![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Previous Article | Next Article 
Journal of Bacteriology, October 2005, p. 7009-7017, Vol. 187, No. 20
0021-9193/05/$08.00+0 doi:10.1128/JB.187.20.7009-7017.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Dipartimento di Biologia Strutturale e Funzionale, Università Federico II, Naples, Italy,1 The Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Cambridge, United Kingdom2
Received 11 April 2005/ Accepted 23 June 2005
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In enteric bacteria, Lrp appears to play more general role as a global regulator of metabolism (2, 6, 10, 17). The 3,000 dimers of Lrp estimated in E. coli cells were shown by two-dimensional electrophoresis comparison of wild-type and lrp mutant cells to affect the expression of at least 30 genes (5). More recently, a microarray analysis has expanded this view, suggesting that at least 10% of all E. coli genes are under Lrp control (24). In addition to genes involved in amino acid metabolism and transport, most of the genes expressed upon entry into the stationary growth phase belong to this expanded group (24). For some of these genes the interaction with leucine is responsible for the modulation of Lrp action, with cases in which leucine determines or potentiates and others in which it inhibits or reduces the Lrp effect. For a third class of genes, which includes the Lrp structural gene, lrp, leucine has no effect on Lrp action (25).
In pathogenic enterobacteria, Lrp controls some virulence-associated genes. Examples are genes required for conjugal transfer of the virulence plasmid of Salmonella enterica (3), the plasmid-encoded spv gene of S. enterica serovar Typhimurium (15), virulence-associated genes of Proteus mirabilis (8), and various fimbrial genes of E. coli, including the fim, sfa, daa, pap, and fan operons (14, 19).
The fim system is composed of a seven-cistron operon (fimAICDFGH), encoding the structural and regulatory components of the type I pilus, and by two independent genes, fimB and fimE, each encoding a specific recombinase needed to control by phase variation the expression of the fimAICDFGH operon (19). Depending on the orientation of FimS, a 314-bp element containing the promoter of the fimAICDFGH operon, pili are formed (phase ON) or not formed (phase OFF), with the FimB recombinase catalyzing both the ON to OFF and OFF to ON switches and FimE only catalyzing the ON to OFF switch. To allow phase variation, the two recombinases, bound to the two ends of FimS, have to physically interact with each other, which is only possible if the DNA separating them is bent (19). In addition to the fimAICDFGH promoter, the 314-bp element also contains multiple binding sites for the integration host factor (IHF) and for Lrp, two regulatory proteins known to promote DNA bending (19). Mutations in each one of three Lrp binding sites present within FimS, whose effect is to prevent Lrp binding, reduce both FimB- and FimE-dependent switching (7).
We studied the lrp gene and the regulatory role of its product on the expression of fim genes in Citrobacter rodentium, a mouse pathogen that belongs to a family of human and animal pathogens, such as the clinically significant enteropathogenic and enterohemorrhagic E. coli. C. rodentium causes transmissible colonic hyperplasia in mice by means of attaching and effacing lesions through which it colonizes the host gastrointestinal tract (12). Since enteropathogenic and enterohemorrhagic E. coli and other human enteropathogens are not able to colonize mice, C. rodentium has been extensively used as a model of human gastrointestinal pathogens in in vivo experiments and has proven useful in revealing phenotypes for proteins, including fimbrial proteins, not revealed by in vitro infection models (26).
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(supE44 
lacU169 [
80lacZM15] hsdR17 recA1) (20) was used for all cloning experiments, while the E. coli strains CV975 (ilvIH::lacZ) and CV1008 (ilvIH::lacZ lrp-35::Tn10) (18) were used for the complementation experiments. Bacteria were grown at 37°C in rich or minimal medium supplemented with thiamine (5 µg/ml), glucose (0.4%), and, when indicated, leucine (100 µg/ml). Antibiotics were 100 µg/ml ampicillin and 35 µg/ml chloramphenicol. For ß-galactosidase or ß-glucuronidase activity assays, overnight cultures were diluted to an optical density at 600 nm of 0.1 in the appropriate medium, grown in shaking conditions at 37°C up to the log phase (optical density at 600 nm of approximately 0.8), collected by centrifugation, and stored at –80°C until the time of assay.
