ABSTRACT
The regulated expression of virulence genes is critical for successful infection by an intestinal pathogen. Bacteria rely on sensing environmental signals to find preferable niches and reach the infectious state. Orally ingested enterohemorrhagic Escherichia coli (EHEC) travels through the gastrointestinal tract and encounters a variety of environmental factors, some of which act as triggering signals for the induction of virulence genes. Butyrate, one of the main short-chain fatty acids (SCFAs), is such a signal, enhancing the expression of genes for intimate attachment and type III secretion. We further explored the role of SCFAs and found a positive effect of SCFAs on flagellar expression. Although EHEC did not produce flagella when grown in Dulbecco's modified Eagle's medium (DMEM), a tissue culture medium that enhances virulence gene expression, the addition of SCFAs to the medium induced the production of flagella, and the EHEC bacteria became motile. Among SCFAs, butyrate simultaneously activates both virulence and flagellar genes. Flagella did not affect initial adherence, and they were not expressed in adherent bacteria during microcolony formation. SCFAs activated flagellar genes via two regulatory steps. Butyrate activated the flhDC regulatory genes through leucine-responsive regulatory protein (Lrp), which is also a regulator of virulence genes. However, butyrate, acetate, and propionate also activated downstream genes independently of flhDC activation. Consequently, when encountering increased concentrations of SCFAs, which are abundant in acetate, in the intestine, EHEC first activates flagellar production and motility, followed by genes involved in adherence and type III secretion, which leads to efficient adherence in a preferable niche.
Enteropathogenic bacteria encounter a variety of environmental conditions during travel through, as well as colonization of and multiplication in, the human intestine. Intestinal environments are composed of complex factors, including dietary components, host secretions, primary metabolites, and secondary metabolites produced by microflora. The expression of both virulence genes and the associated genes necessary for infection is often affected by changes in these factors, which direct the pathogen to a specific niche in the intestine (8). Short-chain fatty acids (SCFAs), primarily acetate, propionate, and butyrate, are end products of the fermentation of dietary carbohydrates, specifically resistant starches and dietary fiber, by gut bacteria. In the large intestine, SCFAs are present at concentrations ranging between 20 and 140 mM in total, of which acetate is the most abundant (5, 28). SCFAs are believed to play important roles in maintaining colonic health and decreasing the risk of colon cancer and irritable bowel syndrome because butyrate is a major energy source for the colonic epithelium and promotes cell differentiation (34). In addition, butyrate has been shown to protect epithelial cell lines from Campylobacter invasion and translocation (30). However, SCFAs have been shown to affect the expression of virulence genes in Salmonella and enterohemorrhagic Escherichia coli (EHEC). For example, acetate provides a signal for invasion gene expression in Salmonella enterica serovar Typhimurium by the production of acetyl phosphate in the bacterial cytoplasm (18). Additionally, exposure of EHEC to butyrate has been shown to enhance the expression of virulence genes involved in adherence to epithelial cells by sensing the regulator leucine-responsive regulatory protein (Lrp) (23).
The adherence of EHEC to epithelial cells induces attaching/effacing (A/E) lesions, which are characterized by the intimate attachment of bacteria to the surface of the host cell and local destruction of microvilli (17). The formation of A/E lesions requires a type III secretion system and adhesin, which are encoded by the chromosomal locus for enterocyte effacement (LEE). The expression of LEE genes is regulated by a combination of two transcriptional regulators, Pch and Ler (1, 14). Butyrate enhances the adherence capacity of EHEC by elevating the expression of LEE genes, while other major SCFAs, such as acetate and propionate, do not markedly affect expression (23). The sensing of and response to butyrate are mediated by the regulatory protein Lrp, which is conserved among pathogenic and nonpathogenic strains of E. coli (23). SCFAs, including butyrate, can be one of the triggers for EHEC in the intestine to start expressing virulence properties through the induction of virulence regulatory systems consisting of Pch and Ler.
The motility of pathogenic bacteria in the intestine may provide an important advantage for reaching a favorable niche and avoiding detrimental locations. Since the intestinal epithelium is covered with mucus glycocalyx, flagella are necessary for successful infection by some bacteria, such as Vibrio cholerae, Yersinia enterocolitica, and Campylobacter jejuni. In addition, flagella have been shown to be able to serve as adhesive appendages in the initial phase of colonization by Salmonella, E. coli, and Bordetella pertussis (15). However, flagella bear strong antigenic properties, since they are recognized by Toll-like receptor 5 and elicit the expression of the proinflammatory response in epithelial cells (10). Flagella are also involved in the modulation of host cell death by delaying apoptosis in Salmonella-infected epithelial cells (31) and inducing pyroptosis in Salmonella-infected macrophages (7), resulting in a less severe Salmonella infection. Flagella of EHEC and EPEC have been shown to mediate adhesion to epithelial cells in vitro and colonization in the bovine intestine (9, 20) and also to stimulate epithelial cells to produce proinflammatory cytokines (37). Thus, flagellum production must be strictly regulated in the intestine, since it is both advantageous and disadvantageous to pathogenic bacteria.
