Citrobacter rodentium is a murine pathogenic bacterium that adheres to intestinal epithelial cells, resulting in loss of microvilli and pedestal formation, and alters multiple cellular processes, including actin dynamics. Translocated intimin receptor (Tir), one of its virulence factors, functions as receptor for intimin, a bacterial adhesin, thereby mediating bacterial adhesion to epithelial cells. Although robust immune responses are induced to eliminate pathogenic bacteria in the host, they are suppressed against harmless commensal bacteria. The mechanism(s) underlying such a differentiation remains unclear. This study sought to determine the roles of intimate adhesion in the induction of specific immune responses upon C. rodentium infection. To this end, microbiota-depleted mice were infected with the Tir-F strain expressing full-length Tir or mutant strains expressing the C-terminal truncated Tir that is defective in intimin binding and host cell actin polymerization. There were no differences in the colonization kinetics and Abs responses against C. rodentium LPS among the strains, whereas Abs against the virulence factors were only produced on Tir-F infection. Although there were no differences in the virulence factors mRNA expression levels, colonic hyperplasia, and bacterial translocation to the systemic organs irrespective of the strain, adhesion to colonic epithelial cells was reduced in the mutant strain–infected mice. Furthermore, transcriptomic analysis indicated that robust inflammatory and immune responses were only induced in the Tir-F–infected group and were suppressed in the mutant-infected groups. Taken together, these findings suggest that Tir-mediated intimate adhesion induces inflammatory and immune responses, resulting in the induction of virulence factor–specific Abs.

Enterohemorrhagic Escherichia coli (EHEC) and enteropathogenic E. coli (EPEC) colonize the intestine and cause diarrhea in humans (1, 2). To colonize the intestine in a stable manner, these bacteria first attach to the epithelia by fimbria, followed by formation of attaching/effacing (A/E) lesions, characterized by intimate bacterial attachment to the intestinal epithelial cells, induction of cellular actin polymerization under the attached bacteria, effacement of the microvilli, and formation of pedestal-like structures (3). To form the A/E lesions, they use the type 3 secretion system (T3SS) to inject various effectors. The effectors are encoded both in the locus of enterocyte effacement (LEE) and in the non-LEE locus of the bacterial genomes and affect signaling pathways for multiple biological targets in host cells (4).

Translocated intimin receptor (Tir), one of the effector proteins of T3SS, is integrated into the plasma membrane of the host epithelial cells, exposing an extracellular central region that functions as a receptor for intimin, a bacterial adhesin (5). Binding of intimin to Tir induces clustering of Tir, which triggers actin polymerization under the attached bacteria via several mechanisms that require the C-terminal intracellular region of Tir (68).

A previous study reported that induction of the EHEC-specific mucosal and systemic IgA responses require the intimate adhesion of the bacteria to host epithelial cells via the Tir–intimin interaction (9). Citrobacter rodentium is a natural mouse A/E pathogen that colonizes the colonic mucosa by a mechanism similar to those of EHEC and EPEC (4). It was reported that infection of germ-free mice with a C. rodentium strain lacking the eae gene for intimin induced mucosal IgA production at a level lower than that of the wild-type (WT) strain (10). These results suggest that adhesion of the A/E pathogens to the intestinal epithelial cells via the Tir–intimin interaction is important for the induction of mucosal IgA responses. However, the underlying mechanism by which the Tir–intimin interaction induces mucosal IgA production, its effect on systemic Ab responses, and the specificity of the induced Abs have been largely unknown.

In this study, to address these questions, we examined Ab responses in mice infected with C. rodentium mutants expressing an authentic form of Tir or truncated forms of Tir that are defective both in binding with intimin and in induction of actin polymerization in epithelial cells.

C. rodentium (American Type Culture Collection 51459) and Tir mutant strains were grown in Luria–Bertani (LB) medium or DMEM supplemented with kanamycin (50 µg/ml) or chloramphenicol (10 µg/ml) when necessary. To determine the growth rate, an overnight culture of C. rodentium was added to fresh LB medium at a 1:100 dilution and incubated at 37°C with shaking. OD600 was measured hourly for 8 h.

We used the RedET-based homologous recombination system to introduce a deletion into the endogenous chromosomal tir gene of luciferase-producing C. rodentium as previously described (11). Briefly, the C. rodentium genome sequence from 3,161,217 to 3,158,515 (GenBank account no. NC_013716) was amplified by PCR and cloned into pSI029 (12), resulting in plasmid pSI029–Tir-F. Next, a kanamycin-resistance gene cassette from pFN21K (Promega) was cloned downstream of the tir gene of pSI029–Tir-F, resulting in pSI029–Tir-F–KanR. The regions of the tir gene encoding aa 412–547 and aa 172–547 were deleted from pS1029–Tir-F–KanR by inverse PCR, resulting in pSI029–Tir-M–KanR and pSI029–Tir-N–KanR, respectively. DNA fragments encoding the authentic, Δ412–547 and Δ172–547 tir genes were PCR amplified from pSI029–Tir-F–KanR, pSI029–Tir-M–KanR, and pSI029–Tir-N–KanR, respectively, and electroporated into luciferase-producing C. rodentium (11) harboring plasmid pRedET (Gene Bridge). Clones resistant to kanamycin and chloramphenicol were obtained, and their mutated tir genes were confirmed by sequencing.

Total proteins were extracted by sonication from C. rodentium, which was cultured in DMEM for 6 h to induce expression of virulence factors, and insoluble fractions were removed by centrifugation (10,000 × g, 10 min, 4°C). Protein concentrations of the soluble fractions were determined using a BCA Protein Assay Kit. Western blotting was performed as previously described using anti-Tir sera (13).

Total RNA specimens were extracted from C. rodentium that were cultured for 6 h to mimic the T3SS-positive (DMEM) or -negative (LB) conditions using a TRIzol Max Bacterial RNA Extraction kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. To analyze virulence factor gene expression in vivo, 5-mm pieces of the cecum were dissected from mice as described previously (14). Total RNA specimens were extracted from the tissues as described above. Total RNA (200 ng) was converted to cDNA using the PrimeScript RT reagent with gDNA Eraser (TAKARA), and quantitative RT-PCR (qRT-PCR) was performed using the GoTaq qPCR Master Mix kit (Promega) according to the manufacturer’s instructions. Luciferase gene (nluc) was used as a calibrator for the analysis. qRT-PCR reactions were performed using Mx3000P (Agilent Technologies). Primers were listed in Supplemental Table I.

The murine rectal carcinoma cell line CMT-93 (American Type Culture Collection no. CCL-223) was grown and maintained in DMEM containing 10% FBS. The C. rodentium mutants cultured overnight in DMEM were added to CMT-93 cells grown in eight-well chamber slides (Nunc) at a multiplicity of infection of 100. The cells were washed at 3 h postinfection with sterile PBS, then incubated for an additional 3 h in DMEM. Unbound bacteria were then removed by washing the cells with PBS five times. The cells were fixed with 4% paraformaldehyde in PBS and permeabilized with 0.1% Triton X-100. C. rodentium was stained with rabbit anti-O152 serum (Denka Seiken) and Alexa Fluor 488 (Alexa488)–conjugated anti-rabbit IgG (Invitrogen). Tir was stained using anti-Tir sera and Alexa647-conjugated anti-mouse IgG (Invitrogen). ActiStain555 (Cytoskeleton) was used to stain F-actin, and DNA was counterstained with Hoechst 33342. Images of the stained cells were acquired using an LSM900 confocal microscopy (Carl Zeiss). Data were analyzed by acquiring 25–30 random digital images that contain around five cells per image from two independent experiments. Z-stack images were collected at 0.13-µm optical sections and reconstituted using ZEN 3.1 software (Carl Zeiss).

Specific pathogen–free female 6-wk-old ICR mice were purchased from Japan SLC. All animals were housed in individually high-efficiency particulate air–filtered cages with sterile bedding and free access to sterilized food and water. All animal experiments were approved by the institutional animal care and use committee.