DNA manipulations.
Plasmid and chromosomal DNA preparations, restriction digestions, ligation, bacterial transformation, and agarose gel electrophoresis were performed as previously described (20). DNA sequencing reactions were performed by using the T7 sequencing kit (USB corporation) with [-35S]thio-dATP (>1,000 Ci mmol–1; MP Biomedicals). Southern blot experiments were performed according to standard procedures (20) and using as probes DNA fragments labeled by means of Ready-ToGo DNA labeling beads kit (without dCTP) (Amersham) and with [
-32P]dCTP (MP Biomedicals).
Plasmid pAC12 was obtained by cloning a 1,228-bp amplification product resulting from a PCR performed with C. rodentium chromosomal DNA as the template and synthetic oligonucleotides A and C as the primers (Table 1) into the commercial vector pGEMT-Easy (Promega). Plasmid pAC12 was used to transform competent cells of the E. coli strain CV1008. Ampicillin-resistant clones were screened on LB agar containing 5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside (X-Gal) and one clone, AC13, checked by nucleotide sequencing reaction, was selected for further study.
|
An entire copy of the lrp gene, released by NotI digestion from plasmid pAC12, was cloned into pAC47 previously linearized by SalI digestion. The noncohesive ends of the DNA fragments were filled by treatment with Klenow fragment (Biolabs), according to the manufacturer's instructions. The resulting plasmid, pAC61, was then used to transform the wild-type C. rodentium, generating strain AC62.
Construction of lrp and fimE null mutations. Null C. rodentium mutations in the fimE and lrp genes were constructed by using the Datsenko and Wanner (4) method. In brief, the chloramphenicol resistance cassette (cat) of plasmid pKD3 was PCR amplified by using oligonucleotides X and Y for lrp or Z and W for fimE (Table 1). These oligonucleotides were designed with 18 and 17 bases, respectively, complementary to cat sequences at their 3' ends, next to 40 bases complementary to regions adjacent to the lrp or fimE gene (–368 to –329 for X, +432 to +469 for Y, +23 to +66 for Z, and + 457 to +497 for W, considering the first translated nucleotide as +1). The 1,091-bp (lrp) and 1,099-bp (fimE) amplification products were used to transform C. rodentium strain EM1, generated by transformation of the wild-type strain ATCC 51459 with the low-copy-number plasmid pKD46, expressing the lambda red recombinase and carrying an ampicillin resistance gene, by electroporation. Chloramphenicol-resistant clones were then cured of the pKD46 plasmid by repeated growth cycles at 37°C in the absence of ampicillin. The presence of the cat cassette within the lrp or fimE gene was verified by PCR with oligonucleotides c1 and O (lrp) or Q (fimE) (Table 1) and Southern blot. One positive clone for each transformation, EM2 (lrp) and EM3 (fimE), was selected for further studies.
ß-Galactosidase and ß-glucuronidase assays. ß-Galactosidase assays were performed as previously described (18). For each strain analyzed, units of activity (nanomoles of o-nitrophenyl galactoside hydrolyzed per minute) were calculated from A420/Ve t, where Ve is the volume of permeabilized cells in ml and t is the time in minutes. Specific activity equals units per A590. ß-Glucuronidase assays were performed as previously described (28). For each sample a graph of A405 (y axis) versus time in minutes (x axis) was designed; the slope S of the graph in A405 units per minute was estimated and units of activity (nanomoles of p-nitrophenyl glucuronide hydrolyzed per minute) were calculated from S/Ve x 0.02, where Ve is the volume of permeabilized cells in ml and 0.02 represents the A405 given relative to 1 nmol of product produced. Specific activity equals units per A590. Values reported here were the average of at least three independent experiments. Statistical significance was determined by Student's t test and the significance level was set at P < 0.05.
