Abstract
Quorum-sensing (QS) signalling pathways are important regulatory networks for controlling the expression of genes promoting adherence of enterohaemorrhagic Escherichia coli (EHEC) O157 : H7 to epithelial cells. A recent study has shown that EHEC O157 : H7 encodes a luxR homologue, called sdiA, which upon overexpression reduces the expression of genes encoding flagellar and locus of enterocyte effacement (LEE) proteins, thus negatively impacting on the motility and intimate adherence phenotypes, respectively. Here, we show that the deletion of sdiA from EHEC O157 : H7 strain 86-24, and from a hha (a negative regulator of ler) mutant of this strain, enhanced bacterial adherence to HEp-2 epithelial cells of the sdiA mutant strains relative to the strains containing a wild-type copy of sdiA. Quantitative reverse transcription PCR showed that the expression of LEE-encoded genes ler, espA and eae in strains with the sdiA deletions was not significantly different from that of the strains wild-type for sdiA. Similarly, no additional increases in the expression of LEE genes were observed in a sdiA hha double mutant strain relative to that observed in the hha deletion mutant. While the expression of fliC, which encodes flagellin, was enhanced in the sdiA mutant strain, the expression of fliC was reduced by several fold in the hha mutant strain, irrespective of the presence or absence of sdiA, indicating that the genes sdiA and hha exert opposing effects on the expression of fliC. The strains with deletions in sdiA or hha showed enhanced expression of csgA, encoding curlin of the curli fimbriae, with the expression of csgA highest in the sdiA hha double mutant, suggesting an additive effect of these two gene deletions on the expression of csgA. No significant differences were observed in the expression of the genes lpfA and fimA of the operons encoding long polar and type 1 fimbriae in the sdiA mutant strain. These data indicate that SdiA has no significant effect on the expression of LEE genes, but that it appears to act as a strong repressor of genes encoding flagella and curli fimbriae, and the alleviation of the SdiA-mediated repression of these genes in an EHEC O157 : H7 sdiA mutant strain contributes to enhanced bacterial motility and increased adherence to HEp-2 epithelial cells.
- EHEC, enterohaemorrhagic Escherichia coli
- LEE, locus of enterocyte effacement
- QRT-PCR, quantitative real-time polymerase chain reaction
- QS, quorum sensing
Edited by: J. G. Shaw
INTRODUCTION
Enterohaemorrhagic Escherichia coli (EHEC) O157 : H7 is responsible for outbreaks of bloody diarrhoea, haemorrhagic colitis and haemolytic-uraemic syndrome (HUS) in humans (Nataro & Kaper, 1998). In addition to Shiga toxins that act on vascular endothelial cells to produce HUS (Karmali, 1989), EHEC O157 : H7 produces characteristic attaching and effacing (A/E) lesions on infected host epithelial cells (Francis et al., 1986). The genes required for the formation of A/E lesions are encoded by the locus of enterocyte effacement (LEE), which contains 41 ORFs organized into operons LEE1–LEE5 (Elliott et al., 2000). The expression of the LEE operons is regulated by the LEE-encoded positive regulator Ler. Several positive and negative regulators have been shown to regulate ler expression. In previous studies, we have reported that Hha represses ler transcription, and that the deletion of hha increases expression of ler and ler-regulated LEE1–LEE5 operons, thereby enhancing the adherence of the mutant strain to HEp-2 cells (Sharma & Zuerner, 2004; Sharma et al., 2005). In addition to LEE-mediated intimate adherence to epithelial cells, EHEC O157 : H7 presumably also uses surface appendages, such as flagella and fimbriae, in the initial stages of the adherence process. For example, recent studies have shown that flagella of EHEC O157 : H7 bind bovine intestinal mucos to promote intestinal colonization, and fliC deletion mutants exhibit reduced adherence to bovine intestinal tissues (Erdem et al., 2007). Although EHEC O157 : H7 does not express type 1 fimbriae because of a 16 bp deletion in the fim switch (Roe et al., 2001), it contains several other fimbrial operons, including the operons encoding curli and long polar fimbriae. Recent studies have implicated long polar fimbriae in adherence of EHEC O157 : H7 to epithelial cells, and intestinal colonization of ruminants (Torres et al., 2007). Similarly, the expression of curli fimbriae has been shown to enhance adherence of EHEC O157 : H7 and O157 : NM strains to HEp-2 and Caco-2 cells, respectively (Rosser et al., 2008; Uhlich et al., 2002).