Mice were treated with an antibiotic mixture containing 1 mg/ml ampicillin, 0.5 mg/ml vancomycin, 1 mg/ml neomycin, and 0.6 mg/ml lincomycin in drinking water for 5 d before inoculation with C. rodentium and then treated with 0.5 mg/ml kanamycin alone with drinking water throughout the experimental course postinfection. C. rodentium was cultured with 15 ml of LB broth for 16 h at 37°C with shaking, and the bacteria were harvested and resuspended with 1.5 ml of sterile PBS to prepare the inoculum as reported (15). Mice were inoculated by oral administration with 0.2 ml of the C. rodentium suspension or PBS for the uninfected group for three consecutive days. The number of live C. rodentium within the inoculum was counted retrospectively by plating dilutions of the remaining portion of the materials used for inoculation (1 × 1010 to 2 × 1010 CFU/mouse). Fresh fecal pellets were collected from individual mice at varying time points after inoculation and frozen at −80°C until use for the determination of the number of C. rodentium in feces by luciferase assay (11). Blood and fecal pellets were also collected at days −5, 14, and 28 postinoculation (pi) for determination of Ab titers. Sera were prepared from the blood specimens by centrifugation (800 × g, 20 min, 4°C) after clotting. Fecal pellets were homogenized (200 mg/ml) in PBS containing a complete protease inhibitor mixture (Roche Diagnostics) and centrifuged (20,000 × g, 10 min, 4°C), and the supernatants were collected as fecal extracts. The sera and fecal extracts were stored at −20°C until use for the measurement of Ab titers. To evaluate colonic hyperplasia, the colonic tissue was weighed after the terminal 2-cm portion of each mouse colon was flushed with 3 ml of sterile PBS to remove luminal contents. To measure bacterial translocation to the visceral organs, the liver was homogenized in PBS. To quantify the number of epithelia-associated C. rodentium, the terminal 2-cm portion of each mouse colon was flushed as described above and then homogenized in PBS containing 5 mM EDTA. The tissue homogenates were serially diluted with PBS and plated onto MacConkey agar plates containing kanamycin, and C. rodentium colonies identified as pink colonies with narrow clear edge were enumerated after incubation for 24 h at 37°C.

The Δ1–257 aa Tir and the C-terminal 280 aa of intimin of C. rodentium was purified from E. coli BL21 transformed with a plasmid encoding N-terminal GST-tagged Δ1–257 aa Tir and 280 aa of intimin, respectively, by Glutathione Sepharose 4B (GE Healthcare), according to the manufacturer’s instructions. The full-length EspB of C. rodentium was purified as inclusion body from E. coli BL21 transformed with a plasmid encoding N-terminal GST-tagged EspB. Protein concentrations were measured using a BCA protein assay kit (Thermo Fisher Scientific). Aliquots of the proteins were stored at −80°C until use.

C. rodentium LPS was extracted by the modified hot phenol method as previously described (16). The extracted LPS was dissolved in distilled water and stored at −80°C until use.

For measurement of the Ab titers against C. rodentium by ELISA 96-well plates (MaxiSorp, Nunc) were coated with the fixed whole bacterial body of C. rodentium as previously described (11). Sera (1:400 for IgG and 1:100 for IgM) or fecal extracts (1:10 for IgA) diluted with PBST containing 5% skim milk were then added and incubated for 1 h at 37°C. HRP-conjugated polyclonal goat anti-mouse IgG, IgA, or IgM Abs (1:1000; all from SouthernBiotech) were added and further incubated for 1 h at 37°C. Plates were developed using o-phenylenediamine substrate, the reaction was stopped by adding H2SO4, and OD492 values were then measured. For adsorption, sera were preincubated with 50 µg of purified LPS from C. rodentium or E. coli B55:O5 (Sigma-Aldrich) as a control for 30 min at 37°C. For measurement of Abs against GST-Tir, -intimin, or -EspB, 96-well plates were coated with 100 ng/well of each recombinant protein.

The terminal 1-cm portion of each mouse colon was fixed overnight in 4% paraformaldehyde at 4°C and then incubated in 20% sucrose solution for 24 h. The fixed colon was embedded in OCT media (Sakura Finetek Japan) and flash frozen with liquid nitrogen. Ten-micrometer-thick sections were cut and rinsed twice in PBS and once in PBS containing 2% BSA. The sections were stained with anti-O152 serum (1:200) for 1 h at room temperature. After washing with PBS, the sections were stained with an Alexa488-conjugated secondary Ab (1:200) and Hoechst 33342. Images of randomly selected sections were acquired as described for in vitro adherence analysis.

For transcriptomic analysis, total RNA was extracted from the terminal 2-cm portion of each mouse colon using TRIzol (Thermo Fisher Scientific) and further purified using NucleoSpin RNA (TAKARA) with DNase I treatment according to the manufacturer’s instructions. RNA library preparation was performed using the NEBNext Ultra RNA Library Prep Kit for Illumina (New England Biolabs) according to the manufacturer’s instructions. After assessing the library quality, sequencing was performed using Illumina Novaseq 6000 with read configuration of 150-bp, paired-ended reads at a commercial service company (Gene Nex). A total of at least 20 million reads was generated for each sample. Two biological replicates for each condition were tested. The DNA Data Bank of Japan accession numbers for the RNA sequencing (RNA-seq) data reported in this paper are DRA011295, DRA011296, DRA011297, DRA011298, DRA011299, DRA011300, DRA011301, and DRA011302 (https://ddbj.nig.ac.jp/DRASearch/). Paired-ended clean reads were aligned to the Mus musculus reference genome (mm10, National Center for Biotechnology Information/University of California, Santa Cruz/Ensembl) using the Spliced Transcripts Alignment to a Reference v2.5 program. HTSeq v0.6.1 was used to count the read numbers mapped for each gene. Fragments per kilobase per million reads of each gene were calculated based on the length of the gene and read count mapped to the gene. Differential expression analysis between two conditions per groups was performed using DESeq2 R v2_1.6.3. Gene Ontology (GO) enrichment analysis of differentially expressed genes (DEGs) was implemented using the clusterProfiler R package. GO terms with a corrected p value <0.05 were considered significantly enriched by DEGs. mRNA levels of several genes were further validated by qRT-PCR using primers listed in Supplemental Table I (n = 3).

Data were analyzed and graphed by GraphPad Prism 8 (GraphPad Software). Data were analyzed by unpaired Student t test or one-way ANOVA or Kruskal–Wallis test followed by Dunnett or Tukey post hoc test. Any p values <0.05 were considered significant.

The extracellular central region (258–359 aa) and the intracellular carboxyl terminal region (383–547 aa) of the Tir protein are required for intimin binding and the induction of actin polymerization, respectively (5, 17, 18). Although many studies have shown that tyrosine residues at the C terminus of Tir are involved in intimate adhesion and A/E lesion formation in vitro, Crepin et al. (8) suggested that an unknown compensatory mechanism is involved in vivo. Therefore, to investigate the role of Tir-mediated bacterial intimate adhesion both in vitro and in vivo and host Ab responses, we constructed mutant strains Tir-M and Tir-N that lack the entire C-terminal region to exclude the effect of such a compensatory mechanism. Tir-M and Tir-N strains lack 412–547 aa and 172–547 aa of Tir, respectively (Fig. 1A). A C. rodentium mutant expressing the 547 aa authentic Tir (Tir-F) was also constructed by the same genetic manipulations (Fig. 1A). The three mutants and the parental WT strain, a C. rodentium strain constitutively producing luciferase (11), had identical growth rates in LB media (Fig. 1B). In addition, the relationships between the luciferase activities and the number of live bacteria were identical in each mutant (Fig. 1C).

FIGURE 1.

Growth C. rodentium mutants and their expression of virulence factors.