In this study, we explored the effect of SCFAs on the expression of flagella in EHEC and their role in the colonization of epithelial cells. During growth in tissue culture medium containing SCFAs, both flagellum production and motility were induced in EHEC. Although mutual negative regulation between flagellar genes and LEE genes has been suggested (13), flagellum production and LEE gene expression were simultaneously enhanced by exposure to butyrate. Additionally, we showed that the response of flagellum expression to SCFAs is mediated through two regulatory steps of the flagellar regulatory cascade.
MATERIALS AND METHODS
Bacterial strains, plasmids, and culture conditions.EHEC O157 Sakai (RIMD 0509952) (11) and its derivative strains used in this study are listed in Table 1. Strains harboring deletion mutations or FLAG-tagged genes were constructed from EHEC O157 Sakai according to the method of Datsenko and Wanner (6) or by an epitope-tagging procedure (29). pMS-PflhDC and pMS-PfliC were constructed by inserting a DNA fragment into pMS437C (12) containing the flhDC operon promoter or the fliC promoter, respectively, which was isolated by PCR using the following primers: CTGAGAATTCCCGGGTCGGAAGCGAGAGTA and CTGAGAATTCTCAGACGCTGTGCAAGTAGT for flhDC and AGTCGGATCCCGACGAGTTAGCCGCGCTGAT and CTCTGAATTCACAGACGCTCGATAGAACTC for fliC. Bacteria were grown overnight in LB, diluted 100-fold, and then incubated at 37°C with shaking to an optical density at 600 nm (OD600) of 0.9 to 1.0 in DMEM (Sigma) containing 0.1 M MOPS (morpholinepropanesulfonic acid) (pH 6.7) and sodium chloride (Wako), sodium acetate (Wako), sodium propionate (Sigma), or sodium butyrate (Wako). To examine the effect of leucine, bacteria were grown in DMEM containing 0.1 M MOPS (pH 6.7) and 20 mM l-leucine (Wako).
Bacterial strains and plasmids used in this study
Motility assay.To measure bacterial motility, 5 μl of a logarithmic-phase bacterial culture grown in DMEM with sodium chloride or SCFAs was spotted onto 0.3% agar plates containing DMEM. The motility halos were observed after 10 to 11 h of incubation at 37°C.
Analysis of proteins in whole-cell lysates.Bacteria were collected from culture using centrifugation (at 10,000 × g for 2 min at 4°C), and the cell pellet was dissolved in SDS sample buffer. The concentration of each sample was normalized to the OD600 of the culture, and the samples were analyzed using immunoblotting after SDS-polyacrylamide (12% or 10%) gel electrophoresis (SDS-PAGE) and transfer onto an Immobilon membrane (Millipore). The proteins were detected with antibodies specific for EspB (27), DnaK (Calbiochem), H7 antigen (FliC) (Denka Seiken), or FLAG (Sigma), followed by incubation with a horseradish peroxidase-conjugated secondary antibody and visualization using an ECL detection kit (Amersham Biosciences). The reproducibility of immunoblotting data was confirmed by replication at least twice with independent experiments.
Immunofluorescent staining of flagella.A drop of prepared bacterial culture was placed on a glass coverslip (13-mm diameter), and the bacteria were fixed in 4% paraformaldehyde for 50 min at 37°C. The coverslips were washed twice with phosphate-buffered saline (PBS) and blocked in PBS containing 4% bovine serum albumin (BSA) for 30 min at room temperature. The washed bacteria were then incubated with a rabbit polyclonal anti-H7 (FliC) antibody for 1 h at 37°C. After two washes, the coverslips were incubated with secondary anti-rabbit IgG fluorescein isothiocyanate (FITC)-conjugated antibodies for 1 h. The coverslips were washed twice and mounted onto glass slides using 2 μl Vectashield (Vector Laboratories). The fluorescence was observed, and images were acquired using the BioRevo fluorescence microscope (Keyence).
Adherence assay.The adherence assay was performed as previously described (22), with a slight modification. Bacteria were grown overnight in LB, diluted 100-fold in DMEM with 0.1 M MOPS (pH 6.7) and 20 mM either sodium chloride or the sodium salt of SCFAs, and incubated at 37°C for 3 h with shaking. A confluent monolayer of Caco-2 cells was infected at a multiplicity of infection of 100 ± 15 for 2 h at 37°C. During infection, 20 mM either sodium chloride or the sodium salt of SCFAs was added. The cells were washed with PBS and then incubated in fresh medium, which included either SCFAs or NaCl, for an additional 3 h at 37°C. After another PBS wash, the cells were fixed and stained with Giemsa. The number of adherent bacteria and the number of microcolonies, which are clusters containing at least eight bacteria, were determined in five microscopic fields. The number of adherent bacteria or microcolonies was adjusted for the number of Caco-2 cells in a given field. Averages and standard deviations were calculated from the results of three experiments.
