboring either one of these plasmids were examined for production of fliC mRNA and FliC protein by Northern blotting with probe 1 and Western blotting with anti-FliC antibody, respectively. As expected, FliC protein was not produced from pICP2M even in phase 1 cells (Fig. 6B). It was shown that the steady-state level of fliC mRNA from pICP2M was much reduced even in phase 1 cells (Fig. 6B). This suggests that fliC mRNA is destabilized in the absence of its translation. In order to confirm this, we compared the half-lives of fliC mRNAs from pICP2 and pICP2M in phase 1 cells (Fig. 6C and D). As expected, the half-life of fliC mRNAs from pICP2M (less than 1 min) was much shorter than that from pICP2 (approximately 5 min).
Binding of FljA to fliC mRNA. DNA binding activity was not observed in the purified His-FljA protein in a gel mobility shift assay using 32P-labeled fliC DNA as a probe (data not shown). The fact that FljA inhibits the translation of fliC mRNA suggested that FljA binds to fliC mRNA. In order to test this, a gel mobility shift assay was performed using 32P-labeled fliC mRNA (RNA C) synthesized in vitro. This RNA was composed of the first 100 nucleotides of fliC mRNA containing the putative operator sequence. A shifted band was detected in the presence of 5 nM His-FljA. When 32P-labeled RNA lacking the putative operator sequence (RNA T) was used, no shifted band was observed even in the presence of 500 nM His-FljA (Fig. 7). Binding specificity was examined by a competition experiment using unlabeled RNAs as competitors. As a result, the unlabeled RNA C behaved as an effective competitor, whereas the unlabeled RNA T did not, indicating a high specificity in binding of FljA to fliC mRNA containing the operator sequence.
DISCUSSION
The dual controlling system governs flagellar phase variation of Salmonella; one part of the system is Hin-mediated inversion of the H segment containing the promoter for the fljBA operon, and the other is FljA-mediated inhibition of the fliC expression. Although the mechanism of the former system has already been well documented at a molecular level (18, 24, 25), there have been very few studies on the mechanism of the latter system. Suzuki and Iino (50) showed that phase 2 cells did not contain mRNA with FliC-synthesizing activity. Since then, FljA-mediated repression of the fliC gene has long been believed to be achieved at a transcriptional level. However, Bonifield and Hughes (2) showed evidence suggesting that this repression might occur mainly at a posttranscriptional level. This study was carried out to clarify this discrepancy. We showed that fliC mRNA was not detected in the presence of FljA, which coincides with the observation previously reported by Suzuki and Iino (50). However, in the fliC-lacZ gene fusion analysis, FljA-dependent repression was not so strong at the transcriptional level but was severe at the translational level, which conforms to the observation reported previously by Bonifield and Hughes (2). In order to explain these conflicting results, we examined the effect of FljA on the stability of fliC mRNA and found that it was much shorter in the presence of FljA than in the absence of FljA. Therefore, we can explain the above-mentioned conflicting results as follows: in phase 2 cells, transcription of the fliC gene occurs, but synthesized fliC mRNA is degraded rapidly in the presence of FljA.
Using purified His-tagged FljA protein, we showed that FljA binds specifically to fliC mRNA and inhibits the translation in vitro from fliC mRNA. Consistent with this, FljA belongs to a superfamily of Lsm ribonucleoproteins, which includes important modulators of RNA biosynthesis and function (17, 27). The binding affinity of FljA to fliC mRNA seems to be very low, because a complete shift could not be attained in a gel mobility shift assay in the presence of FljA, even at a 1,000-fold molar excess over fliC mRNA (Fig. 7). This suggests a possibility that another protein(s) may be required for maximal interaction between FljA and fliC mRNA.
Previously isolated operator mutations that rendered fliC expression insensitive to FljA defined the putative operator site around the SD sequence of the fliC gene (21). It is reasonable to postulate that this putative operator sequence acts as the FljA-binding site. This site locates within the 5'-UTR of the fliC gene and thus is transcribed into mRNA. Therefore, we propose that FljA binds to fliC mRNA at this site. Consistent with this, the fliC mRNA lacking this sequence showed no activity to bind FljA in the gel mobility shift assay (Fig. 6). Because the SD sequence should act as a ribosome-binding site (46), binding of FljA to this site may prevent ribosomes from binding to this site, which leads to translation inhibition of the fliC gene. Therefore, like some ribosomal proteins such as S4 and S15 of E. coli (10, 11, 41) and Reg protein of bacteriophage T4 (52), FljA must act as a translational repressor.