(A) Schematic representation of genome structures of the parental WT and the Tir mutants of C. rodentium used in this study. Filled boxes show the two transmembrane helices (TM) spanning aa 231–257 and aa 360–382, respectively. C. rodentium mutant strains encoding authentic (547 aa) or truncated (Δ412–547 or Δ172–547) forms of Tir are indicated as Tir-F, Tir-M, or Tir-N, respectively. The striped box shows the Δ1–257-aa recombinant Tir protein used to prepare anti-Tir sera. Kmr, kanamycin-resistance gene cassette. (B) Growth rates of WT and Tir-F, -M, and -N mutants cultured in LB medium. Overnight culture of the WT strain or the mutants was subcultured in fresh LB medium at 1:100, then OD600 was measured hourly for 8 h. Data represent means ± SD (n = 3, error bars are behind the symbols due to their small values). (C) Correlations between CFUs and the luciferase activities (relative light units [RLUs]) of the mutants. The WT strain and the mutants cultured in LB media for 8 h were serially diluted, and each of CFU and RLU were measured (n = 3). (D) Soluble fraction of cell lysates of the WT strain and Tir-F, -M, and -N mutants cultured in DMEM for 6 h were separated by SDS-PAGE, followed by Coomassie Brilliant Blue (CBB) staining (left) and Western blotting (WB) with anti-Tir sera (right). An arrowhead indicates the full-length Tir, and an arrow indicates Tir-M. (E) mRNA expression levels of tir, ces-T, eae, ler, and espB were determined by qRT-PCR using total RNA extracted from the WT strain and Tir-F, -M, and -N mutants cultured in LB medium or DMEM for 6 h (means ± SD, n = 3). Data from one representative experiment of two independent experiments are shown.

FIGURE 1.

Growth C. rodentium mutants and their expression of virulence factors.

(A) Schematic representation of genome structures of the parental WT and the Tir mutants of C. rodentium used in this study. Filled boxes show the two transmembrane helices (TM) spanning aa 231–257 and aa 360–382, respectively. C. rodentium mutant strains encoding authentic (547 aa) or truncated (Δ412–547 or Δ172–547) forms of Tir are indicated as Tir-F, Tir-M, or Tir-N, respectively. The striped box shows the Δ1–257-aa recombinant Tir protein used to prepare anti-Tir sera. Kmr, kanamycin-resistance gene cassette. (B) Growth rates of WT and Tir-F, -M, and -N mutants cultured in LB medium. Overnight culture of the WT strain or the mutants was subcultured in fresh LB medium at 1:100, then OD600 was measured hourly for 8 h. Data represent means ± SD (n = 3, error bars are behind the symbols due to their small values). (C) Correlations between CFUs and the luciferase activities (relative light units [RLUs]) of the mutants. The WT strain and the mutants cultured in LB media for 8 h were serially diluted, and each of CFU and RLU were measured (n = 3). (D) Soluble fraction of cell lysates of the WT strain and Tir-F, -M, and -N mutants cultured in DMEM for 6 h were separated by SDS-PAGE, followed by Coomassie Brilliant Blue (CBB) staining (left) and Western blotting (WB) with anti-Tir sera (right). An arrowhead indicates the full-length Tir, and an arrow indicates Tir-M. (E) mRNA expression levels of tir, ces-T, eae, ler, and espB were determined by qRT-PCR using total RNA extracted from the WT strain and Tir-F, -M, and -N mutants cultured in LB medium or DMEM for 6 h (means ± SD, n = 3). Data from one representative experiment of two independent experiments are shown.

Close modal

Immunoblot analysis confirmed that the Tir-F and Tir-M mutants expressed the intended forms of Tir, although Tir-N was undetectable because of the lack of reactive capacity of the anti-Tir sera that raised against Δ1–257 aa recombinant Tir (Fig. 1A, 1D). Immunoblot analysis also showed the reduced expression of Tir protein in the Tir-M strain in comparison with that in Tir-F (Fig. 1D). To test the potential pleotropic effects of insertion of a kanamycin-resistance gene cassette into the tir gene locus to the gene expression of LEE virulence factors, the mRNA levels of tir and other virulence factors induced by DMEM stimulation were evaluated. The parental WT stain and all mutants showed comparable mRNA induction levels of tir, suggesting that reduced expression of Tir protein in the Tir-M strain was not due to a decrease at the transcriptional level (Fig. 1E). Although the genes that lie directly downstream of the inserted kanamycin-resistance gene cassette (including ces-T that encodes a chaperon protein and eae) were constitutively expressed irrespective of DMEM stimulation because of the constitutively active nature of the kanamycin promoter, their expression levels were similar in the three generated strains (Fig. 1E). In contrast, mRNA induction of ler, a global transcriptional regulator of LEE-encoded genes located in LEE1, and espB, a secretory component of T3SS located in LEE4, was dependent on DMEM stimulation, whereas mRNA level of those genes was comparable in all strains, suggesting that the insertion of kanamycin-resistance gene cassette into the genome did not affect the overall gene expression of LEE virulence factors (Fig. 1E).

To characterize the adhesion phenotype of the mutants in vitro, CMT-93, a murine rectal carcinoma cell line, was cocultured with each mutant and subjected to immunostaining. In the cells cocultured with the Tir-F strain, polymerized actin was colocalized with Tir as well as with the bacteria and formed a pedestal-like structure, suggesting that actin polymerization occurred under the adhered bacteria as in the cells cocultured with the parental WT strain (Fig. 2A–C). Although the Tir-M strain did not induce actin polymerization and the number of adhered bacteria per cells was significantly decreased than Tir-F strain (Fig. 2A, 2B), half of the adhered bacteria were apparently colocalized with Tir (Fig. 2C), suggesting that a part of Tir-M–expressing bacteria bound to truncated Tir present on the cells in the absence of actin polymerization. In the cells cocultured with the Tir-N strain, the number of adhered bacteria was comparable with that of Tir-M strain (Fig. 2A, 2B). Again, no polymerized actin was observed under the Tir-N strain (Fig. 2A, 2C). The lack of actin polymerization in the cells cocultured with the Tir-N strain was also confirmed in higher-powered Z-stack images (Fig. 3). These observations indicated that the mutants had the intended adherence phenotypes.

FIGURE 2.

Adherent phenotype of the mutants.

(A) Representative images of C. rodentium WT or Tir-F, -M, and -N mutants bound to CMT-93 cells. The number of cells used for quantification analysis in (B) are shown in each merged image panel. Markers used were as follows: C. rodentium (anti-O152, green), F-actin (phalloidin, red), Tir (anti-Tir, violet) and cell nuclei (Hoechst, blue). NA, Not applicable (the Tir-N protein does not contain the region that reacts with the anti-Tir sera used). Scale bars, 5 µm. (B) Enumeration of adhered bacteria per infected cells. Bars indicate median of the data. ****p < 0.0001. ns, not significant. (C) Characteristics of C. rodentium mutants (the numbers of Tir-F, -M, and -N were 3946, 946, and 676, respectively) adhered to cells were classified based on the presence (positive or negative) of tir and polymerized actin.

FIGURE 2.

Adherent phenotype of the mutants.

(A) Representative images of C. rodentium WT or Tir-F, -M, and -N mutants bound to CMT-93 cells. The number of cells used for quantification analysis in (B) are shown in each merged image panel. Markers used were as follows: C. rodentium (anti-O152, green), F-actin (phalloidin, red), Tir (anti-Tir, violet) and cell nuclei (Hoechst, blue). NA, Not applicable (the Tir-N protein does not contain the region that reacts with the anti-Tir sera used). Scale bars, 5 µm. (B) Enumeration of adhered bacteria per infected cells. Bars indicate median of the data. ****p < 0.0001. ns, not significant. (C) Characteristics of C. rodentium mutants (the numbers of Tir-F, -M, and -N were 3946, 946, and 676, respectively) adhered to cells were classified based on the presence (positive or negative) of tir and polymerized actin.