Immunofluorescent staining of flagella of adherent bacteria.For immunofluorescent staining, Caco-2 cells were infected as in the adherence assay. The cells were infected with EHEC expressing DsRed fluorescent protein and then fixed in 4% paraformaldehyde at 2 or 5 h postinfection. For preparation of the 5-h time point sample, nonadherent EHEC cells were removed at 2 h postinfection and the Caco-2 cells were further incubated with fresh medium for an additional 3 h. The fixed cells were blocked in 4% BSA and then incubated with a polyclonal rabbit anti-H7 (FliC) antibody for 1 h at 37°C. After three washes, the coverslips were incubated with secondary anti-rabbit IgG FITC-conjugated antibodies and DAPI (4′,6-diamidino-2-phenylindole) for 1 h. The cells were washed extensively, mounted in SlowFade (Invitrogen), and visualized under a fluorescence microscope (Keyence; BioRevo BZ-9000).
Transcript analysis by DNA microarray.The total RNA from EHEC was isolated using an RNeasy kit (Qiagen) according to the manufacturer's instructions. The RNA was further purified after treatment with RNase-free DNase I (Takara). The GeneChip E. coli Genome 2.0 array of the Affymetrix system was used to compare the transcriptomes of EHEC O157 Sakai grown with either sodium butyrate or leucine. The processing of extracted RNA, cDNA labeling, hybridization, and slide-scanning procedures were performed according to the manufacturer's instructions (Affymetrix). Two independent experiments were performed for each condition. The data corresponding to EHEC O157 Sakai were analyzed and normalized using the total signal intensity. The ratio of transcript levels for each gene was calculated as an average of the ratio for four combinations of two experiments under each condition. The raw and processed data are available in the supplemental material.
RESULTS
Activation of flagellum synthesis by short-chain fatty acids.When grown in DMEM, EHEC does not express flagella and is nonmotile. In contrast, when grown in LB, EHEC is motile and expresses flagella. However, the expression of LEE and related genes, including non-LEE effector genes, is upregulated when EHEC is grown in DMEM compared to that of EHEC grown in LB (13). The intestinal environment contains SCFAs, such as acetate, propionate, and butyrate, and the addition of butyrate has been shown to enhance the expression of LEE genes in EHEC (23). Therefore, we examined the effect of SCFAs on the expression of flagella and motility in EHEC. When EHEC O157:H7 Sakai (O157 Sakai) was grown in DMEM containing 20 mM NaCl as a control, EHEC was nonmotile (Fig. 1A). When grown in DMEM containing 20 mM sodium acetate, sodium propionate, or sodium butyrate, EHEC O157 Sakai became motile (Fig. 1A). The motility of EHEC grown with butyrate and propionate was greater than that of the EHEC grown with acetate, as shown by comparing the sizes of the halos around the colonies. Consistent with the motility results, when O157 Sakai was grown with SCFAs, we observed the production of flagella and elevated levels of FliC, a protein that codes for flagellin, a major component of flagella (Fig. 1B, C, and D). Furthermore, the level of FliC protein and the number of bacteria expressing flagella after growth in acetate were lower than those after growth in propionate or butyrate (Fig. 1B and C). These observations indicated that SCFAs stimulated the expression of flagella and motility in EHEC.
Induction of flagellum production and motility in EHEC by SCFAs. Wild-type EHEC O157 Sakai was grown in DMEM containing NaCl (N), sodium acetate (A), sodium propionate (P), or sodium butyrate (B) at 37°C until late logarithmic phase. (A) Effect of SCFAs on EHEC motility. A drop of bacterial culture was spotted on a soft-agar DMEM plate and incubated at 37°C. (B) Flagellum production by EHEC. Bacteria were fixed on a coverglass, and flagella were visualized using immunofluorescent staining with an anti-H7 antibody. (C) Frequency of EHEC producing flagella. The ratio of flagellated vs. total bacteria was determined by microscopic observation, as in panel B. At least 3 fields containing more than 30 bacteria were counted. The error bars represent standard errors of the mean (SEM). (D) Production of the major flagellar subunit FliC. A whole-cell lysate of EHEC was subjected to immunoblotting for the FliC protein using an anti-H7 antibody. DnaK is shown as a loading reference.
Simultaneous induction of LEE gene expression and flagellum synthesis by butyrate.The expression of LEE genes is enhanced by the presence of butyrate in DMEM, and the adherence capacity is increased approximately 10-fold compared to EHEC grown in DMEM without butyrate (23). However, the induction of LEE gene expression has been shown to repress flagellar gene expression through the action of the LEE-encoded regulator GrlA and constitutive expression of flhD-flhC genes, which encode master regulators of flagellar genes, reducing the adherence capacity of EHEC (13). Therefore, we examined the expression of flagella in ler, grlR, and grlA mutants of EHEC in response to butyrate. GrlR, encoded by the grlR-grlA operon in LEE, is an inhibitor of GrlA. When bacteria were grown in DMEM containing sodium butyrate, the amount of FliC produced by the mutants was similar to that produced by the wild type (Fig. 2A). Although GrlA was in its active state in the grlR mutant, as shown by the expression of EspB, production of the FliC protein in the grlR mutant was enhanced by the presence of butyrate compared to sodium chloride in the medium. In addition, the butyrate-stimulated induction of FliC in the grlA and grlRA mutants was observed, as in the wild type. Next, we examined the involvement of the flhD and flhC genes in the butyrate-stimulated production of FliC and the expression of LEE genes. As shown in Fig. 2B, deletion of the flhD-flhC genes completely abrogated the production of FliC in response to butyrate. However, the butyrate-enhanced production of LEE-encoded EspB was not affected by the presence or absence of flhD-flhC genes. These results indicated that GrlA did not affect the activation of flagellar gene expression, at least in response to butyrate, and that the expression of both LEE and flagellar genes was simultaneously stimulated by butyrate.