Quorum-sensing (QS) systems have emerged as important regulatory pathways in controlling the expression of multiple cellular functions in E. coli O157 : H7 such as flagellation, motility, and intimate adherence to epithelial cells (Reading et al., 2007; Sperandio et al., 2002, 2003). In some Gram-negative bacterial species, QS requires LuxR, a transcriptional factor that recognizes and binds to specific QS signalling molecules called autoinducers (Fuqua et al., 1994). These autoinducers, which belong to a family of secreted molecules called acylhomoserine lactones, are synthesized via the pathways requiring LuxI-encoded autoinducer synthase (Fuqua et al., 1994).
Like E. coli K-12 and Salmonella, EHEC O157 : H7 contains a LuxR homologue called SdiA, but lacks its cognate autoinducer-synthase-encoding gene (Henikoff et al., 1990). Therefore, only a few studies have focused on the understanding of the biological significance of SdiA in controlling specific cellular functions in E. coli strains. For example, Kanamaru et al. (2000a) reported that overproduction of SdiA from a high-copy-number plasmid reduces motility and the expression of some of the LEE-encoded genes of EHEC O157 : H7. Other studies have implicated sdiA as an important component of the regulatory cascade that governs the formation of biofilms by E. coli K-12 by regulating the expression of motility and chemotaxis genes in response to indole, which is a by-product of tryptophan metabolism (Lee et al., 2007). However, types of regulatory controls exerted by the chromosomal copy of sdiA on the expression of genes, such as those encoded by the flagellar, fimbrial and LEE operons, which are essential for motility and adherence of EHEC O157 : H7 to epithelial cells, are not completely understood.
In this study, we describe the effects of sdiA deletion on the adherence of EHEC O157 : H7, with and without the gene hha, to HEp-2 cells, and correlate the degree of adherence of these mutants to the relative expression of LEE-encoded genes and the genes involved in the biosynthesis of flagella and fimbriae.
METHODS
Media and growth conditions.
Bacterial strains were grown in either Luria–Bertani broth (LB; 10 g tryptone, 5 g yeast extract, 10 g NaCl per litre of medium) or Dulbecco's modified Eagle's medium (DMEM), at 37 °C, with or without shaking (200 r.p.m.). Media were supplemented with appropriate antibiotics.
Construction of sdiA and sdiA hha mutants.
EHEC O157 : H7 strain 86-24 Δstx2 Δlac (Sharma & Zuerner, 2004) was used for construction of the mutants by using the lambda-Red-mediated recombineering method (Murphy & Campellone, 2003). Briefly, a 1.539 kb DNA fragment containing a 1.1 kb fragment encoding kanamycin resistance, flanked at its 5′ and 3′ ends with 254 bp and 175 bp sequences representing nucleotide regions immediately upstream and downstream of sdiA, respectively, was isolated from a recombinant plasmid using primers sdiA-F and sdiA-R (Table 1⇓). This 1.539 kb fragment was purified from an agarose gel using a gel extraction kit, according to the manufacturer's instructions (Qiagen). The gel-purified fragment was electroporated into electrocompetent cells of EHEC O157 : H7 strain 86-24, and its hha mutant (Sharma & Zuerner, 2004) carrying plasmid pKM208; kanamycin-resistant colonies were isolated, and screened by PCR to confirm that sdiA had been deleted from the chromosome of the electroporated strain.
Primers used in this study
RNA isolation.
An overnight culture of a bacterial strain grown in LB broth plus 100 μg streptomycin ml–1 was diluted 1 : 100 in DMEM, and incubated at 37 °C with shaking (200 r.p.m.). At an OD600 of approximately 1.0, the culture was mixed with 2 vols RNAprotect (Qiagen) by vortexing for 5 s, then incubated for 5 min at room temperature, and centrifuged at 5000 g for 10 min at 4 °C. Total RNA was isolated from cell pellets by using an RNeasy Mini kit, according to the manufacturer's instructions (Qiagen). RNA was electrophoresed in 1 % standard agarose gels using Tris/borate/EDTA buffer. Following electrophoresis, the gel was stained with ethidium bromide to make sure that the RNA appeared as a smear containing two prominent rRNA bands. Any contaminating DNA present in the purified RNA preparation was removed by DNase treatment using the TURBO RNase-free DNase kit, according to the manufacturer's instructions (Ambion). Purified RNA was checked for its concentration using a spectrophotometer (NanoDrop Technologies), and it was used as a template in a standard PCR (PE Biosystems) containing primers for rpoA amplification to verify that it was free of DNA contamination.