Close modal
FIGURE 3.

Z-stack analysis of adhesion of the mutants in vitro.

Representative Z-stack images of C. rodentium Tir-F, -M, and -N mutants bound to CMT 93 cells. Markers used were as follows: C. rodentium (anti-O152, green), F-actin (phalloidin, red), Tir (anti-Tir, violet), and cell nuclei (Hoechst, blue). Scale bars, 5 µm.

FIGURE 3.

Z-stack analysis of adhesion of the mutants in vitro.

Representative Z-stack images of C. rodentium Tir-F, -M, and -N mutants bound to CMT 93 cells. Markers used were as follows: C. rodentium (anti-O152, green), F-actin (phalloidin, red), Tir (anti-Tir, violet), and cell nuclei (Hoechst, blue). Scale bars, 5 µm.

Close modal

It is known that a mutant strain lacking the C-terminal intracellular region or the extracellular region of Tir shows reduced or deficient colonization in specific pathogen–free mice, respectively (17, 19). Therefore, to forcibly colonize the mutants, mice were treated with an antibiotic mixture containing four antibiotics, followed by oral inoculation of the mutants for three consecutive days (Fig. 4A). Although inoculation of germ-free mice with Tir mutant strain is usually performed with a single dose, the inoculation in this model was performed for three consecutive days to exclude any effects of remaining commensal bacteria in antibiotic-treated mice on the capacities of colonization of Tir mutants. The number of fecal C. rodentium reached >1 × 109 CFU/g of feces immediately after the third inoculation in all mutants, thereafter gradually decreasing, although it remained above 1 × 107 CFU/g even at day 20 pi, suggesting that all mutants colonize the gut with similar kinetics under our experimental conditions (Fig. 4B).

FIGURE 4.

Colonization of the mutants in the gut of mice.

(A) Scheme of the animal experimental course. Mice were treated with an antibiotic (Abx) mixture containing ampicillin, neomycin, lincomycin, and vancomycin in drinking water for 5 d. C. rodentium was orally inoculated for three consecutive days (days 0–2, 1 × 1010 to 2 × 1010 CFU/dose). After day 0, mice were treated with kanamycin in drinking water over the experimental course. (B) Mice were infected orally with Tir-F, -M, or -N or PBS (Uninfected), and the bacterial number in feces was determined at days −1, 3, 6, 9, 13, and 20 pi (means ± SEM, n = 7–10). The dashed line indicates the detection limit. Results are a pool of two independent experiments.

FIGURE 4.

Colonization of the mutants in the gut of mice.

(A) Scheme of the animal experimental course. Mice were treated with an antibiotic (Abx) mixture containing ampicillin, neomycin, lincomycin, and vancomycin in drinking water for 5 d. C. rodentium was orally inoculated for three consecutive days (days 0–2, 1 × 1010 to 2 × 1010 CFU/dose). After day 0, mice were treated with kanamycin in drinking water over the experimental course. (B) Mice were infected orally with Tir-F, -M, or -N or PBS (Uninfected), and the bacterial number in feces was determined at days −1, 3, 6, 9, 13, and 20 pi (means ± SEM, n = 7–10). The dashed line indicates the detection limit. Results are a pool of two independent experiments.

Close modal

Serum IgG and mucosal IgA against C. rodentium in mice infected with each mutant showed a significant increase, and no significant differences in those Ab titers were observed among the strains (Fig. 5A). The binding of the serum IgG to C. rodentium was robustly inhibited by pretreatment of the sera with LPS purified from C. rodentium, but not by pretreatment with that from E. coli, suggesting that LPS is the primary target of C. rodentium–specific IgG (Fig. 5B). As it has been reported that Abs against virulence factors are produced during C. rodentium infection in mice (20), the titers of serum IgG and fecal IgA Abs against virulence factors (Tir, EspB, and intimin) were determined. Abs against the virulence factors were detected in the Tir-F–infected mice, but those in the Tir-M– or Tir-N–infected mice were almost undetectable (Fig. 5C, 5D). In addition, the level of serum IgM Abs specific to those Ags showed a trend similar to those of serum IgG and fecal IgA (Supplemental Fig. 1), suggesting that the early steps of the immune responses that lead to the activation of adaptive immune responses, such as Ag capturing, presentation to T cells, germinal center reactions, and so on may not occur in the Tir-M– or Tir-N–infected mice. To determine the mechanism underlying the phenomenon by which only Tir-F induced antivirulence factor Ab production, the following experiments were performed.

FIGURE 5.

Ab responses against whole C. rodentium in mice infected with the mutants.

Mice were infected orally with Tir-F, -M, or -N or uninfected. (A) The levels of serum IgG (left panel) and fecal IgA (right panel) against C. rodentium determined by ELISA (means ± SEM, n = 6–10). The samples collected at day 28 pi were used. Results are a pool of two independent experiments. (B) The levels of serum IgG against C. rodentium in Tir-F, -M, or -N–infected mice determined by ELISA after adsorption with E. coli (EC)– or C. rodentium–derived LPS. OD values before adsorption of each sample are used as 100% controls (n = 5). Data from one representative experiment of two independent experiments are shown. The levels of serum IgG (C) and fecal IgA (D) against Tir, intimin, and EspB determined by ELISA. The samples collected at day 28 pi were used. Data represent means ± SEM (n = 6–10). Significance was determined by the Kruskal–Wallis test and Dunnett post hoc test for (A), (C), and (D) and unpaired Student t test for (B). Results are a pool of two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

FIGURE 5.

Ab responses against whole C. rodentium in mice infected with the mutants.

Mice were infected orally with Tir-F, -M, or -N or uninfected. (A) The levels of serum IgG (left panel) and fecal IgA (right panel) against C. rodentium determined by ELISA (means ± SEM, n = 6–10). The samples collected at day 28 pi were used. Results are a pool of two independent experiments. (B) The levels of serum IgG against C. rodentium in Tir-F, -M, or -N–infected mice determined by ELISA after adsorption with E. coli (EC)– or C. rodentium–derived LPS. OD values before adsorption of each sample are used as 100% controls (n = 5). Data from one representative experiment of two independent experiments are shown. The levels of serum IgG (C) and fecal IgA (D) against Tir, intimin, and EspB determined by ELISA. The samples collected at day 28 pi were used. Data represent means ± SEM (n = 6–10). Significance was determined by the Kruskal–Wallis test and Dunnett post hoc test for (A), (C), and (D) and unpaired Student t test for (B). Results are a pool of two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Close modal

To assess the pathogenicity of mutants, colonic hyperplasia and the number of the bacteria in the systemic organs were evaluated. Colonic tissue weights were significantly increased in mice infected with the Tir-F or Tir-M strain (Fig 6A). The Tir-N strain–infected mice also showed a tendency toward increased colon weight (Fig 5A). Although significant crypt elongation was seen only in mice infected with the Tir-F strain (Fig. 6B, 6C), hypertrophy of muscularis propria was observed in mice infected with the Tir-M or Tir-N stains as Tir-F strain (Fig. 6B, 6D). In addition, not only the Tir-F strain but also the Tir-M or Tir-N strain was detected in the liver at day 9 pi, suggesting that these bacteria translocated to the systemic organs by breaching the intestinal epithelial barriers (Fig. 6E). These results clearly indicate that the Tir-M and Tir-N strains are still pathogenic in mice at least under our experimental conditions. Therefore, it seems that the pathogenic characteristics are irrelevant to the production of virulence factor–specific Abs.

FIGURE 6.

Pathogenicity of the mutants.