Isolation of total RNA and RT-PCR. Total RNA was extracted using the QIAGEN mini kit (QIAGEN, Milan, Italy) using the manufacturer's instructions. Total RNAs were dissolved in 50 µl of RNase-free water containing recombinant porcine RNasin (1U/µg of total RNA; Amersham Pharmacia Biotech) and stored at –80°C until used for the reverse transcription (RT)-PCR analysis. The final concentration and quality of the RNA samples were estimated using either a spectrophotometer or by agarose gel electrophoresis with ethidium bromide staining.
Prior to RT-PCR total RNAs were treated with RNase-free DNase (1 U/µg of total RNA; Turbo DNA-free, Ambion) for 30 min at 37°C and the reaction was stopped with DNase inactivation reagent; 2 µM of oligonucleotide E was used to prime 4 µg of total RNAs in a final volume of 15.5 µl. The template and primer were incubated at 70°C for 10 min and then added to a reverse transcriptase reaction mixture containing 1x reaction buffer, 3 mM MgCl2, 0.5 mM each of the deoxynucleoside triphosphates, 1 U/µl recombinant porcine RNasin (1 U/µg of total RNA; Amersham Pharmacia Biotech), and 1 µl Improm II reverse transcriptase (Promega). The mixture was incubated at 25°C for 5 min, at 42°C for 60 min, and at 70° for 10 min to inactivate the enzyme.
After reverse transcription the cDNA obtained was primed with oligonucleotide pairs as indicated in Fig. 1A and 3A. For each oligonucleotide pair, a positive (chromosomal DNA as template) and two negative (no RT and no cDNA) control reactions were performed. PCR conditions were 5 min at 94°C, followed by 30 cycles of 95°C for 30 s, 52°C for 40 s, and 72°C for 1 min, concluding with extension at 72°C for 5 min. The products were separated on 1.0% agarose gels, stained with ethidium bromide, and visualized by UV transillumination.
|
|
Ct method was used to calculate the relative amount of specific RNA present in a sample, from which the fold induction of transcription of the gene was estimated by comparison to values relative to the wild-type strain grown in minimal medium at early stationary phase. Data were expressed as the mean ± standard error of the mean. Statistical significance was determined by Student's unpaired t test and the significance levels are reported in the text.
Primer pair efficiency was tested by looking at how Ct (the difference between the two Ct values of two PCRs for the same template amount) varies with template dilution, as suggested by the manufacturer's instruction guide (PE Applied Biosystems). According to the instruction guide, the efficiency of amplicons was considered 100% when plots of log template amount versus Ct originated parallel lines with slope values between –3.1 and –3.7.
Yeast agglutination experiment.
The capacity of bacteria to express an -D-mannose binding phenotype was assayed by their ability to agglutinate Saccharomyces cerevisiae cells on glass slides, as previously reported (23); 10 µl of liquid bacterial cultures at an optical density of 595 nm of 3 and 10 µl of 3% (wt/vol) yeast cells, dissolved in phosphate-buffered saline, were gently mixed and the resulting suspension was incubated at room temperature for 2 min.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A 1,228-bp PCR product, obtained using synthetic oligonucleotides A and C (Table 1 and Fig. 1A) to prime for the amplification of the lrp gene from the C. rodentium chromosome, was cloned in a pGEM-T-easy (Promega) vector. The recombinant plasmid obtained, pAC12, was used to transform the E. coli strain CV1008 (ilvIH::lacZ lrp::Tn10) (18). The resulting recombinant strain, AC13, contained the lrp-controlled operon ilvIH (18) translationally fused to the lacZ reporter gene on the chromosome and the C. rodentium lrp gene on a plasmid as the only entire copy of this gene, since a Tn10 chromosomal insertion disrupted lrp of E. coli (18). The lrp gene of C. rodentium was able to complement the lrp null mutation of strain CV1008, restoring ilvIH expression to levels similar to those due to endogenous lrp (strain CV975) (Fig. 1B). In addition, like the E. coli protein (5), C. rodentium Lrp responded to the presence of leucine with a strong reduction of the transcriptional activation of the ilvIH operon (Fig. 1B).