Quantification of gene expression by QRT-PCR.
Quantitative real-time RT-PCR (QRT-PCR) was performed in 15 μl reaction volumes containing 2 μl (25 ng RNA) DNA-free RNA, 7.5 μl 2× QRT-PCR master mix, 0.6 μl RT/RNase block mixture (Stratagene), 0.225 μl 0.066 M reference dye ROX, 1.5 μl 1× SYBR Green (Molecular Probes), appropriate volumes of forward and reverse primers (Table 1⇑), and RNase-free water. QRT-PCR was carried out in a single-step reaction format using Mx3005P (Stratagene) programmed to generate cDNA in one cycle of 30 min at 50 °C, followed by a single 10 min cycle at 95 °C, and 40 cycles of cDNA amplification (95 °C for 30 s, 55 °C for 60 s, and 72 °C for 30 s). A comparative quantification option provided in the Mx3005P system set-up procedure was used for QRT-PCR data acquisition and analysis. The sample containing RNA from the parent strain was used as a calibrator to which all other RNA samples were compared, with respect to the expression of the target gene. To account for any variations in the amounts of RNA across samples, QRT-PCR amplification of each sample was normalized to samples containing primers for rpoA (Table 1⇑). The passive reference dye ROX allowed compensation for non-PCR related variations in the fluorescence data acquisition. QRT-PCR data are presented as fold change in the expression level of the target gene relative to the expression level for the same gene in the calibrator, where the expression level of the calibrator is set to 1.0. Final data represent expression levels as means (±sem) of the data collected from three independent experiments.
Determination of adherence to HEp-2 cells.
Bacterial adherence to HEp-2 cells was performed by a previously described procedure (Sharma et al., 2005). Briefly, fresh cultures (0.5 ml) of HEp-2 cells (5.0×104) were added to a well of a chamber slide (Nalge Nunc International), and incubated at 37 °C overnight in an atmosphere of 5 % CO2 in air. The wells were washed three times in Dulbecco's PBS, followed by addition of 0.5 ml RPMI 1640 containing 1 % fetal bovine serum (Invitrogen), and 50 μl of an overnight bacterial culture prepared as follows: a few colonies of a bacterial strain grown overnight on LB agar were inoculated into LB broth, incubated at 37 °C on a shaker (200 r.p.m.) for 8 h, diluted 1 : 500 in fresh LB broth, and incubated overnight at 37 °C, without shaking.
The chamber slides were incubated at 37 °C for 3 h in an atmosphere of 5 % CO2 in air. The chamber portion of chamber slide was removed, and the slide was washed four times in PBS, fixed for 1 min in 95 % ethanol, stained in toluidine blue for 15 s, washed in distilled H2O, dipped in 95 % ethanol to fix the stain, washed briefly in water, and air-dried. A coverslip was mounted on the air-dried slide using VectaMount, and the slide was visualized under a microscope at ×100 magnification. Ten visual fields containing confluent HEp-2 cells were randomly selected for enumerating the number of bacterial cells adhered per HEp-2 cell. The adherence data are represented as the mean number of bacteria per HEp-2 cell in a given visual field obtained from three independent assays.
Determination of cellular morphology and motility.
Bacterial wet mounts prepared from cultures grown in DMEM were examined, using a phase-contrast microscope at ×400 magnification, to determine any gross changes in cellular morphology, such as the appearance of filamentous cells as opposed to small rod-shaped bacterial cells typical of E. coli strains. For determination of bacterial motility, overnight cultures were centrifuged at 5000 g for 5 min, washed in PBS, and spotted (2 μl per spot) on motility agar plates, which were prepared by adding 0.32 % agarose to DMEM medium. The diameters of motility haloes were measured after incubation for 18 h at 37 °C.
Construction of sdiA complementation strains.