Next, we consider a possible mechanism whereby fliC mRNA is destabilized in phase 2 cells. In several genes, translating ribosomes have been known to interfere with the degradation of mRNAs (3, 6, 22, 23, 42, 54). For example, ribosomes inhibit an RNase E cleavage of rpsO mRNA in E. coli (3). By analogy with this, we propose that in phase 2 cells, FljA binds to fliC mRNA and inhibits its translation, which in turn facilitates its degradation. This is supported by our observation that an engineered untranslatable fliC mRNA was degraded rapidly even in the absence of FljA (Fig. 6). It is unlikely that FljA itself may act as an RNase responsible for the degradation of fliC mRNA, because FljA did not facilitate the degradation of fliC mRNA in vitro (data not shown). In our preliminary experiment, fliC mRNA was shown to be stabilized in the mutant defective in the RNase E gene even in the presence of FljA (our unpublished results), suggesting that the degradosome including RNase E (4) may be implicated in this regulation.
In the fliC-lacZ transcriptional fusion gene, a weak decrease in -galactosidase activity was observed in the presence of FljA (Table 4). This suggests that transcriptional repression may also be involved in FljA-mediated inhibition of fliC expression. The fljA gene is known to be transcribed inefficiently, owing to the -independent terminator between the fljB and fljA genes (16). This is also supported by our observation that a full-length fljBA mRNA was hardly observed in phase 2 cells (Fig. 3). Therefore, the cellular level of FljA should be significantly low. However, in this study, the chromosomal fljA gene was found to be able to inhibit the expression of the fliC gene on multicopy plasmids (Fig. 3). This indicates that multiple DNA copies of the operator sequence cannot efficiently titrate the intracellular FljA proteins. Consistent with this, the DNA fragment containing the operator sequence showed no activity to bind FljA in a gel mobility shift assay (data not shown). Therefore, it is unlikely that FljA may repress transcription of the fliC gene by binding to the operator site on DNA. At present, we cannot think of a special mechanism for transcriptional repression by FljA. However, there is a possibility that the fliC-lacZ hybrid mRNA transcribed from the fliC-lacZ transcriptional fusion gene may be susceptible to RNase attack due to the presence of an FljA-binding site, which may reduce -galactosidase activity to some extent in the presence of FljA.
We also showed that the chromosomal fljA gene is sufficient to inhibit fliC expression from the trc promoter, which is known to be one of the strongest promoters in bacteria (38). This suggests that overproduced fliC mRNAs cannot titrate intracellular FljA proteins either. This is not a surprising result, because fliC mRNA is degraded rapidly in the presence of FljA (Fig. 4). However, the possibility that the binding affinity of FljA to fliC mRNA is too low to titrate out the cellular FljA proteins cannot be ruled out.
In Fig. 8, we summarize the present understanding of the mechanism controlling flagellar phase variation of Salmonella. H inversion is an on-off switch of transcription of the fljBA operon, whereas FljA is a posttranscriptional regulator of the fliC operon. It is believed that posttranscriptional control enables bacteria to quickly adjust their gene expression to environmental changes (8, 14, 43, 44, 49). Flagellar filaments or flagellin subunits act as potent antigens in host organisms, and there have been several reports suggesting that flagellar phase variation might influence the virulence phenotype of Salmonella (7, 20, 45, 51). Therefore, it is possible that the mode of its genetic control may have some effect on the growth or survival of Salmonella cells in hosts. However, at present, the biological significance of Salmonella employing posttranscriptional control for FljA-mediated inhibition of fliC expression remains unknown.