Mice were uninfected or infected with the Tir-F, -M, or -N mutant. (A) Terminal colon weights (day 28 pi) are shown. Each symbol represents a single mouse (means ± SEM, n = 6–10). Results are a pool of two independent experiments. Significance was determined by the ANOVA and Dunnett post hoc test. **p < 0.01, ***p < 0.001. (B) Representative images of colon sections of mice uninfected or infected with Tir-F, -M, or -N mutant at day 28 pi. Markers used are as follows: nuclei (Hoechst 33342, blue) and F-actin (phalloidin, white). Scale bars, 100 μm. M, muscularis propria. (C and D) Colonic crypt length (C) and muscularis propria (D) of mice uninfected or infected with Tir-F, -M, or -N mutant at day 28 pi. Two slides per mouse from different locations of tissue and more than five crypt per section were used to measurement (each symbol represents a measured crypt length or muscularis propria thickness). Bars indicate median of the data. Significance was determined by the ANOVA and Dunnett post hoc test. ***p < 0.001, ****p < 0.0001. ns, not significant. (E) The numbers of C. rodentium in the liver at day 9 pi are showed. Each symbol represents a single mouse (means ± SEM, n = 6–8). The dashed line indicates the detection limit. Significance was determined by the ANOVA and Tukey post hoc test. ns, not significant. (F) mRNA levels of ler, tir, eae, and espB determined by qRT-PCR in the cecum of mice infected with the Tir-F, -M, or -N mutant at day 9 pi are shown (means ± SEM, n = 13–15). Expression levels of the luciferase gene encoded by the C. rodentium strains were used for normalization. Data represent the mRNA expression relative to those in WT C. rodentium cultured in LB medium (WT [LB in vitro], n = 3). Significance was determined by the ANOVA and Tukey post hoc test. Results are a pool of two independent experiments. ns, not significant.

FIGURE 6.

Pathogenicity of the mutants.

Mice were uninfected or infected with the Tir-F, -M, or -N mutant. (A) Terminal colon weights (day 28 pi) are shown. Each symbol represents a single mouse (means ± SEM, n = 6–10). Results are a pool of two independent experiments. Significance was determined by the ANOVA and Dunnett post hoc test. **p < 0.01, ***p < 0.001. (B) Representative images of colon sections of mice uninfected or infected with Tir-F, -M, or -N mutant at day 28 pi. Markers used are as follows: nuclei (Hoechst 33342, blue) and F-actin (phalloidin, white). Scale bars, 100 μm. M, muscularis propria. (C and D) Colonic crypt length (C) and muscularis propria (D) of mice uninfected or infected with Tir-F, -M, or -N mutant at day 28 pi. Two slides per mouse from different locations of tissue and more than five crypt per section were used to measurement (each symbol represents a measured crypt length or muscularis propria thickness). Bars indicate median of the data. Significance was determined by the ANOVA and Dunnett post hoc test. ***p < 0.001, ****p < 0.0001. ns, not significant. (E) The numbers of C. rodentium in the liver at day 9 pi are showed. Each symbol represents a single mouse (means ± SEM, n = 6–8). The dashed line indicates the detection limit. Significance was determined by the ANOVA and Tukey post hoc test. ns, not significant. (F) mRNA levels of ler, tir, eae, and espB determined by qRT-PCR in the cecum of mice infected with the Tir-F, -M, or -N mutant at day 9 pi are shown (means ± SEM, n = 13–15). Expression levels of the luciferase gene encoded by the C. rodentium strains were used for normalization. Data represent the mRNA expression relative to those in WT C. rodentium cultured in LB medium (WT [LB in vitro], n = 3). Significance was determined by the ANOVA and Tukey post hoc test. Results are a pool of two independent experiments. ns, not significant.

Close modal

Commensal bacteria (21), microbiota-derived metabolites (21, 22), and host responses (20) affect the production of virulence factors in various complicated ways. As the amount of virulence factors produced in the intestine affects virulence factor–specific Abs production, the mRNA levels of virulence factors of each strain in the cecum were compared. There were no significant differences in the mRNA levels of the ler, tir, eae, and espB genes among the strains, although there were considerable variations among the mRNA expression levels (Fig. 6F). Therefore, it is unlikely that the amount of the virulence factors is the main determinant for the production of virulence factor–specific Abs.

Previous studies have reported that specific Abs are preferentially produced against potentially pathogenic bacteria that invade the normally sterile inner mucus layer (2325). Therefore, we analyzed the localization of the mutants in the colonic tissue. Immunofluorescence microscopic analyses of colon cross-sections showed that the Tir-F strain adhered to colonic epithelial cells, spread widely on the epithelial surfaces, and formed numerous microcolonies at day 9 pi (Fig. 7A, 7D). Although adhesion of the Tir-M or Tir-N bacteria to the colonic epithelial cells was also observed, most of these mutants bound to the luminal contents and the number of epithelia adherent bacteria was evidently smaller than that of Tir-F (Fig. 7B, 7C, 7E, 7F). Quantitative analysis after washing out the nonadherent luminal bacteria showed that the number of tissue-associated adherent Tir-M or Tir-N bacteria was significantly decreased in comparison with that of Tir-F (Fig. 7G). These results suggest that luminal localization of Tir-M or Tir-N strain is one possible reason for their defective virulence factor–specific Ab production in mice.

FIGURE 7.

Localization of the mutants in sections of colonic tissue.

Representative images of immunofluorescence-stained colon sections from mice infected with Tir-F (A and D), -M (B and E), or -N (C and F) at day 9 pi are shown. Markers used are as follows: C. rodentium (anti-O152, green) and F-actin (phalloidin, white). (D), (E), and (F) show magnified images of the regions indicated with boxes in (A), (B), and (C), respectively. Scale bars, 100 μm. (G) The number of colonic tissue–associated C. rodentium obtained from the mice uninfected or infected with the Tir-F, -M, or -N mutant at day 9 pi (means ± SEM, n = 4–9). The dashed line indicates the detection limit. Significance was determined by the ANOVA and Tukey post hoc test. Results are a pool of two independent experiments. ****p < 0.0001. ns, not significant.

FIGURE 7.

Localization of the mutants in sections of colonic tissue.

Representative images of immunofluorescence-stained colon sections from mice infected with Tir-F (A and D), -M (B and E), or -N (C and F) at day 9 pi are shown. Markers used are as follows: C. rodentium (anti-O152, green) and F-actin (phalloidin, white). (D), (E), and (F) show magnified images of the regions indicated with boxes in (A), (B), and (C), respectively. Scale bars, 100 μm. (G) The number of colonic tissue–associated C. rodentium obtained from the mice uninfected or infected with the Tir-F, -M, or -N mutant at day 9 pi (means ± SEM, n = 4–9). The dashed line indicates the detection limit. Significance was determined by the ANOVA and Tukey post hoc test. Results are a pool of two independent experiments. ****p < 0.0001. ns, not significant.

Close modal

To analyze the effects of bacterial adhesion to the colonic epithelia on the host colonic gene expression profile, transcriptomic analysis was performed. Principal component analysis showed that the gene expression patterns of Tir-F, Tir-M, and Tir-N were completely different (Fig. 8A). In the Tir-F–infected mice, robust changes in gene expression were observed in comparison with those in uninfected mice or Tir-M– or Tir-N–infected mice (Fig. 8B). GO analysis of DEGs between Tir-infected and -uninfected mice indicated the upregulation of the genes related to inflammatory and immune responses, such as acute phase protein serum amyloid A (Saa)3, Saa4, Lrg1, Steap4, reactive oxygen species–related genes (Duox2 and Duoxa2), Nos2, leukocyte migration–related genes (chemokines, chemokine receptors, and Plet1), T cell immunity–related genes (Cd4, Cd8, Cd28, Gzma, Gzmb, and Il17a/f), and Ab genes (Igha, Ighm, Ighg, and L chain genes), in addition to the genes involved in the cell cycle (Fig. 8C, 8D, Supplemental Table II). The small number of DEGs in comparison of uninfected mice with the Tir-M– or Tir-N–infected mice indicated that the infection of these mutants did not result in robust changes in gene expression (Fig. 8B, Supplemental Fig. 2A, 2B). DEGs analysis also showed a small number of DEGs between the Tir-M– and Tir-N–infected mice (Fig. 8B, Supplemental Fig. 2C). These results support our data that Tir-M and Tir-N infection resulted in similar phenotypes. GO analysis suggested that the expression of genes related to immune reactions were rather decreased in the Tir-M– or Tir-N–infected mice compared with that in the uninfected mice, whereas the genes for several complement components, Abs, neutrophil markers, and chemokines tended to increase (Supplemental Fig. 2D–F). Validation of the transcriptome data by qRT-PCR showed that the relative expression levels of the tested genes exhibited profiles similar to those extracted from the RNA-seq datasets (Fig. 8E). Thus, it is plausible that intimate adhesion induces inflammation in the colon and promotes Ab production against virulence factors by the recruitment of immune cells, including APCs, B cells, and T cells, to the sites of bacterial adhesion. It is noteworthy that nonadherent infection or adhesion by Tir–intimin binding in the absence of actin polymerization may rather suppress the inflammatory and immune responses to the infected bacteria.