Simultaneous enhancement of adherence capacity and flagellum synthesis. (A) Enhancement of the LEE gene-encoded virulence factor EspB and flagellum expression in response to butyrate. Bacteria were grown in DMEM containing NaCl (N) or sodium butyrate (B) until late logarithmic phase. EspB and FliC were detected by immunoblotting using specific antibodies. WT, wild type. (B) Expression of EspB and FliC in the flhDC mutant. EspB and FliC were detected as in panel A. (C) Expression of flagella in adherent EHEC. Caco-2 cells were infected with EHEC constitutively expressing DsRed (Red) after growth with NaCl (N), sodium acetate (A), sodium propionate (P), or sodium butyrate (B). After 2 h, the cells were washed and fixed in paraformaldehyde, and flagella were detected with an anti-H7 antibody and FITC-conjugated anti-rabbit IgG (green). Cell nuclei were detected using DAPI (blue). (D) Adherence of flagellum-expressing EHEC. The adherent EHEC bacteria that did and did not possess flagella were counted after the adherence assay shown in panel C. Adherent bacteria were counted in five microscopic fields of each sample, and the averages ± standard deviations of three independent experiments were determined (*, P > 0.05). (E) Adherence of wild-type EHEC and the flhDC mutant. Caco-2 cells were infected with wild-type EHEC (WT) or the flhDC mutant (ΔflhDC) for 2 h. After the removal of nonadherent bacteria, the cells were incubated for an additional 3 h to allow the bacteria to form microcolonies. The cells were washed, fixed, and stained in Giemsa solution. Microcolonies with at least five bacteria in five microscopic fields were counted, and the average of three independent experiments was determined (*, P > 0.05).
Since it has been previously reported that constitutive expression of flhD and flhC genes greatly reduces the adherence capacity of EHEC (13), we examined EHEC bacteria that adhered to Caco-2 cells for the expression of flagella. To characterize adherent EHEC at an early stage of adherence to Caco-2 cells, infected cells were washed and fixed at 2 h postinfection, when microcolonies with multiple bacteria had not yet formed. The presence of butyrate enhanced the adherence capacity more than 10-fold compared to the control sodium chloride, while acetate or propionate did not (Fig. 2C and D). To determine the expression of flagella by adherent EHEC, flagella were visualized using immunostaining. As shown in Fig. 2C, most of the adherent EHEC bacteria expressed flagella after growth in the medium containing SCFA. Of the adherent EHEC, 50 to 76% produced flagella (Fig. 2D). The results suggest that expression of flagella did not affect adherence capacity. To confirm these results, we compared the adherence efficiency of the flhD-flhC deletion mutant of EHEC, which is deficient in flagellum synthesis, to that of wild-type EHEC after growth in DMEM containing sodium butyrate. The wild type and the flhD-flhC mutant adhered to Caco-2 cells with the same efficiency after growth in DMEM containing sodium butyrate (Fig. 2E). Consequently, the expression of flagella in response to butyrate did not affect the adherence capacity of EHEC.
GrlA-regulated disappearance of flagella from adherent bacteria.Although flagellum production did not affect the initial adherence of EHEC, we further explored flagellum expression after adherence to epithelial cells in the presence of SCFA. To allow the adherent bacteria to form microcolonies, Caco-2 cells infected with wild-type EHEC were further incubated in fresh medium containing SCFA after removal of nonadherent bacteria at 2 h postinfection. At 5 h postinfection, microcolonies were observed, while flagellum-expressing bacteria were rarely observed (Fig. 3A and C). The results suggested that flagellar synthesis was turned off after attachment to epithelial cells even in the medium containing SCFA. Since flagellar gene expression has been shown to be negatively regulated by GrlA, which is encoded by LEE genes (13), we examined the involvement of the grlA gene in the repression of flagellar expression during microcolony formation in the medium containing butyrate. At 5 h postinfection, although flagella were rarely detected in the wild-type microcolonies, most of the grlA mutant EHEC bacteria produced flagella, even in microcolonies (Fig. 3B). Ninety percent of grlA microcolonies comprised bacteria that possessed flagella, while only 5% of the wild-type microcolonies were associated with flagella (Fig. 3C). Introduction of plasmids harboring the grlRA operon into the grlA mutant restored the phenotype (Fig. 3B and C, ΔgrlA/pgrlRA). The sizes of the microcolonies were different in the wild type and the grlA mutant; the grlA mutant microcolonies contained only three to five bacteria, and the wild-type microcolonies contained more than eight bacteria (Fig. 3B). To further compare the capacities for colonization, the change in the number of adherent bacteria after the removal of nonadherent bacteria was determined. Although nonadherent bacteria were removed at 2 h postinfection, the total number of adherent wild-type EHEC bacteria increased, indicating multiplication of adherent bacteria and expanded microcolony size (Fig. 3D). In contrast, the total number of adherent grlA EHEC bacteria was only slightly increased, even at 5 h postinfection (Fig. 3D). These phenotypic changes in the grlA mutant were restored by the introduction of a plasmid harboring the grlRA operon. Since the number of grlA mutants that adhered to Caco-2 cells at 2 h postinfection was the same as that for the wild type, the grlA mutation did not affect the initial adherence capacity. However, the small increase in adherent grlA bacteria 5 h postinfection and the small size of the microcolonies suggest that the grlA gene is required for the efficient formation of microcolonies.