Two FljA paralogues are known to exist: one is HCM2.0081, encoded by a plasmid, pHCM2, found in S. enterica serovar Typhi CT18 (39), and the other is Y1019, encoded by a plasmid, pMT1, found in Yersinia pestis KIM (35). More than 30% of their amino acids are identical to those of FljA (data not shown). Therefore, although their function has not yet been characterized, it is reasonable to postulate that they may regulate gene expression through a mechanism similar to FljA.
ACKNOWLEDGMENTS
We thank Naoya Tanaka for his assistance in construction of some plasmids used in this study. Radioisotope facilities were provided by the Advanced Science Research Center Tsushima Division, Okayama University.
REFERENCES
Amann, E., B. Ochs, and K.-J. Abel. 1988. Tightly regulated tac promoter vectors useful for the expression of unfused and fused proteins in Escherichia coli. Gene 69:301-315.
Bonifield, H. R., and K. T. Hughes. 2003. Flagellar phase variation in Salmonella enterica is mediated by a posttranscriptional control mechanism. J. Bacteriol. 185:3567-3574.
Braun, F., J. Le Derout, and P. Regnier. 1998. Ribosomes inhibit an RNase E cleavage which induces the decay of the rpsO mRNA of Escherichia coli. EMBO J. 17:4790-4797.
Carpousis, A. J. 2002. The Escherichia coli RNA degradosome: structure, function and relationship to other ribonucleolytic multienzyme complexes. Biochem. Soc. Trans. 30:150-155.
Casadaban, M., A. Martinez-Arias, S. Shapiro, and J. Chou. 1983. -Galactosidase gene fusions for analyzing gene expression in E. coli and yeast. Methods Enzymol. 100:293-308.
Cho, K.-O., and C. Yanofsky. 1986. Sequence changes preceding a Shine-Dalgarno region influence trpE mRNA translation and decay. J. Mol. Biol. 204:51-60.
Ciacci-Woolwine, F., I. C. Blomfield, S. H. Richardson, and S. B. Mizel. 1998. Salmonella flagellin induces tumor necrosis factor alpha in a human promonocytic cell line. Infect. Immun. 66:1127-1134.
Coburn, G. A., and G. A. Mackie. 1999. Degradation of mRNA in Escherichia coli: an old problem with some new twists. Prog. Nucleic Acid Res. Mol. Biol. 62:55-108.
Collado-Vides, J., B. Magasanik, and J. D. Gralla. 1991. Control site location and transcriptional regulation in Escherichia coli. Microbiol. Rev. 55:371-394.
Deckman, I. C., and D. E. Draper. 1987. S4-alpha mRNA translation regulation complex. II. Secondary structures of the RNA regulatory site in the presence and absence of S4. J. Mol. Biol. 196:323-332.
Deckman, I. C., D. E. Draper, and M. S. Thomas. 1987. S4-alpha mRNA translation regulation complex. I. Thermodynamics of formation. J. Mol. Biol. 196:313-322.
Fujita, H., S. Yamaguchi, and T. Iino. 1973. Studies on H-O variants in Salmonella in relation to phase variation. J. Gen. Microbiol. 76:127-134.
Fujita, H., S. Yamaguchi, T. Taira, and T. Iino. 1987. Isolation and genetic analysis of operator-constitutive mutants of the H1 operon in Salmonella typhimurium. J. Gen. Microbiol. 133:3071-3080.
Grunberg-Manago, M. 1999. Messenger RNA stability and its role in control of gene expression in bacteria and phages. Annu. Rev. Genet. 33:193-227.
Guzman, L. M., D. Belin, M. J. Carson, and J. Beckwith. 1995. Tight regulation, modulation, and High-level expression by vectors containing the arabinose PBAD promoter. J. Bacteriol. 177:4121-4130.
Hanafusa, T., K. Saito, A. Tominaga, and M. Enomoto. 1993. Nucleotide sequence and regulated expression of the Salmonella fljA gene encoding a repressor of the phase 1 flagellin gene. Mol. Gen. Genet. 236:260-266.
He, W., and R. Parker. 2000. Functions of Lsm proteins in mRNA degradation and splicing. Curr. Opin. Cell Biol. 12:346-350.
Heichman, K. A., and R. C. Johnson. 1990. The Hin invertasome: protein-mediated joining of distant recombination sites at the enhancer. Science 249:511-517.