To characterize the lrp promoter region we performed a series of RT-PCR experiments. Oligonucleotide E (Table 1, Fig. 1A) was used to prime total RNA with reverse transcriptase. The cDNA obtained was then primed with oligonucleotide pairs as indicated in Fig. 1A. Primer-pairs L-E, I-E, and H-E but not G-E, F-E, and A-E originated an amplification product of the expected size (data not shown), indicating that a DNA region up to 250 bp upstream of the translational start site is transcribed and that the transcriptional start point is in the region between oligonucleotides G and H. A similar situation has been observed for E. coli, where the lrp transcriptional start site has been mapped 267 bp upstream of the translational start site (25).
Lrp negatively controls the expression of its own structural gene.
To study the expression of the lrp gene of C. rodentium we constructed an lrp null mutant by replacing the lrp gene with a chloramphenicol resistance cassette (cat) on the C. rodentium chromosome. The low-copy-number plasmid pKD46, encoding the 
Red recombinase (4), was used to transform wild-type C. rodentium. The resulting strain, EM1, was then transformed with a 1,091-bp PCR product, containing the cat cassette flanked by 40 bp of homologous to DNA adjacent to lrp. Chloramphenicol-resistant clones were the result of a double-crossover event. Several clones were checked by PCR and Southern blot, and one clone, EM2, was selected for further analysis.
An lrp::gusA translational fusion was then obtained as follows: a 734-bp DNA fragment containing the lrp promoter region and six N-terminal codons of the lrp open reading frame was amplified from C. rodentium chromosomal DNA by using oligonucleotides E and N as primers (Table 1). The PCR product was then fused in frame to the gusA gene of E. coli carried by plasmid pGusA, yielding plasmid pAC47. Plasmid pAC47 was then introduced into the wild-type C. rodentium strain ATCC 51459 and into its isogenic lrp null mutant EM2, yielding strains AC49 and AC52, respectively. AC49 showed a ß-glucuronidase activity significantly higher than that observed with ATCC 51459 cells transformed with the vector plasmid pGusA (data not shown), excluding a potential interference of endogenous enzymes in our assays. As shown in Fig. 2, lrp-directed ß-glucuronidase activity was slightly higher (P < 0.05) in the lrp null mutant (AC52, white bar) than in the wild type (AC49, white bar), suggesting that Lrp autogenously repressed the expression of its own structural gene. Addition of 100 µg/ml of exogenous leucine to the growth medium did not significantly (P > 0.05) affect ß-glucuronidase activity in either strain (Fig. 2, AC49 and AC52, gray bars), suggesting that the Lrp control on lrp was leucine independent.
|
Lrp positively controls the expression of fim genes. In order to verify whether the C. rodentium chromosome contains genes encoding the type 1 pilus, we compared the nucleotide sequence of the fim chromosomal region of Escherichia coli K-12 (accession number U14003) with sequences present in the data bank of the Sanger Institute. This search identified a C. rodentium genome region highly homologous to the fimAICDFGH operon and to the two adjacent fimB and fimE genes of E. coli. In addition, a 314-bp region bounded by left and right inverted repeats (IRL and IRR) and identified as fimS in Fig. 3, located between the fimAICDFGH and fimE genes and containing the fimAICDFGH promoter, was also found. By homology with the same region of E. coli, we assumed that in C. rodentium fimS is involved in a phase variation mechanism controlling fimAICDFGHI expression.
To verify whether the C. rodentium fim genes are transcriptionally active, we performed a series of RT-PCR experiments priming the total RNA of C. rodentium with the synthetic oligonucleotides listed in Table 1 and graphically indicated in Fig. 3. These experiments, summarized in Table 2, indicated that fimB and fimE are transcribed but not cotranscribed (see Table 2, oligonucleotide pairs B1-B2, E1-E2, and B1-E2) and that fimAICDFGH is not cotranscribed with fimE (see Table 2, oligonucleotide pair E1-C4). Indication that fimAICDFGH is transcribed and forms a single transcriptional unit came from DNA sequence data revealing the presence of the open reading frames indicated in Fig. 3 and from RT-PCR data (see Table 2, oligonucleotide pairs A5-C4, Cs-D1, Ds-Da, D6-G5, and G4-H3). These results allowed us to conclude that the fim genes are transcribed and organized into three units of 6,596 bp (fimAICDFGH), 606 bp (fimB), and 597 bp (fimE).