The gene encoding sdiA was amplified by PCR from EHEC O157 : H7 strain 86-24 Δstx2 Δlac using primers sdiA-F915 and sdiA-R864, which contain SalI restriction sites at their 5′ ends (Table 1⇑). These primers were used to amplify a 1.1 kb fragment containing the promoter and the ORF encoding SdiA. This 1.1 kb fragment was cloned into the SalI site of a previously described temperature-sensitive plasmid (pSM80) used for deleting the lac operon in strain 86-24 (Sharma & Zuerner, 2004). The plasmid pSM80 containing the 1.1 kb sdiA gene fragment was introduced into sdiA and sdiA hha mutant strains by electroporation, and clones containing the sdiA gene in the lac region were selected by using a previously described allelic replacement procedure (Sharma & Zuerner, 2004). The insertion of the 1.1 kb fragment in the lac regions of sdiA and sdiA hha mutant strains was confirmed by PCR using primers lac-F824 and lac-R881 (Table 1⇑).
Congo red binding and biofilm assays.
The abilities of bacterial cells to bind the dye Congo red, and produce biofilms on plastic surfaces, were determined by the minor modifications of previously described procedures (Hammar et al., 1996; Uhlich et al., 2006a). For Congo red binding assays, Congo red and Coomassie brilliant blue dyes were used at 50 μg ml−1 and 6.25 μg ml−1, respectively, in YESCA agar medium. Bacterial strains grown overnight in LB broth were streaked onto this medium, and, following incubation at 28 °C for 48 h, the plates were checked for growth, and photographed under identical settings for exposure, background light and contrast to eliminate any digital biasing of the resulting images. For biofilm assays, 96-well polystyrene plates were inoculated with 200 μl of 1 : 100-fold diluted, overnight bacterial cultures grown in YESCA broth at 37 °C with shaking (175 r.p.m.). After incubation at 28 °C for 48 h, the plates were inverted on paper towels to drain the wells of the culture. The plates were heat-fixed, stained with 0.1 % crystal violet solution, and scanned at 590 nm to assess the amount of biofilm production. Biofilm production is given as the mean (±sem) of three independent experiments.
Statistics.
One-way ANOVA with Tukey–Kramer multiple comparisons test was performed using GraphPad InStat version 3.00 for Windows 95 (GraphPad Software).
RESULTS
Deletion of sdiA enhances cellular adherence
As shown in Fig. 1⇓, deletion of sdiA from EHEC O157 : H7 strain 86-24 enhanced adherence to HEp-2 cells twofold. Furthermore, deletion of sdiA from the hha mutant strain resulted in an additional twofold increase in bacterial adherence of the sdiA hha double mutant relative to the hyperadherent hha mutant strain, indicating that SdiA inhibits the adherence of EHEC O157 : H7 to epithelial cells.
Effect of sdiA on adherence of EHEC O157 : H7. Adherence of sdiA deletion mutants of EHEC O157 : H7 with or without hha was monitored by exposing HEp-2 cells to overnight bacterial cultures for 3 h at 37 °C. After washing and fixing, the HEp-2 cells were stained with toluidine blue and adherent bacteria were counted at ×100 magnification. Assays were performed in triplicate in which adherent bacteria were enumerated from 10 HEp-2 cells for each replicate. Error bars indicate sem. The P-values, indicated as ** (<0.01) and *** (<0.001), are reported in relation to the adherence level exhibited by the parent strain.
Effect of sdiA on expression of LEE-encoded genes
In order to understand the genetic basis for the increased adherence of the sdiA deletion mutant strains to HEp-2 cells, we first analysed total RNA by QRT-PCR to determine expression levels for several LEE-encoded genes, as their increased expression has been shown to facilitate increased adherence to epithelial cells (Sharma & Zuerner, 2004; Sharma et al., 2005). As shown in Table 2⇓, sdiA deletion resulted in very low-level increases, which were determined to be statistically insignificant, in the expression of LEE-encoded ler (1.3-fold), espA (1.2-fold) and eae (1.6-fold), relative to the parent strain. However, in comparison, high levels of increase in expression observed for ler (15-fold), espA (22-fold), and eae (166-fold) in the hha mutant strain were independent of the presence or absence of the sdiA deletion, indicating that SdiA has either no or a very weak effect on the expression of LEE-encoded genes, in comparison with the stronger repression exerted by Hha on the same genes.