Iino, T. 1969. Genetics and chemistry of bacterial flagella. Bacteriol. Rev. 33:454-475.
Ikeda, J. S., C. K. Schmitt, S. C. Darnell, P. R. Watson, J. Bispham, T. S. Wallis, D. L. Weinstein, E. S. Metcalf, P. Adams, C. D. O'Connor, and A. D. O'Brien. 2001. Flagellar phase variation of Salmonella enterica serovar Typhimurium contributes to virulence in the murine typhoid infection model but does not influence Salmonella-induced enteropathogenesis. Infect. Immun. 69:3021-3030.
Inoue, Y. H., K. Kutsukake, T. Iino, and S. Yamaguchi. 1989. Sequence analysis of operator mutants of the phase-1 flagellin-encoding gene, fliC, in Salmonella typhimurium. Gene 85:221-226.
Iost, I., and M. Dreyfus. 1995. The stability of Escherichia coli lacZ mRNA depends upon the simultaneity of it synthesis and translation. EMBO J. 14:3252-3261.
Jain, C., and N. Kleckner. 1993. IS10 mRNA stability and steady-state levels in Escherichia coli: indirect effects of translation and role of rne function. Mol. Microbiol. 9:233-247.
Johnson, R. C., A. C. Glasgow, and M. I. Simon. 1987. Spatial relationship of the Fis binding sites for Hin recombinational enhancer activity. Nature 329:462-465.
Johnson, R. C., and M. I. Simon. 1985. Hin-mediated site specific recombination requires two 26 bp recombination sites and a 60 bp recombination enhancer. Cell 41:781-791.
Joys, T. M. 1985. The covalent structure of the phase-1 flagellar filament protein of Salmonella typhimurium and its comparison with other flagellins. J. Biol. Chem. 260:15758-15761.
Kufel, J., C. Allmang, E. Petfalski, J. Beggs, and D. Tollervey. 2002. Lsm proteins are required for normal processing and stability of ribosomal RNAs. J. Biol. Chem. 278:2147-2156.
Kutsukake, K., and T. Iino. 1980. A trans-acting factor mediates inversion of a specific DNA segment in flagellar phase variation of Salmonella. Nature 284:479-481.
Kutsukake, K., and T. Iino. 1980. Inversions of specific DNA segments in flagellar phase variation of Salmonella and inversion systems of bacteriophages P1 and Mu. Proc. Natl. Acad. Sci. USA 77:7338-7341.
Kutsukake, K., and T. Iino. 1994. Role of the FliA-FlgM regulatory system on the transcriptional control of the flagellar regulon and flagellar formation in Salmonella typhimurium. J. Bacteriol. 176:3598-3605.
Kutsukake, K., S. Iyoda, K. Ohnishi, and T. Iino. 1994. Genetic and molecular analyses of the interaction between the flagellum-specific sigma and anti-sigma factors in Salmonella typhimurium. EMBO J. 13:4568-4576.
Kutsukake, K., H. Nakashima, A. Tominaga, and T. Abo. 2006. Two DNA invertases contribute to flagellar phase variation in Salmonella enterica serovar Typhimurium strain LT2. J. Bacteriol. 188:950-957.
Kutsukake, K., and T. Nambu. 2000. Bacterial flagellum: a paradigm for biogenesis of transenvelope supramolecular structures. Recent Res. Dev. Microbiol. 4:607-615.
Kutsukake, K., Y. Ohya, S. Yamaguchi, and T. Iino. 1988. Operon structure of flagellar genes in Salmonella typhimurium. Mol. Gen. Genet. 214:11-15.
Lindler, L. E., G. V. Plano, V. Burland, G. F. Mayhew, and F. R. Blattner. 1998. Complete DNA sequence and detailed analysis of the Yersinia pestis KIM5 plasmid encoding murine toxin and capsular antigen. Infect. Immun. 66:5731-5742.
Malo, M. S., and R. L. Loughlin. 1988. Promoter-detection vectors for Escherichia coli with multiple useful features. Gene 64:207-215.