Disappearance of flagella from adherent EHEC. (A) Flagellum production by adherent bacteria that formed a microcolony. Wild-type EHEC O157 Sakai expressing DsRed were grown in DMEM containing NaCl (N), sodium acetate (A), sodium propionate (P), or sodium butyrate (B) at 37°C and then used to infect Caco-2 cells for 5 h, including 3 h after the removal of nonadherent bacteria. Flagella were stained with an anti-H7 antibody and FITC-conjugated anti-rabbit IgG, and cell nuclei were stained with DAPI. (B) Flagellum production by the grlA mutant adhered to Caco-2 cells. Wild-type EHEC (WT), the grlA mutant (ΔgrlA), or the grlA mutant harboring a grlRA plasmid (ΔgrlA/pgrlRA) were allowed to infect Caco-2 cells, and the flagella and cell nuclei were visualized as in panel A. (C) Numbers of microcolonies with flagellum-expressing bacteria. Microcolonies composed of at least 3 bacteria from the experiments shown in panel B were counted (*, P > 0.05 compared to WT). (D) Numbers of adherent bacteria after initial adherence. Caco-2 cells were infected with wild-type EHEC (closed circles), the grlA mutant (closed triangles), or the grlA mutant harboring a grlRA plasmid (ΔgrlA/pgrlRA) (open triangles) for 2 h, and nonadherent bacteria were removed by washing the cells. The cells were further incubated in fresh medium for 3.5 or 5 h. Bacteria were collected in PBS-1% Triton X-100, and the viable bacterial were counted by plating them on LB agar. The numbers of bacteria per well in a 24-well plate are shown. The averages of three independent experiments are indicated, with standard deviations.
Lrp-dependent and -independent regulation of flagellar genes in response to butyrate.The expression of LEE genes in response to butyrate is regulated through sensing and activation of Lrp. Loss of the lrp gene completely abrogates the response of LEE genes to butyrate (23). To examine the role of Lrp in the butyrate-regulated expression of flagella in EHEC, FliC production in the lrp mutant was compared to that of wild-type EHEC during growth in DMEM containing either sodium butyrate or sodium chloride (Fig. 4A). As previously reported, the production of EspB in the lrp mutant remained at a low level even in the presence of butyrate, and the response to butyrate was restored by expression of the lrp gene. In contrast, FliC expression in the lrp mutant was enhanced by growth with butyrate compared with sodium chloride (Fig. 4A). Leucine is a ligand of Lrp (3), and it enhances the expression of LEE genes, similar to butyrate (23). In contrast to the response of LEE genes, such as espB, the expression of fliC was not enhanced by the addition of leucine to the medium (Fig. 4B), supporting a Lrp-independent response of flagellar genes to butyrate.
Role of Lrp in the response of flagellar synthesis to butyrate. (A) Responses in the lrp mutants. Wild-type EHEC (WT), the lrp mutant (Δlrp), and the lrp mutant harboring the Para-lrp fusion gene on a plasmid (Δlrp/Para-lrp) were grown in DMEM containing NaCl or sodium butyrate. FliC, flagellin, and EspB, a LEE gene-encoded virulence factor, were detected using immunoblotting with specific antibodies. To induce a low level of lrp gene expression, 0.05% arabinose was added to the culture of the lrp mutant harboring the Para-lrp gene. (B) Response of the LEE and flagellar genes to leucine or butyrate. Wild-type EHEC bacteria were grown in DMEM containing NaCl (N), sodium butyrate (B), or leucine (Leu) to late logarithmic phase. EspB and FliC were detected in the whole-cell lysate using immunoblotting. (C) Transcriptional levels of flagellar and LEE genes in EHEC grown with butyrate or leucine. Wild type EHEC bacteria were grown in DMEM containing sodium butyrate or leucine, and the total RNA was extracted. The amount of mRNA for each gene was determined using GeneChip (Affymetrix), and the ratios of the mRNA levels of EHEC bacteria grown with butyrate compared to those grown with leucine are shown. (D) Promoter activity of the fliC gene. Wild-type EHEC bacteria harboring the PfliC-lacZ fusion were grown in DMEM containing NaCl (N), leucine (Leu), or sodium butyrate (B). The promoter activity was measured as β-galactosidase activity (*, P < 0.05 compared to NaCl). (E) Promoter activity of the flhDC operon. Wild-type EHEC bacteria harboring the PflhDC-lacZ fusion were grown in DMEM containing NaCl (N), leucine (Leu), or sodium butyrate (B). Promoter activity was measured as β-galactosidase activity (*, P < 0.05 compared to NaCl).