Effect of sdiA on the expression of adherence and motility genes
SdiA reduces flagellar and fimbrial gene expression
Several reports have described that the expression of flagellar and various fimbrial operons affects adherence of EHEC O157 : H7 to epithelial cells (Kim & Kim, 2004; Rosser et al., 2008; Saldana et al., 2009; Uhlich et al., 2002). Therefore, we used QRT-PCR to examine the effects of sdiA deletion on the expression of fliC, lpfA, csgA and fimA, genes representing operons encoding flagella, long polar fimbriae, curli fimbriae and type 1 fimbriae, respectively. Since type 1 fimbriae are not expressed in EHEC O157 : H7 strains because of a 16 bp deletion in the fimA regulatory region (Roe et al., 2001), the expression of fimA in sdiA, hha and sdiA hha mutant strains was expected to be identical to that of the parent strain. As shown in Table 2⇑, fliC expression in the sdiA deletion mutant was significantly increased by 2.7-fold relative to the parent strain 86-24. On the other hand, fliC expression in the hha mutant strain was reduced by 5-fold compared with that in the parent strain. It was also apparent that the deletion of sdiA did not alter the expression of fliC in the hha deletion strain, suggesting that Hha is necessary for the expression of fliC, and the positive effect of sdiA deletion on fliC expression is evident in a hha+ strain only. However, the expression of lpfA, the structural gene for the long polar fimbriae, in the sdiA (1.124-fold), hha (0.81-fold) and sdiA hha (0.934-fold) mutant strains was not significantly different from that of the parent (1.00-fold) strain. However, the expression of csgA, encoding the structural protein curlin of the curli fimbriae, was increased by approximately three- to fourfold in sdiA and hha deletion mutants relative to the parent strain. In the sdiA hha double mutant strain, the expression of csgA was increased by ninefold relative to the parent strain, and by two- to threefold relative to the csgA expression observed in either sdiA or hha deletion mutants. The increased expression of csgA in sdiA and hha deletion mutants indicates that both SdiA and Hha exert a negative effect on the expression csgA. As expected, the expression of fimA, encoding the major structural protein FimA of type 1 fimbriae, was not significantly different between the parent strain (1.00-fold) and sdiA (0.98-fold), hha (0.68-fold) and sdiA hha (0.86-fold) deletion mutant strains.
sdiA deletion affects bacterial motility, but has no effect on cellular morphology
Kanamaru et al. (2000a) have shown that the expression of SdiA from a high-copy-number plasmid not only results in the appearance of elongated cells due to abnormal cell division, but also reduces the motility of EHEC O157 : H7. However, examination of the wet mounts of sdiA, hha and sdiA hha mutants under a phase-contrast microscope showed the presence of small, rod-shaped cells that are considered typical of E. coli strains (data not shown). On the other hand, visual examination of soft-agar motility plates revealed that motility of the sdiA mutant was increased by approximately 30 % relative to the parent strain, indicating that SdiA exerts a negative effect on the motility of EHEC O157 : H7 (Fig. 2⇓). It was also apparent that the positive effect of sdiA deletion on motility was dependent on the presence of hha, since the motility of the sdiA hha double mutant was reduced to the levels that were observed for the hha mutant strain (approximately 35 % lower than the parent strain).
Effect of sdiA deletion on the motility of EHEC O157 : H7. Each strain was spot inoculated onto soft DMEM agar. The diameters of the motility haloes (means±sem of three independent experiments) were as follows: A (parent), 9.38±1.21 mm; B (hha mutant), 6.33±0.73 mm; C (sdiA mutant), 12.28±1.25 mm; D (sdiA hha mutant), 6.04±0.82 mm. The difference in the motility diameter for each mutant strain was determined to be significant (P<0.05) relative to the parent strain.