Miller, J. 1972. Experiments in molecular genetics, p. 352-355. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Mulligan, M. E., J. Brosius, and W. R. McClure. 1985. Characterization in vitro of the effect of spacer length on the activity of Escherichia coli RNA polymerase at the TAC promoter. J. Biol. Chem. 260:3529-3538.
Parkhill, J., G. Dougan, K. D. James, N. R. Thomson, D. Pickard, J. Wain, C. Churcher, K. L. Mungall, S. D. Bentley, M. T. G. Holden, M. Sebaihia, S. Baker, D. Basham, K. Brooks, T. Chillingworth, P. Connerton, A. Cronin, P. Davis, R. M. Davies, L. Dowd, N. White, J. Farrar, T. Feltwell, N. Hamlin, A. Haque, T. T. Hien, S. Holroyd, K. Jagels, A. Krogh, T. S. Larsen, S. Leather, S. Moule, P. o'Gaora, C. Parry, M. Quail, K. Rutherford, M. Simmonds, J. Skelton, K. Stevens, S. Whitehead, and B. G. Barrell. 2001. Complete genome sequence of a multiple drug resistant Salmonella enterica serovar Typhi CT18. Nature 413:848-852.
Pearce, U. B., and B. A. D. Stocker. 1967. Phase variation of flagellar antigens in Salmonella: abortive transduction studies. J. Gen. Microbiol. 49:335-349.
Philippe, C., F. Eyermann, L. Benard, C. Portier, B. Ehresmann, and C. Ehresmann. 1993. Ribosomal protein S15 from Escherichia coli modulates its own translation by trapping the ribosome on the mRNA initiation loading site. Proc. Natl. Acad. Sci. USA 90:4394-4398.
Rapaport, L. R., and G. A. Mackie. 1994. Influence of translation efficiency on the stability of the mRNA for ribosomal protein S20 in Escherichia coli. J. Bacteriol. 176:992-998.
Rauhut, R., and G. Klug. 1999. mRNA degradation in bacteria. FEMS Microbiol. Rev. 23:353-370.
Regnier, P., and C. M. Arraiano. 2000. Degradation of mRNA in bacteria: emergence of ubiquitous features. Bioessays 22:235-244.
Schmitt, C. K., S. C. Darnell, and A. D. O'Brien. 1996. The attenuated phenotype of a Salmonella typhimurium flgM mutant is related to expression of FliC flagellin. J. Bacteriol. 178:2911-2915.
Shine, J., and L. Dalgarno. 1974. The 3'-terminal sequence of Escherichia coli 16S ribosomal RNA: complementarity to nonsense triplets and ribosome binding sites. Proc. Natl. Acad. Sci. USA 71:1342-1346.
Silverman, M., and M. Simon. 1980. Phase variation: genetic analysis of switching mutants. Cell 19:845-854.
Silverman, M., J. Zieg, and M. Simon. 1979. Flagellar phase variation: isolation of the rh1 gene. J. Bacteriol. 137:517-523.
Steege, D. A. 2000. Emerging features of mRNA decay in bacteria. RNA 6:1079-1090.
Suzuki, H., and T. Iino. 1973. In vitro synthesis of phase-specific flagellin of Salmonella. J. Mol. Biol. 81:57-70.
Van der Woude, M. W., and A. J. Baumler. 2004. Phase and antigenic variation in bacteria. Clin. Microbiol. Rev. 17:581-611.
Winter, R. B., L. Morrisey, P. Gauss, L. Gold, T. Hsu, and J. Karam. 1987. Bacteriophage RegA protein binds to mRNAs and prevents translation initiation. Proc. Natl. Acad. Sci. USA 84:7822-7826.
Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13 mp18 and pUC19 vectors. Gene 33:103-119.
Yarchuk, O., I. Iost, and M. Dreyfus. 1991. The relation between translation and mRNA degradation in the lacZ gene. Biochimie 73:1533-1541.
Zieg, J., M. Silverman, M. Hilmen, and M. Simon. 1977. Recombinational switch for gene expression. Science 196:170-172.
Zieg, J., and M. Simon. 1980. Analysis of the nucleotide sequence of an invertible controlling element. Proc. Natl. Acad. Sci. USA 77:4196-4200.(Shouji Yamamoto1 and Kazu)