To further explore the response of flagellar genes, we performed a transcriptome analysis to compare the transcript level of each gene in EHEC grown with butyrate to that in EHEC grown with leucine. Total RNA from EHEC grown with butyrate or with leucine was extracted, and the relative amounts of transcript were determined by using DNA microarray analysis. Flagellar genes are organized in a transcriptional regulatory hierarchy that is based on three promoter classes. A class I promoter regulates the flagellar master operon flhDC, which is at the top of this hierarchy and encodes the transcriptional activator of class II promoters. Class III promoters are transcribed by RNA polymerase with σ28, which is encoded by one of the genes with a class II promoter. Since transcript levels of LEE genes were enhanced by both butyrate and leucine through the activation of Lrp, there was no difference between the transcript levels under the two conditions (Fig. 4C). However, the transcript levels of all known genes under class II and class III promoters were higher in EHEC bacteria grown in butyrate than in those grown in leucine. In contrast to genes with class II and class III promoters, the transcript levels of two genes, flhD and flhC, which have a class I promoter, were the same under both growth conditions (Fig. 4C). To confirm the differences in the responses of class I and class III promoters, we measured the activities of the promoters using a promoter-lacZ fusion. The class III promoter activity of the fliC gene was markedly activated by butyrate, but not by leucine (Fig. 4D), whereas the flhDC promoter was enhanced by both butyrate and leucine (Fig. 4E). These results indicated that flagellar genes under class III promoters were responsive only to butyrate, not leucine, while the flhDC operon was activated by leucine and butyrate. This result suggests that flagellar genes are activated by butyrate in the following two distinct steps: (i) at the flhDC promoter through the activation of Lrp and (ii) at class II promoters in which activation is independent of Lrp.
Activation of flagellum expression by SCFAs using a class II regulatory step.Class II promoters are activated by FlhD4FlhC2, which is encoded by the flhDC operon. To explore the role of Lrp in the expression of the flhDC operon, the response of the flhDC operon promoter was examined in the wild type and a lrp mutant of EHEC. The activity of the flhDC promoter in wild-type EHEC was enhanced by growth with butyrate compared to growth with sodium chloride. However, the promoter did not respond to butyrate in the lrp mutant (Fig. 5A). This observation suggests that Lrp was only involved in the butyrate response and not at the basal level of flhDC promoter activity. Next, we examined the production of the FlhC protein in the lrp mutant. When the Lrp-positive strain was grown in DMEM containing butyrate, the amount of FlhC protein was greatly increased compared to that grown with sodium chloride. However, levels of the FlhC protein in the lrp mutant were not increased in response to butyrate (Fig. 4B). The amount of FlhC protein once more paralleled the promoter activity of the flhDC operon, indicating that neither butyrate nor Lrp affected flhDC expression in a posttranscriptional step. In contrast to FlhC production, the amount of FliC was increased by growth with butyrate, even in the lrp mutant (Fig. 5B). Since FlhD and FlhC are activators for class II promoters, we examined the responses of genes under class II promoters in the lrp mutant. The flgN gene, which is the last gene in the flgAMN operon, was FLAG tagged in the Lrp-positive and Lrp-negative strains. The strains were grown in DMEM containing sodium chloride or sodium butyrate, and the expression levels of FlgN were compared. As shown in Fig. 5C, the expression of flgN was enhanced by growth with butyrate compared to growth with sodium chloride in the lrp mutant, as well as the Lrp-positive strain. These results indicated that butyrate enhanced expression of the flagellar genes under the control of class II/III promoters without increasing the amounts of the FlhD and FlhC activators. To confirm that the response of the class II promoter-regulated gene was distinct from the response of the flhDC genes, the effect of SCFAs on the production of the FlgN protein in the lrp mutant was examined. Even in the absence of the lrp gene, the amounts of the FlgN and FliC proteins were increased by all of the SCFAs compared to that of sodium chloride (Fig. 5D). The production of functional flagella was further confirmed using immunofluorescent staining of FliC. As shown in Fig. 5E, we observed the expression of multiple flagella in the lrp mutant grown in DMEM containing butyrate or propionate in over 90% of bacteria. When grown in DMEM containing acetate, the mutant also expressed flagella, although the number of flagellum-containing bacteria was lower than for those grown with butyrate or propionate. These results indicated that SCFAs could enhance the expression of flagella in EHEC without a response of the flhD-flhC genes and that a response using the class II promoter regulatory step was enough for the production of functional flagella.