|
|
|
To confirm the transcriptional results in Table 3 and verify the role of Lrp in the formation of the type I pilus, we also performed a yeast agglutination assay (23). This assay is indicative of the presence of type I pili on bacterial cells since their tip is formed by FimH, an adhesin that mediates attachment to mannose-containing receptors, abundantly present on the surface of yeast cells. As shown in Fig. 5, lrp mutant cells of C. rodentium (Fig. 5A) have a reduced ability to agglutinate Saccharomyces cerevisiae cells compared with wild-type cells (Fig. 5B), as evidenced by the smaller number of bacteria-yeast aggregates. This result is in agreement with the transcriptional data summarized in Table 3 and indicate that Lrp has a positive role in the formation of the type I pilus in C. rodentium.
|
As reported in Table 4, the number of FimS molecules in phase ON was 37-fold (P < 0.05) higher in wild-type cells than in lrp null mutant cells, indicating that Lrp controls the phase variation by favoring the ON orientation. In the same conditions, the number of fimAICDFGH mRNA molecules was 16-fold (P < 0.005) higher in the wild type than in the lrp mutant (Table 4). The different increase in the number of potentially active promoters (FimS in phase ON) and of synthesized mRNA molecules observed in the presence of Lrp suggests that fimAICDFGH expression is not constitutive but somehow negatively regulated.
|
Yeast agglutination assays performed with strain EM3 showed that in cells not expressing fimE the agglutination activity is stronger and, as a consequence, the number of type I pili on the cell surface is much higher than on wild-type cells (Fig. 5), suggesting that in C. rodentium, as in E. coli and other enterobacteria, fimE encodes the recombinase that specifically catalyzes the ON-to-OFF switch.
A real-time PCR analysis of fimAICDFGH expression was performed on chromosomal DNA and on cDNA synthesized from total RNA extracted from cells of strain EM3 grown in minimal medium. As reported in Table 4, in EM3 the number of FimS molecules in phase ON was 86-fold (P < 0.05) higher than in wild-type cells, while the number of fimAICDFGH mRNA molecules was 50-fold (P < 0.05) higher than in the wild type. Consistent with data derived from the lrp mutant, the results obtained with strain EM3 (fimE) indicate a different increase in the number of DNA molecules in phase ON and of mRNA molecules synthesized, suggesting that fimAICDFGH expression is not constitutive and is instead regulated by a dual control at the levels of both phase variation and transcription.
Although statistically significant, the difference between the number of fimS DNA and fimAICDFGH mRNA molecules shown in Table 4 could be affected by a different efficiency of the oligonucleotides used to prime the real time PCR amplification. To check the efficiency of our primer pairs, wild-type cells were used to extract chromosomal DNA and total RNA and cDNA were synthesized from the latter. DNA and cDNA were then serially diluted in a range including the dilutions applied in the experiments of Table 4 and used as templates in real-time PCR experiments. Reactions were primed with oligonucleotide pairs ONs and ONa (amplifying fimS in the ON orientation; Table 1), fimA-rt1 and fimA-rt2 (amplifying a coding region of fimA; Table 1), and sig70s and sig70a (amplifying a coding region of the normalizing gene rpoD; Table 1). Parallel straight lines with similar slopes (–3.70 for ONs-ONa; –3.64 for fimA-rt1-fimA-rt2; and –3.35 and –3.55 for sig70s-sig70a amplifying rpoD chromosomal DNA and cDNA, respectively) were obtained with the template dilutions used for all oligonucleotide pairs (data not shown). The slope values were in the range considered optimal (Materials and Methods) and indicated that Ct did not vary in different PCR conditions and therefore all oligonucleotide pairs tested were working with the same efficiency.