FIGURE 8.

Transcriptomic analysis of the colonic tissues.

Mice were uninfected or infected with the Tir-F, -M, or -N mutant. At day 9 pi, RNA-seq was performed using total RNA extracted from the terminal colonic tissues (biological replicate = 2 per group). (A) Principal component analysis of the RNA-seq data. (B) The number of DEGs calculated using DEseq2 (false discovery rate [FDR] cutoff = 0.1, minimum (min) log fold change [FC] = 1). F, Tir-F; M, Tir-M; N, Tir-N; U, uninfected. (C) Volcano plot represents DEGs between the uninfected and Tir-F–infected groups. (D) GO analysis of genes upregulated in Tir-F–infected mice compared with those in uninfected mice. The x-axis indicates −Log10 adjusted p values (Padj), and the y-axis indicates the top 15 enriched biological processes. (E) Colonic tissue mRNA expression levels measured by qRT-PCR (means ± SEM, n = 3). Significance was determined by the ANOVA and Dunnett post hoc test. *p < 0.05, **p < 0.01, ****p < 0.0001.

FIGURE 8.

Transcriptomic analysis of the colonic tissues.

Mice were uninfected or infected with the Tir-F, -M, or -N mutant. At day 9 pi, RNA-seq was performed using total RNA extracted from the terminal colonic tissues (biological replicate = 2 per group). (A) Principal component analysis of the RNA-seq data. (B) The number of DEGs calculated using DEseq2 (false discovery rate [FDR] cutoff = 0.1, minimum (min) log fold change [FC] = 1). F, Tir-F; M, Tir-M; N, Tir-N; U, uninfected. (C) Volcano plot represents DEGs between the uninfected and Tir-F–infected groups. (D) GO analysis of genes upregulated in Tir-F–infected mice compared with those in uninfected mice. The x-axis indicates −Log10 adjusted p values (Padj), and the y-axis indicates the top 15 enriched biological processes. (E) Colonic tissue mRNA expression levels measured by qRT-PCR (means ± SEM, n = 3). Significance was determined by the ANOVA and Dunnett post hoc test. *p < 0.05, **p < 0.01, ****p < 0.0001.

Close modal

In this study, we constructed C. rodentium strains with the same genetic backgrounds except for the truncations in the tir gene, product of which is responsible for the intimate adhesion of the bacteria to colonic epithelial cells. Using these mutants, we examined the effects of their adhesion to epithelial cells on specific Ab production. As a result, and as suggested in previous reports (2325), we clearly showed that intimate bacterial adhesion to the colonic epithelial cells plays an essential role in Ab production by inducing robust inflammatory and immune responses. Surprisingly, the lack of Ab production in the mutants with defective adhesion was seen only for protein Ags, but not for LPS, a typical T cell–independent Ag.

To determine the mechanisms underlying the reason why virulence factor–specific Ab production occurred only in Tir-F, we performed transcriptomic analysis, and it clearly showed that robust inflammatory and immune responses occurred only in the Tir-F–infected mice. Many of the genes elevated in the Tir-F–infected mice in our experiment were consistent with those elevated in germ-free mice infected with C. rodentium or segmented filamentous bacteria (10). Tir-F–infected mice showed strong enhancement of gene expression of Plet1, a recently characterized intestinal dendritic cell (DC) marker (26), suggesting that DC-mediated Ag presentation to T cells is enhanced in the Tir-F–infected mice. Although the numbers of C. rodentium in the fecal samples were similar among the mutant strains, the number of bacteria adhered to the colonic epithelia decreased significantly in Tir-M and Tir-N in comparison with Tir-F. These observations allow us to derive the following scenario. In Tir-F–infected mice, intimate adhesion of the bacteria induces the production of acute phase protein Saa3/4 and cytokines/chemokines from epithelial cells, followed by recruitment of various immune cells to the site of adhesion, resulting in the induction of Ag-specific T cell–dependent (TD) immunity, which includes virulence factor–specific Ab production. In contrast to infection of mice with Tir-F, infection with Tir mutants (Tir-M and Tir-N) resulted both in little induction of virulence factor–specific Abs in the systemic and mucosal tissues and in little induction of immune response–related genes expression. In addition to the IgG and IgA Ab levels, IgM Abs against virulence factors were also barely induced in Tir mutant–infected mice. We speculate that the loss of Ag capturing and presentation by DC, and the absence of T cell activation in lymphoid tissues, such as Peyer’s patches, mesenteric lymph nodes, or isolated lymphoid follicles, cause the reduced Ab responses in Tir mutant–infected mice. Alternatively, the amount of virulence factor that migrate to the lymph nodes may be limited, normal B cell responses, including initial IgM responses, class switching, and somatic hypermutations, may not occur. This aspect could be further in investigated by BCR IgH sequencing.

It would be important to consider what the key event that cues host immune responses against adherent bacteria is. Actin remodeling in the intestinal epithelial cells is a common phenomenon observed during the adherence of various pathogenic and a part of nonpathogenic bacterial strains, including A/E pathogens, Salmonella, Shigella, segmented filamentous bacteria, and Akkermansia muciniphila, to the host epithelial cells (27, 28). In addition, the Tir–intimin interaction associated with A/E pathogen infection affects the gene expression of the epithelial cells involved in host immune responses (19, 29). In this study, Tir-M infection induced almost no Abs against virulence factors. These studies and our findings suggested that actin remodeling associated with intimate adhesion plays an important role in eliciting specific immune responses against invading pathogens.

However, some features of A/E pathogens complicate the interpretation of our data regarding to the role of actin polymerization. EPEC adherence to epithelial cells via Tir–intimin interaction alters bacterial gene expression on a global scale to prepare the pathogen for an adherent lifestyle (30). Tir-mediated actin assembly also enhances type III secretion by promoting bacterial attachment to host cells (31) and mucosal attachment during animal infection (32). It is likely that Tir-M strain is incapable of activating such positive feedback loops, which results in significant decrease of effector proteins translocated into host cells. As these effectors alter multiple cellular processes, the defect of Tir-M is considered to influence immune responses to the virulence factors. In addition, the C terminus of Tir has activities that alter cellular signaling beyond actin assembly (3335). Further studies to identify the signaling pathway(s) or factor(s) for the initiation of inflammatory and immune responses to C. rodentium will be required.

Surprisingly, in the Tir-M– or Tir-N–infected mice, the entire immune responses were rather suppressed compared with those in uninfected mice, although the induction of Abs against LPS, expression of genes related to immune responses, such as some Ab genes, neutrophil markers, and chemokines, was upregulated, and significant colonic tissue hypertrophy and translocation to the systemic organ occurred. These findings are consistent with a recent study reporting that the expression of genes involved in inflammation and apoptosis, which were enhanced in the EPEC-infected HeLa cells, is suppressed when HeLa cells are infected with a ΔEspFu EPEC mutant defective in induction of actin polymerization (29). Bunker et al. (27) have demonstrated that T cell–independent IgA responses occurs against most commensal bacteria residing in the small intestine. Commensal bacteria, in contrast, do not elicit TD IgA responses. These findings imply that unnecessary immune responses are suppressed against Tir-M, Tir-N, and harmless commensal bacteria, based on the lack of robust adhesion. This hypothesis explains the selective lack of response to TD Ags, such as virulence factors, but not the response to LPS due to the direct activation of B cells by LPS.