Lrp-dependent and -independent responses of flagellar synthesis to SCFAs. (A) Response of the flhDC operon promoter to butyrate in the lrp mutant. Wild-type EHEC (WT) and the lrp mutant (Δlrp) harboring PflhDC-lacZ were grown in DMEM containing NaCl or butyrate, and the activity of PflhDC was measured as β-galactosidase activity (*, P < 0.05, and **, P > 0.05 compared to WT with NaCl). (B) Expression of FlhC in the lrp mutant. Expression of the flhC-FLAG gene in the wild-type (WT) or lrp mutant (Δlrp) background was detected using immunoblotting with an anti-FLAG antibody. In addition, FliC production in the same strain was detected using an anti-H7 antibody. (C) Expression of FlgN in the lrp mutant. Expression of flgN-FLAG in the wild-type (WT) or the lrp mutant (Δlrp) background was detected by immunoblotting using an anti-FLAG antibody. FliC production in the same strain was detected using an anti-H7 antibody. (D) Response of the lrp mutant to SCFAs. Expression of flgN-FLAG and fliC in the lrp mutant grown in DMEM containing NaCl (N), sodium acetate (A), sodium propionate (P), or sodium butyrate (B) was detected using immunoblotting. (E) Production of flagella by the lrp mutant in response to SCFAs. The lrp mutant was grown in DMEM containing NaCl or one of the SCFAs, as in panel D. Bacteria were fixed, and flagella were detected using immunofluorescent staining with an anti-H7 antibody and FITC-anti-rabbit IgG. The frequency of bacteria with flagella in the total number of bacteria is shown on the right.
DISCUSSION
Bacterial motility and the expression of flagella are strictly regulated by a cascade of three transcriptional steps and a response to environmental factors. Growth in DMEM enhances virulence gene expression and completely shuts off flagellum expression. However, the addition of short-chain fatty acids to DMEM markedly activated the expression of flagellar genes and motility. The activation of flagellar genes was achieved via a response of two of three regulatory steps for flagellar genes: class I (flhDC operon) and class II (Fig. 6). Activation at a class I promoter requires Lrp and occurs in response to butyrate, but not the other SCFAs, acetate and propionate. However, activation at a class II promoter does not depend on Lrp and responds to acetate and propionate, as well as butyrate. The response of flagellum expression is independent of regulators of LEE genes, such as ler, grlA, and grlR, and flagellum production does not affect the adherence capacity of EHEC. Hence, when EHEC encounters SCFAs in the intestine, the expression of both flagellar and LEE genes is induced, which may differentially contribute to EHEC infection at a mucosal surface. Furthermore, because they have no effect on adherence and disappear during microcolony formation, flagella may contribute to the early stage of infection, such as localizing EHEC to a mucosal surface by activating motility.
Model depicting the SCFA-regulated expression of flagellar and LEE genes. Genes for flagellar biosynthesis are regulated via three steps: class I, class II, and class III. Class II gene expression is activated by SCFAs, such as acetate, propionate, and butyrate, and class I gene expression is activated by butyrate through Lrp. LEE gene expression is also activated by butyrate through Lrp. The LEE gene-encoded regulator GrlA represses flagellar gene expression in adherent bacteria. Eσ28 RNA polymerase, RNA polymerase with flagellar sigma factor.
Although the expression of LEE genes was enhanced by butyrate but not by the other SCFAs, acetate and propionate, flagellar synthesis was enhanced by all three SCFAs. The response of LEE genes to butyrate is dependent upon the Lrp regulator, which senses butyrate, as well as leucine, and activates transcription of a set of genes. Transcriptome analysis comparing transcript levels in EHEC grown with leucine or butyrate clearly indicated that LEE genes and flhD-flhC genes were expressed to comparable levels whereas flagellar genes belonging to class II and III were expressed at higher levels when grown with butyrate than when grown with leucine. As with regulation of LEE genes, the response of the flhDC operon promoter to butyrate was mediated by Lrp, and the response of the flhDC operon promoter was completely abrogated by deletion of the lrp gene. Furthermore, the activity of the flhDC promoter was enhanced by butyrate, but not by acetate or propionate. As shown for LEE genes, the flhDC promoter does not seem to be directly regulated by Lrp, since we could not detect binding of Lrp to the chromosomal region of the flhDC operon (T. Tobe and T. Oshima, unpublished data). The flhDC operon is regulated by many regulatory proteins, such as LrhA, QseB, UvrY, RcsB, CitA, and EnvZ (4, 19, 24). Although we examined flagellar synthesis in mutants of each regulatory gene, none of these genes affected the response. Flagellum expression and motility of E. coli are stimulated by acid (21). Whereas, since the pH of the medium was not altered by the addition of 20 mM sodium butyrate and transcriptome data did not show changes of expression of acid-responding genes (23), it is unlikely that the response to butyrate is due to the acid response. This is further supported by simultaneous enhancement of flagellar gene expression and LEE gene expression by butyrate, since the expression of LEE genes has been reported to be repressed by acid (16, 26). Although the precise mechanism remains unknown, a coordinated response of the flagellar regulatory genes and LEE genes suggests that flagellum production and motility play important roles in EHEC infection.