sdiA deletion increases Congo red binding by bacterial cells
It has been reported that the ability of E. coli strains to bind Congo red is affected by the levels of curli fibres expressed at the bacterial cell surface (Kim & Kim, 2004; Rosser et al., 2008; Uhlich et al., 2006b; Vidal et al., 1998). Since the deletion of sdiA enhanced the expression of csgA (the gene encoding major structural component of curli fimbriae) in a hha-independent manner, we wanted to determine if an increased expression of csgA in strains deleted of sdiA would enable them to bind increased amounts of Congo red. In addition, we inserted a single copy of the wild-type sdiA gene in the lac region of the chromosome of sdiA and sdiA hha mutant strains to determine whether SdiA would complement the abilities of these strains to bind Congo red in amounts comparable with those bound by the sdiA+ or sdiA+ hha strains, respectively. Visual examination, comparing growth of sdiA (Fig. 3a⇓, panel C) and sdiA hha (Fig. 3b⇓, panel C) mutants on Congo red medium, revealed the presence of dark-pink colonies with dark-red centres compared with the light-pink colonies with red centres (fish-eye appearance) produced by the sdiA+ (Fig. 3a⇓, panel A) and sdiA+ hha (Fig. 3b⇓, panel A) strains. Complementation of the sdiA (Fig. 3a⇓, panel B) and sdiA hha (Fig. 3b⇓, panel B) mutant strains with SdiA decreased Congo red binding to a level similar to that observed for colonies of the sdiA+ (Fig. 3a⇓, panel A) and sdiA+ hha (Fig. 3b⇓, panel A) strains.
Phenotypic production of curli on medium containing Congo red. Overnight bacterial cultures were streaked on YESCA agar containing Congo red. After incubation for 48 h at 28 °C, the plates were photographed to capture coloured phenotypes of colonies that grew on these plates. (a) Coloured phenotypes of the colonies produced by the parent strain (panel A), sdiA deletion mutant complemented with a wild-type copy of sdiA (panel B), and sdiA deletion mutant (panel C). (b) Coloured phenotypes expressed by the parent strain (panel A), sdiA complemented sdiA hha deletion mutant (panel B), and the sdiA hha deletion mutant (panel C).
Reduction of biofilm formation by sdiA in EHEC O157 : H7
The formation of biofilms in E. coli is influenced by motility and the expression of fimbrial structures, which are required for reversible and irreversible phases of adherence during the genesis of biofilms, respectively (Kim & Kim, 2004; Pratt & Kolter, 1998; Ryu & Beuchat, 2005; Wood et al., 2006). Since we determined that the sdiA deletion caused increased expression of fliC and csgA, which correlated positively with enhanced motility and increased Congo red binding, we hypothesized that the sdiA deletion mutant might show increased biofilm production. As shown in Fig. 4⇓, the amount of biofilm produced by the sdiA deletion mutant was only 1.56-fold higher than that produced by the sdiA+ parent strain. On the other hand, the amount of biofilm produced by the hha mutant was 3.56-fold higher than the sdiA+ parent and 2.3-fold higher than the sdiA deletion mutant strains. The deletion of sdiA in the hha deletion mutant caused only a small increase (1.22-fold) in biofilm formation over the hha mutant strain, suggesting that SdiA exerts effects of lower magnitude than Hha on the production of biofilms in EHEC O157 : H7. The fact that the hha deletion mutant displayed reduced expression of fliC (Table 2⇑), reduced motility (Fig. 2⇑), and increased csgA expression (Table 2⇑) suggests that the repression of motility and increased expression of curli fimbriae are necessary for increased biofilm formation by EHEC O157 : H7. Thus, the increased motility caused by deletion of sdiA might be responsible for preventing enhanced biofilm formation by the sdiA mutant strain, despite its ability to express higher levels of csgA and increased Congo red binding compared with the sdiA+ parent strain.
Evaluation of biofilm formation in polystyrene plates. Overnight bacterial cultures were diluted 1 : 100 in YESCA broth and pipetted into the wells of a 96-well polystyrene plate. Following 48 h of incubation at 28 °C, the amount of biofilm produced was determined by crystal violet staining as described in Methods. Biofilm production is shown as the mean±sem of three independent experiments.
DISCUSSION
In this study, we have demonstrated that the deletion of a single chromosomal copy of sdiA significantly increases the adherence of EHEC O157 : H7 and its isogenic hha mutant strain to HEp-2 cells, suggesting that SdiA exerts a negative effect on the adherence phenotype. These findings are in agreement with the results reported in a previous study showing that the expression of sdiA from a high-copy-number plasmid causes a dramatic reduction in the ability of EHEC O157 : H7 to adhere to epithelial cells (Kanamaru et al., 2000a). In addition, Kanamaru et al. (2000a) demonstrated that an increased expression of sdiA from a high-copy-number plasmid results in abnormal cell division. However, deletion of sdiA from the EHEC O157 : H7 chromosome, as we have described in this study, had no affect on cellular morphology.