It is known that virulence factor–specific IgG Abs play a major role in elimination of C. rodentium infection (20). In our model, infection with the Tir-M or Tir-N did not induce the production of virulence factor–specific IgG Abs. One potential way to confirm the significance of reduced virulence factor–specific Abs in protection is the reinfection of mice with WT C. rodentium previously colonized with WT or Tir-mutated C. rodentium and to assess whether these animals develop colitis and when these animals clear the infection. In addition, these experiments may clarify the role of LPS-specific Abs, which has not been clarified extensively so far, in protection against C. rodentium.

In conclusion, intimate adhesion mediated by Tir–intimin binding plays a critical role in the induction of robust inflammatory and immune responses at the sites of adhesion, resulting in the induction of virulence factor–specific, but not LPS-specific, Ab production in the C. rodentium infection model. Our findings warrant further elucidation of the processes from adhesion to the initiation of inflammatory and immune responses that will increase our understanding of the host responses to the A/E pathogen infection.

We thank Division of Animal Experiment, Life Science Research Center, Gifu University for the use of its animal facility.

The sequences presented in this article have been submitted to the DNA Data Bank of Japan (https://ddbj.nig.ac.jp/DRASearch/) under accession numbers DRA011295, DRA011296, DRA011297, DRA011298, DRA011299, DRA011300, DRA011301, and DRA011302.

This work was supported by a Grant-in-Aid for Young Scientists B (JP19K16649) from the Japan Society for the Promotion of Science to K.T.

The online version of this article contains supplemental material.