The flhDC operon encodes master regulators of flagellar synthesis. The necessity for FlhC and FlhD regulators in flagellar synthesis in EHEC was clearly shown by the complete loss of flagellum production in the flhDC mutant, while regulation of flagellum expression by SCFA at least in part seems to be independent of regulation at the flhDC promoter. Although butyrate enhanced the expression of the flhD-flhC genes, the contribution of the flhDC operon to the butyrate response of flagellum expression may be relatively small. This is supported by the observation that flagellum expression was still enhanced by butyrate even in the absence of lrp. In addition to regulation via the class I regulatory step, flagellar gene expression is also enhanced by class II regulatory steps in response to SCFAs, including acetate, propionate, and butyrate. The response via class II gene expression plays a major role in SCFA-regulated expression of flagellar genes compared to regulation through class I promoters. The expression of flagella was enhanced by all three SCFAs even without increasing flhD-flhC gene expression, as observed in the lrp mutant. We have examined the expression of flagella in EHEC with single or combined mutations of chemoreceptors, such as tsr, tar, trg, and aer, but all of the mutants showed a response to SCFAs comparable to that of the wild type. The results indicated that major chemoreceptors were not involved in sensing SCFAs. Since a deletion mutation of the flhDC operon completely abrogated the expression of flagellin, as well as the response to SCFAs, FlhD and FlhC are required for the response via class II promoters. FliT is a posttranscriptional regulator of FlhD4FlhC2 (35), and FliZ is an FlhD4FlhC2-dependent activator of class II promoters (25). We examined the responses in mutants of these genes, but neither the fliT nor the fliZ mutation affected the response to SCFAs (T. Tobe, unpublished data). Since the amount of FlhC protein in the lrp mutant was not increased by butyrate, it is unlikely that SCFAs affect the stability of FlhC and FlhD proteins. One possible mechanism is that SCFAs affect FlhD4FlhC2 activity by changing disulfide bond formation (2) or zinc binding (32) or through an unknown modification. The response to SCFAs via the class II promoter seems to be important for the induction of flagellar gene expression in the intestine. Acetate is highly abundant among SCFAs in the intestine. Therefore, flagellum expression could be enhanced by a relatively low concentration of total SCFAs prior to encountering a higher concentration of SCFAs that contain enough butyrate for the induction of flhDC and LEE gene expression. Furthermore, encountering a higher concentration of SCFAs further enhances flagellum expression, since the induction of flagellum production by propionate and butyrate was stronger than the induction by acetate.
Although most adherent bacteria possess flagella in the initial adherence stage after growth with SCFAs, adherent bacteria that form a microcolony at a later stage do not produce flagella. It is likely that the expression of flagellar genes is turned off after adherence to epithelial cells. This result is consistent with a previous observation by Yona-Nadler et al. (36). These authors observed the disappearance of flagella from adherent EPEC bacteria in a later stage of adherence to HeLa cells. Flagellum production did not stop in the grlA mutant of EHEC; even at 5 h postinfection, adherent bacteria that formed a microcolony still expressed flagella. Overexpression of grlA repressed flagellum expression by EHEC, and GrlA is proposed to be a negative regulator of flagellar genes (13). Therefore, it is possible that GrlA is a key regulator for repressing flagellar genes during microcolony formation in the later stage of adherence. In contrast, butyrate stimulated the expression of flagellin in the grlR, grlAR, and grlA mutants to the same level as in the wild type, indicating that GrlA does not affect the butyrate-activated expression of flagellar genes. During growth in liquid medium, flagellar genes are activated by SCFAs without the inhibitory effect of a LEE product, whose synthesis is also enhanced by butyrate. Since contact with epithelial cells provides other signals to bacteria, as indicated by the activation of type III secretion (33), additional signals after the initial attachment could be necessary for GrlA to repress the flagellar gene.
The intestinal surface is covered with mucus glycocalyx, and bacteria must penetrate through it to reach the surfaces of epithelial cells. The motility conferred on pathogens by flagella is helpful for efficient infection. EHEC might also use flagellum-driven motility to reach the surface of the intestinal mucosa. Since the concentration of SCFAs produced by microflora gradually increases toward the large intestine, we propose that upon encountering increasing amounts of SCFAs at the distal ileum, EHEC bacteria start to produce flagella and become motile. Then, when the butyrate concentration reaches a sufficient level, the elevated expression of LEE and associated genes prepares EHEC for adherence. Once the EHEC bacteria reach the surfaces of epithelial cells and establish initial attachment, flagellum production is turned off because flagella are no longer necessary for further steps in the infection. Instead, flagella at this stage would become disadvantageous to successful infection, since flagellin is a strong agent in inducing a host proinflammatory response. Flagella are lost in the process of colonization, which reduces the chance that the bacteria will be eradicated by the host immune system. Therefore, the regulated expression of flagella is crucial for successful infection by EHEC. SCFAs are important signals for EHEC, as the recognition of proximity to preferable niches triggers flagellum expression and enhances the capacity for intimate attachment.
ACKNOWLEDGMENTS
This work was supported by grant 22590389 from The Ministry of Education, Culture, Sports, Science, and Technology of Japan and a grant from the Yakult Foundation.
FOOTNOTES
- Received 24 August 2010.
- Returned for modification 29 September 2010.
- Accepted 5 December 2010.
- Accepted manuscript posted online 13 December 2010.
† Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.00927-10.
- Copyright © 2011, American Society for Microbiology