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Type I pili, encoded by the fim gene system, are the most common virulence factors of uropathogenic bacteria and allow bacterial adhesion to oligosaccharides containing mannose (21). The adhesive properties of the type I pilus are determined by FimH, a lectin-like protein associated with the fimbrial tip and encoded by the last gene of the fimAICDFGH operon. The type I pilus has recently been identified as a virulence factor of the invasive pathogen Citrobacter freundii (9). However, although they belong to the same genus, C. freundii and C. rodentium have fim genes with different chromosomal organization and low DNA homology. The fim genes of C. freundii are instead similar to those of Salmonella enterica serovar Typhimurium (9) and, as shown here, the C. rodentium ones are instead similar to those of E. coli.
Yeast agglutination assays indicated that lrp null mutant cells of C. rodentium are strongly impaired in the ability to agglutinate yeast cells, showing a positive effect of Lrp on pilus formation. A real-time PCR approach allowed us to observe that the expression of the fimAICDFGH operon, but not that of fimB and fimE, was strongly enhanced by the presence of Lrp, suggesting that the positive role of Lrp on pilus formation is exerted through the control of fimAICDFGH expression. When the Lrp ligand, leucine, is present the positive Lrp effect on fimAICDFGH is enhanced further.
The positive Lrp control of fimAICDFGH expression might either influence the phase variation mechanism by favoring the ON orientation or enhance transcription of the operon. We observed that Lrp strongly favors the ON orientation of the fimS switch and that this effect is not mediated by an action of Lrp on the expression of the fimB or fimE gene, encoding the specific FimS recombinases. Therefore, we propose that the Lrp action on the phase variation mechanism is most probably due to the DNA-bending activity of Lrp (25), which allows the physical interaction of the recombinases bound to the two ends of FimS, as previously proposed for E. coli (19). However, fimAICDFGH transcription does not appear to be constitutive but rather negatively controlled. Whether this transcriptional level of control is also dependent on Lrp remains to be clarified.
Our real time PCR experiments, performed in growth conditions in which fimAICDFGH is highly expressed (Fig. 4), suggested that to an increased number of FimS molecules in the ON orientation did not correspond to an equal increase in the number of fimAICDFGH-specific mRNA molecules. The statistically significant difference between the increase in DNA molecules ready to be transcribed and of fimAICDFGH-specific mRNA molecules induced us to conclude that not all molecules in the ON orientation were transcribed, as expected from an unregulated, constitutive promoter and, as a consequence, that the fimAICDFGH promoter is somehow negatively regulated.
Consistent with this conclusion is the analysis of a mutant that does not produce the recombinase that specifically catalyzes the ON-to-OFF switch. In this mutant the number of FimS molecules in the ON orientation is 86-fold higher than in the isogenic wild type, while the number of fimAICDFGH-specific molecules is only 50-fold higher in the mutant. Taken together, these statistically significant differences suggest that, in addition to the Lrp-mediated control of the fim switch, fimAICDFGH expression is also negatively controlled by a transcriptional mechanism.
Since C. rodentium causes a transmissible colonic hyperplasia in mice similar to that induced in humans by enteropathogenic and enterohemorrhagic E. coli strains (12), it has recently been selected as an in vivo model system to study human infections (26). The characterization in this model organism of Lrp, a transcriptional regulator that affects a large variety of genes including those coding for potential virulence factors, opens a new field of investigation in C. rodentium pathogenesis. The construction of an lrp null mutant will allow the identification of Lrp-controlled virulence factors and a better understanding of the role in the infection process of Lrp and the virulence factors that it controls.
|
ACKNOWLEDGMENTS |
|---|
This work was supported by European Union grant QLK5-CT-2001-01729 to E.R.
|
FOOTNOTES |
|---|
¶ These two authors contributed equally to the work.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Appl. Environ. Microbiol. | Infect. Immun. | Eukaryot. Cell |
|---|---|---|
| Mol. Cell. Biol. | J. Virol. | Microbiol. Mol. Biol. Rev. |
| ALL ASM JOURNALS |