Abbreviations used in this article

A/E

attaching/effacing

Alexa488

Alexa Fluor 488

DC

dendritic cell

DEG

differentially expressed gene

EHEC

enterohemorrhagic Escherichia coli

EPEC

enteropathogenic E. coli

GO

Gene Ontology

LB

Luria–Bertani

LEE

locus of enterocyte effacement

pi

postinoculation

qRT-PCR

quantitative RT-PCR

RNA-seq

RNA sequencing

Saa

serum amyloid A

TD

T cell–dependent

Tir

translocated intimin receptor

T3SS

type 3 secretion system

WT

wild-type

1.
Croxen
M. A.
,
R. J.
Law
,
R.
Scholz
,
K. M.
Keeney
,
M.
Wlodarska
,
B. B.
Finlay
.
2013
.
Recent advances in understanding enteric pathogenic Escherichia coli.
Clin. Microbiol. Rev.
26
:
822
880
.
2.
Wong
A. R. C.
,
J. S.
Pearson
,
M. D.
Bright
,
D.
Munera
,
K. S.
Robinson
,
S. F.
Lee
,
G.
Frankel
,
E. L.
Hartland
.
2011
.
Enteropathogenic and enterohaemorrhagic Escherichia coli: even more subversive elements.
Mol. Microbiol.
80
:
1420
1438
.
3.
McDaniel
T. K.
,
K. G.
Jarvis
,
M. S.
Donnenberg
,
J. B.
Kaper
.
1995
.
A genetic locus of enterocyte effacement conserved among diverse enterobacterial pathogens.
Proc. Natl. Acad. Sci. USA
92
:
1664
1668
.
4.
Collins
J. W.
,
K. M.
Keeney
,
V. F.
Crepin
,
V. A. K.
Rathinam
,
K. A.
Fitzgerald
,
B. B.
Finlay
,
G.
Frankel
.
2014
.
Citrobacter rodentium: infection, inflammation and the microbiota.
Nat. Rev. Microbiol.
12
:
612
623
.
5.
Hartland
E. L.
,
M.
Batchelor
,
R. M.
Delahay
,
C.
Hale
,
S.
Matthews
,
G.
Dougan
,
S.
Knutton
,
I.
Connerton
,
G.
Frankel
.
1999
.
Binding of intimin from enteropathogenic Escherichia coli to Tir and to host cells.
Mol. Microbiol.
32
:
151
158
.
6.
DeVinney
R.
,
M.
Stein
,
D.
Reinscheid
,
A.
Abe
,
S.
Ruschkowski
,
B. B.
Finlay
.
1999
.
Enterohemorrhagic Escherichia coli O157:H7 produces Tir, which is translocated to the host cell membrane but is not tyrosine phosphorylated.
Infect. Immun.
67
:
2389
2398
.
7.
Brady
M. J.
,
K. G.
Campellone
,
M.
Ghildiyal
,
J. M.
Leong
.
2007
.
Enterohaemorrhagic and enteropathogenic Escherichia coli Tir proteins trigger a common Nck-independent actin assembly pathway.
Cell. Microbiol.
9
:
2242
2253
.
8.
Crepin
V. F.
,
F.
Girard
,
S.
Schüller
,
A. D.
Phillips
,
A.
Mousnier
,
G.
Frankel
.
2010
.
Dissecting the role of the Tir:Nck and Tir:IRTKS/IRSp53 signalling pathways in vivo.
Mol. Microbiol.
75
:
308
323
.
9.
Nagano
K.
,
K.
Taguchi
,
S.
Tokoro
,
I.
Tatsuno
,
H.
Mori
.
2014
.
Adhesion of enterohemorrhagic Escherichia coli O157:H7 to the intestinal epithelia is essential for inducing secretory IgA antibody production in the intestine of mice.
Biol. Pharm. Bull.
37
:
409
416
.
10.
Atarashi
K.
,
T.
Tanoue
,
M.
Ando
,
N.
Kamada
,
Y.
Nagano
,
S.
Narushima
,
W.
Suda
,
A.
Imaoka
,
H.
Setoyama
,
T.
Nagamori
, et al
2015
.
Th17 cell induction by adhesion of microbes to intestinal epithelial cells.
Cell
163
:
367
380
.
11.
Takahashi
K.
,
Y.
Hanamura
,
N.
Tokunoh
,
K.
Kassai
,
M.
Matsunishi
,
S.
Watanabe
,
T.
Sugiyama
,
N.
Inoue
.
2019
.
Protective effects of oral immunization with formalin-inactivated whole-cell Citrobacter rodentium on Citrobacter rodentium infection in mice.
J. Microbiol. Methods
159
:
62
68
.
12.
Yokoyama
S.
,
T.
Suzuki
,
S.
Shiraishi
,
N.
Ohishi
,
K.
Yagi
,
S.
Ichihara
,
S.
Itoh
,
H.
Mori
.
2001
.
Construction of deletion mutants of Shiga (-like) toxin genes (stx-1 and/or stx-2) on enterohemorrhagic Escherichia coli (O157:H7).
J. Clin. Biochem. Nutr.
30
:
33
42
.
13.
Takahashi
K.
,
A.
Yano
,
S.
Watanabe
,
P.
Langella
,
L. G.
Bermúdez-Humarán
,
N.
Inoue
.
2018
.
M cell-targeting strategy enhances systemic and mucosal immune responses induced by oral administration of nuclease-producing L. lactis.
Appl. Microbiol. Biotechnol.
102
:
10703
10711
.
14.
Connolly
J. P. R.
,
S. L.
Slater
,
N.
O’Boyle
,
R. J.
Goldstone
,
V. F.
Crepin
,
D.
Ruano-Gallego
,
P.
Herzyk
,
D. G. E.
Smith
,
G. R.
Douce
,
G.
Frankel
,
A. J.
Roe
.
2018
.
Host-associated niche metabolism controls enteric infection through fine-tuning the regulation of type 3 secretion.
Nat. Commun.
DOI: .
15.
Crepin
V. F.
,
J. W.
Collins
,
M.
Habibzay
,
G.
Frankel
.
2016
.
Citrobacter rodentium mouse model of bacterial infection.
Nat. Protoc.
11
:
1851
1876
.
16.
MacLean
L. L.
,
M. B.
Perry
.
2001
.
Structural studies on the O-polysaccharide of the lipopolysaccharide produced by Citrobacter rodentium (ATCC 51459).
Eur. J. Biochem.
268
:
5740
5746
.
17.
Deng
W.
,
B. A.
Vallance
,
Y.
Li
,
J. L.
Puente
,
B. B.
Finlay
.
2003
.
Citrobacter rodentium translocated intimin receptor (Tir) is an essential virulence factor needed for actin condensation, intestinal colonization and colonic hyperplasia in mice.
Mol. Microbiol.
48
:
95
115
.
18.
Kenny
B.
,
R.
DeVinney
,
M.
Stein
,
D. J.
Reinscheid
,
E. A.
Frey
,
B. B.
Finlay
.
1997
.
Enteropathogenic E. coli (EPEC) transfers its receptor for intimate adherence into mammalian cells.
Cell
91
:
511
520
.
19.
Crepin
V. F.
,
M.
Habibzay
,
I.
Glegola-Madejska
,
M.
Guenot
,
J. W.
Collins
,
G.
Frankel
.
2015
.
Tir triggers expression of CXCL1 in enterocytes and neutrophil recruitment during Citrobacter rodentium infection.
Infect. Immun.
83
:
3342
3354
.
20.
Kamada
N.
,
K.
Sakamoto
,
S. U.
Seo
,
M. Y.
Zeng
,
Y. G.
Kim
,
M.
Cascalho
,
B. A.
Vallance
,
J. L.
Puente
,
G.
Núñez
.
2015
.
Humoral immunity in the gut selectively targets phenotypically virulent attaching-and-effacing bacteria for intraluminal elimination.
Cell Host Microbe
17
:
617
627
.
21.
Curtis
M. M.
,
Z.
Hu
,
C.
Klimko
,
S.
Narayanan
,
R.
Deberardinis
,
V.
Sperandio
.
2014
.
The gut commensal Bacteroides thetaiotaomicron exacerbates enteric infection through modification of the metabolic landscape.
Cell Host Microbe
16
:
759
769
.
22.
Pacheco
A. R.
,
M. M.
Curtis
,
J. M.
Ritchie
,
D.
Munera
,
M. K.
Waldor
,
C. G.
Moreira
,
V.
Sperandio
.
2012
.
Fucose sensing regulates bacterial intestinal colonization.
Nature
492
:
113
117
.
23.
Palm
N. W.
,
M. R.
de Zoete
,
T. W.
Cullen
,
N. A.
Barry
,
J.
Stefanowski
,
L.
Hao
,
P. H.
Degnan
,
J.
Hu
,
I.
Peter
,
W.
Zhang
, et al
2014
.
Immunoglobulin A coating identifies colitogenic bacteria in inflammatory bowel disease.
Cell
158
:
1000
1010
.
24.
Elinav
E.
,
T.
Strowig
,
A. L.
Kau
,
J.
Henao-Mejia
,
C. A.
Thaiss
,
C. J.
Booth
,
D. R.
Peaper
,
J.
Bertin
,
S. C.
Eisenbarth
,
J. I.
Gordon
,
R. A.
Flavell
.
2011
.
NLRP6 inflammasome regulates colonic microbial ecology and risk for colitis.
Cell
145
:
745
757
.
25.
Ivanov
I. I.
,
K.
Atarashi
,
N.
Manel
,
E. L.
Brodie
,
T.
Shima
,
U.
Karaoz
,
D.
Wei
,
K. C.
Goldfarb
,
C. A.
Santee
,
S. V.
Lynch
, et al
2009
.
Induction of intestinal Th17 cells by segmented filamentous bacteria.
Cell
139
:
485
498
.
26.
Karrich
J. J.
,
M.
Romera-Hernández
,
N.
Papazian
,
S.
Veenbergen
,
F.
Cornelissen
,
P.
Aparicio-Domingo
,
F. H.
Stenhouse
,
C. D.
Peddie
,
R. M.
Hoogenboezem
,
C. W. J.
den Hollander
, et al
2019
.
Expression of Plet1 controls interstitial migration of murine small intestinal dendritic cells.
Eur. J. Immunol.
49
:
290
301
.
27.
Bunker
J. J.
,
T. M.
Flynn
,
J. C.
Koval
,
D. G.
Shaw
,
M.
Meisel
,
B. D.
McDonald
,
I. E.
Ishizuka
,
A. L.
Dent
,
P. C.
Wilson
,
B.
Jabri
, et al
2015
.
Innate and adaptive humoral responses coat distinct commensal bacteria with immunoglobulin A.
Immunity
43
:
541
553
.
28.
Jepson
M. A.
,
M. A. N. N.
Clark
,
N. L.
Simmons
,
B. H.
Hirst
.
1993
.
Actin accumulation at sites of attachment of indigenous apathogenic segmented filamentous bacteria to mouse ileal epithelial cells.
Infect. Immun.
61
:
4001
4004
.
29.
Martins
F. H.
,
A.
Kumar
,
C. M.
Abe
,
E.
Carvalho
,
M.
Nishiyama-Jr
,
C.
Xing
,
V.
Sperandio
,
W. P.
Elias
.
2020
.
EspFu-mediated actin assembly enhances enteropathogenic Escherichia coli adherence and activates host cell inflammatory signaling pathways.
mBio
11
:
1
18
.
30.
Elbaz
N.
,
Y.
Socol
,
N.
Katsowich
,
I.
Rosenshine
.
2019
.
Control of type III secretion system effector/chaperone ratio fosters pathogen adaptation to host-adherent lifestyle.
mBio
DOI: .
31.
Battle
S. E.
,
M. J.
Brady
,
S. K.
Vanaja
,
J. M.
Leong
,
G. A.
Hecht
.
2014
.
Actin pedestal formation by enterohemorrhagic Escherichia coli enhances bacterial host cell attachment and concomitant type III translocation.
Infect. Immun.
82
:
3713
3722
.
32.
Mallick
E. M.
,
J. J.
Garber
,
V. K.
Vanguri
,
S.
Balasubramanian
,
T.
Blood
,
S.
Clark
,
D.
Vingadassalom
,
C.
Louissaint
,
B.
McCormick
,
S. B.
Snapper
,
J. M.
Leong
.
2014
.
The ability of an attaching and effacing pathogen to trigger localized actin assembly contributes to virulence by promoting mucosal attachment.
Cell. Microbiol.
16
:
1405
1424
.
33.
Smith
K.
,
D.
Humphreys
,
P. J.
Hume
,
V.
Koronakis
.
2010
.
Enteropathogenic Escherichia coli recruits the cellular inositol phosphatase SHIP2 to regulate actin-pedestal formation.
Cell Host Microbe
7
:
13
24
.
34.
Yan
D.
,
X.
Wang
,
L.
Luo
,
X.
Cao
,
B.
Ge
.
2012
.
Inhibition of TLR signaling by a bacterial protein containing immunoreceptor tyrosine-based inhibitory motifs.
Nat. Immunol.
13
:
1063
1071
.
35.
Mokas
S.
,
J. R.
Mills
,
C.
Garreau
,
M.-J.
Fournier
,
F.
Robert
,
P.
Arya
,
R. J.
Kaufman
,
J.
Pelletier
,
R.
Mazroui
.
2009
.
Uncoupling stress granule assembly and translation initiation inhibition.
Mol. Biol. Cell
20
:
2673
2683
.

The authors have no financial conflicts of interest.

This article is distributed under the terms of the CC BY 4.0 Unported license.

Supplementary data