Unlike most of the known bacterial QS systems that activate virulence gene expression in EHEC O157 : H7 (Sperandio et al., 1999, 2002), SdiA-mediated QS had been shown to repress transcription of the flagellar gene fliC, and the virulence genes espD and eae encoded by the LEE4 and LEE5 operons, respectively (Kanamaru et al., 2000b). However, these negative effects of SdiA were demonstrated upon increased expression of sdiA from a high-copy-number plasmid. By using QRT-PCR, we demonstrated very low levels of increase in the expression of LEE-encoded genes ler, espA and eae in the sdiA deletion strain compared with the high levels of increase observed for these genes in the hha mutant strain. These low levels of increase in the expression of LEE-encoded genes observed in our study were statistically insignificant compared with the effect of SdiA reported for the repression of LEE-encoded genes in the study by Kanamaru et al. (2000a). Thus, the lack of a significant effect on the expression of LEE-encoded genes in the sdiA mutant in our study may be due to the reduced intracellular amounts of SdiA produced from a single chromosomal copy of sdiA. For example, it has been postulated that the majority of SdiA expressed from the chromosomal copy of sdiA is a monomer, and that expression of SdiA from a high-copy-number plasmid increases the levels of SdiA dimers, which are required for repressing transcriptional activities of the target gene promoters (Henikoff et al., 1990; Michael et al., 2001). The proposal of a repressor function of SdiA is supported by the presence of a helix–turn–helix DNA-binding motif, and, like most repressors, SdiA dimerizes or oligomerizes to affect the repression of their target gene promoters (Burz & Ackers, 1994; Burz et al., 1994; Harrison & Aggarwal, 1990; Sharma et al., 1998). Thus, the highly impaired adherence of EHEC O157 : H7 reported by Kanamaru et al., (2000a) could very well have resulted from an increased intracellular availability of SdiA, so that concentrations of dimers reached levels that were high enough to inhibit transcriptional activities initiated from the affected promoters of the LEE4 and LEE5 operons.
Several studies have shown that the ability of EHEC O157 : H7 to produce flagella and certain types of fimbriae enhances its adherence to epithelial cells in animal models and in tissue culture assays (Erdem et al., 2007; Torres et al., 2007; Uhlich et al., 2002). In the current study, we found that the strain deleted of sdiA showed increased transcriptional levels of fliC and enhanced motility compared with the parent strain. The overall smaller sizes of motility haloes produced in our studies could be attributed to the use of a minimal medium containing a higher percentage of agar (0.32 %). For example, strain 86-24 has been used by other investigators to examine the impact of QS signalling molecules (epinephrine and norepinephrine) on its motility (Sperandio et al., 2003). In those studies, the sizes of motility haloes produced by strain 86-24 in the absence of epinephrine or norepinephrine were not substantially different from the ones observed in our studies, despite the fact that the motility haloes in those studies were determined on media containing a lower percentage of agar (0.25 %). Although the negative effects of SdiA on fliC expression and/or on motility were qualitatively similar to those reported for a strain of EHEC O157 : H7 producing high levels of SdiA (Kanamaru et al., 2000a; Wei et al., 2001), the negative effect of SdiA on fliC expression was only apparent in the strain wild-type for hha, because strains deleted of hha showed reduced expression of fliC, irrespective of the presence or absence of sdiA. These findings suggest that Hha plays a positive role, via an unknown mechanism, in the expression of fliC. Deletion of hha resulted in reduced fliC expression (and reduced bacterial motility), but enhanced expression of LEE-encoded genes, and bacterial adherence to HEp-2 cells; these findings are similar to those reported in a previous study (Sharma et al., 2005), and suggest that the enhanced in vitro adherence of hha or sdiA hha mutants to HEp-2 cells may not require optimal expression of flagella. Since biosynthesis of both flagella and the type III secretion system requires investment of large amounts of cellular resources, it is reasonable to assume that an increase in the expression of the type III secretion system in the hha mutant might compensate for the reduced production of flagella during bacterial adherence to HEp-2 cells.
Unlike the negative effect of SdiA on fliC gene expression that was apparent only in a hha-positive strain, the negative effect of SdiA on the expression of csgA, the gene encoding curlin of the curli fimbriae, was of a similar magnitude in strains with or without hha. However, the sdiA hha double mutant had the highest level of csgA expression, indicating an additive effect on csgA expression due to deletion of sdiA and hha. In light of a previous report demonstrating that a highly curliated strain of EHEC O157 : H7 displays a highly aggregated adherence pattern on HEp-2 cells, as opposed to the more localized adherence pattern observed for the lesser curliated strain (Kim & Kim, 2004), the increased expression of csgA might be one of the major factors, considering that the expression of LEE-encoded genes was not significantly affected by sdiA, responsible for the enhanced adherence of sdiA and hha mutant strains to HEp-2 cells. The enhanced expression of csgA contributing to an increased adherence of sdiA mutant strains to HEp-2 cells was corroborated by the ability of these strains to bind increased amounts of Congo red, and to produce slightly higher levels of biofilm than the strains with a wild-type sdiA gene. Since increases in Congo red binding and biofilm production are indicative of enhanced curli production (Kim & Kim, 2004; Uhlich et al., 2006a), a reduction in the levels of Congo red binding and biofilm production by the sdiA mutant strains via complementation with a wild-type copy of sdiA provided additional evidence for the negative role of SdiA in csgA transcription and curli production. These results are corroborated by a study that utilized a whole-transcriptomic approach to show that SdiA represses curli formation in E. coli K-12 (Lee et al., 2009).
Interestingly, sdiA deletion resulted in increased fliC expression and enhanced motility, but only a small increase in the amount of biofilm formation compared with larger increases in the levels of biofilm production in hha and sdiA hha deletion mutants. These results indicate that while the increased production of curli promotes adherence to HEp-2 cells, increased motility, on the other hand, reduces the ability of the sdiA mutant strains to produce biofilms. These findings suggest that in EHEC O157 : H7 strain 86-24, unlike other E. coli strains (Barrios et al., 2006; Pratt & Kolter, 1998; Wood et al., 2006), motility inhibits biofilm formation on abiotic surfaces, whereas increased production of curli allows increased adherence to both abiotic and biotic surfaces in motility-compromised hha mutant strains.
It is also apparent from the data presented in this report that SdiA serves as a strong repressor for the transcriptional regulation of fliC and csgA, and that it appears to have no effect on the expression of lpfA and fimA. Moreover, the additive effects of SdiA and Hha on the expression of csgA appears to suggest that these two repressor proteins either bind to independent sites in the csgA promoter region, or the binding of the one repressor protein precedes the binding of the other repressor to the same promoter-binding site to cause effective repression of the csgA promoter.
In conclusion, our study provides direct genetic evidence that expression of a single chromosomal copy of sdiA represses multiple phenotypes, such as flagellation, fimbriation, motility and adherence in EHEC O157 : H7. Based on the data provided in this report, SdiA exerts negative effects of varied magnitudes on different sets of genes. For example, the expression of LEE-encoded genes was not significantly affected compared with the expression of fliC and csgA, which were strongly repressed. Although in vitro gel-shift assays have shown that purified SdiA binds to its target promoters with the same affinity (Kanamaru et al., 2000a), the lack of a consensus SdiA-binding sequence in the SdiA-regulated promoters might determine the relative affinities of these promoters for SdiA in vivo. Since EHEC O157 : H7, like E. coli K-12 and Salmonella spp., encodes sdiA, but lacks the ability to synthesize acylhomoserine lactones, which are the natural inducers of SdiA, understanding the nature of signals influencing expression or repressor activity of SdiA is important for clarifying the role of SdiA in virulence gene expression in these bacterial species. One potential signal could be indole, the metabolic by-product of tryptophan metabolism, which regulates biofilm formation by influencing the activity of SdiA (Lee et al., 2007). Thus, correlating the expression of indole with SdiA effects on virulence gene expression might provide greater insight into the mechanism of SdiA-mediated regulation of gene expression in EHEC O157 : H7.
Acknowledgments
We thank Robert Morgan for technical assistance, and Nancy Cornick and Louisa Tabatabai for critical review of the manuscript.
Disclaimer: Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture.