Tir Triggers Expression of CXCL1 in Enterocytes and Neutrophil Recruitment during Citrobacter rodentium Infection

Bacterial strains and growth conditions.

The bacterial strains, plasmids, and primers used in this study are listed in Table 1. Bacteria were grown in Luria-Bertani (LB) medium, M9 minimum medium (35), or Dulbecco's modified Eagle's medium (DMEM) supplemented with kanamycin (50 mg ml⁻¹), ampicillin (100 mg ml⁻¹), and nalidixic acid (50 mg ml⁻¹) as required.

Introduction of site-directed tir mutations into the C. rodentium chromosome.

We used the lambda red-based mutagenesis system (36) to introduce site-directed tir alterations into the endogenous chromosomal tir gene, together with a kanamycin cassette, in the tir-cesT intergenic region for 3′ mutagenesis as described before (14).

Deletion of the DNA segment encoding amino acids 478 to 547 (Δ478-547) within the tir gene was made by inverse PCR on pICC433 (carrying Tir_Y451/Y471) and pICC438 (carrying Tir_Y451A/Y471A) templates, as previously described (14), using primer pair Tir-P478DSV-stop-EcoRI-Rv and down-Tir-EcoRI-Fw. The inverse-PCR product was then digested with EcoRI and the aphT gene cloned into the tir-cesT intergenic region to confer kanamycin resistance, resulting in plasmids pICC1842 and pICC1843, respectively (Table 1).

Deletion of the DNA segment encoding amino acids 382 to 462 (Δ382-462) within tir was made by overlapping PCR. C. rodentium genomic DNA was used to amplify tir base pairs 664 to 1164 using primer pair Tir-upTM1-Fw and Tir-down-TM2-Rv. The primer pair Tir-down-TM2-Y471-Fw and Tir-EcoRI-Rv was used to amplify tir base pairs 1383 to 1644 from pICC433 (carrying Tir_Y471) and pICC438 (carrying Tir_Y471A) templates. The two PCR fragments ([664-1164] and [1383-1644/Y471]) and ([664-1164] and [1383-1644/Y471A]) were PCR overlapped, EcoRI digested, and ligated to the tir-cesT intergenic region (PCR) amplified as previously described by Crepin et al. (14), using primers EcoRI-(tir-cesT)-Fw/NcesT-Rv. The ligated PCR product was then reamplified using primers Tir-upTM1-Fw and NcesT-Rv and cloned into pGEMT vector. The constructs were digested with EcoRI and the aphT gene cloned into the tir-cesT intergenic region to confer kanamycin resistance, resulting in plasmids pICC1844 and pICC1845, respectively (Table 1). All plasmid derivatives were checked by DNA sequencing.

The various deletions (Δ478-547_Y451/Y471 and Δ478-547_Y451A/Y471A; Δ382-462_Y471 and Δ382-462_Y471A) were PCR amplified from pICC1842 and pICC1843, using primers NcesT-Rv/Tir-Up-YY-Fw, and from pICC1844 and pICC1845, using primers NcesT-Rv/Tir-upTM1-Fw, respectively. The PCR products were electroporated into wild-type C. rodentium expressing the lambda red recombinase from plasmid pKD46 (36). The presence of the mutation was confirmed by PCR and DNA sequencing among the kanamycin-resistant clones.

Cell culture.

The Swiss 3T3 cell line was grown, maintained, and infected with the different C. rodentium strains at a multiplicity of infection (MOI) of 50, as described previously (14). Cells were washed at 6 h postinfection with phosphate-buffered saline (PBS), fixed for 15 min in 4% paraformaldehyde, and permeabilized with 0.1% Triton for 4 min. Phalloidin-tetramethyl rhodamine isocyanate (TRITC) (Sigma) was used to stain F actin, while bacterial DNA was counterstained with Hoechst 33342. Tir was stained using rabbit anti-Tir enterohemorrhagic E. coli (EHEC) (37, 38), recognizing the N-terminal domain, and carbocyanine-2-conjugated donkey anti-rabbit IgG (Jackson ImmunoResearch Europe) secondary antibody. Images were acquired using an AxioCam MRm monochrome camera and processed using AxioVision (Carl Zeiss MicroImaging GmbH, Germany).

Tir translocation by C. rodentium was qualitatively assessed by calculating the ratio between the number of total visible bacterial nuclei and the number of bacteria which show Tir staining concentrated in a straight line at the interface between the bacteria. A minimum of 100 bacteria were counted per experiment, and the experiment was performed twice.

Oral infection of mice.

Pathogen-free female 18- to 20-g C57BL/6 mice were purchased from Charles River. All animals were housed in individually HEPA-filtered cages with sterile bedding and free access to sterilized food and water. All animal experiments were performed in accordance with the Animals Scientific Procedures Act (1986) and were approved by the local Ethical Review Committee. Infections were performed twice using four to eight mice per group. Mice inoculated with mock mutant and nonsense mutant strains were included in every experiment. Mice inoculated with the wild-type strain and uninfected mice were included in parallel with mutant strains.

Mice were inoculated by oral gavage with 200 μl of overnight LB-grown C. rodentium suspension in PBS (≈5 × 10⁹ CFU). The number of viable bacteria used as inoculum was determined by retrospective plating onto LB agar containing antibiotics. Stool samples were recovered aseptically at various time points after inoculation, and the number of viable bacteria per gram of stool was determined by plating onto LB agar (39). At days 7 and 14 postinoculation, the mice were culled and the colonic tissues were collected for further analyses.

Sample collection and colonic crypt hyperplasia measurement.

Segments of the terminal colon (0.5 cm) of each mouse were collected, flushed, and fixed in 10% neutral buffered formalin. Formalin fixed tissues were then processed, paraffin-embedded, sectioned at 5 μm and stained with hematoxylin and eosin (H&E) using standard techniques. H&E-stained tissues were evaluated for colonic crypt hyperplasia microscopically without knowledge of the treatment condition used in the study, and the lengths of at least 20 well-oriented crypts from each section from all of the mice per treatment group (n = 4 to 6) were evaluated. H&E-stained tissues were imaged with an Axio Lab.A1 microscope (Carl Zeiss MicroImaging GmbH, Germany), and images were acquired using an Axio Cam ERc5s color camera and computer processed using AxioVision (Carl Zeiss MicroImaging GmbH, Germany).

Additional colonic segments were embedded in optimal cutting temperature (OCT) medium (Raymond A Lamb Limited, United Kingdom) and frozen in dry-ice/ethanol slush for further cryosectioning. Cryosections were then fixed in 3%-paraformaldehyde (PFA) in PBS as previously described (38, 40) and immunostained using primary antibodies at a 1/50 dilution: chicken anti-intimin (14), Ly-6G (RB6-8C5; Santa Cruz), and E-cadherin (CD324; BD Biosciences). Secondary antibodies were used at a 1/100 dilution: Cy3 (103-165-175; Jackson ImmunoResearch), Alexa 488 (712-546-150; Jackson ImmunoResearch), and Cy5 (715-175-150; Jackson ImmunoResearch), respectively. Images were acquired using an AxioCam MRm monochrome camera and processed using AxioVision (Carl Zeiss MicroImaging GmbH, Germany).

Extraction of enterocytes and immunostaining.

A 4-cm segment of the terminal colon was cut longitudinally, placed in 4 ml enterocyte dissociation buffer (1× Hanks' balanced salt solution without Mg and Ca containing 10 mM HEPES, 1 mM EDTA and 5 μl/ml 2-β-mercaptoethanol), and incubated at 37°C with shaking, for 40 min. Leftover tissue was removed by centrifugation (1,900 × g for 5 min) before the samples for each group were pooled and fixed with 1% formaldehyde. Fixed enterocytes (CD45⁻ CD326⁺) were analyzed for purity by flow cytometry using leukocyte marker CD45 and epithelial cell marker CD326 (EpCAM). For immunofluorescence staining, fixed enterocytes were permeabilized with 0.1% Triton and blocked with 1% bovine serum albumin in PBS. Enterocytes were stained with polyclonal rabbit anti-Tir EHEC (GB1320, SK1786) for 20 min, followed by 30 min of incubation with secondary donkey anti-rabbit IgG(H+L) Alexa 488; phalloidin-TRITC (Sigma, P1951) was used for actin and DAPI (4′,6′-diamidino-2-phenylindole) (Invitrogen, D3571) to visualize the nucleus. Tir staining was visualized with an Axio Imager M1 microscope (Carl Zeiss MicroImaging GmbH, Germany), and images were acquired using an AxioCam MRm monochrome camera and computer processed using AxioVision (Carl Zeiss MicroImaging GmbH, Germany).

Isolation of mRNA and qRT-PCR.

mRNA of enterocytes was isolated using an RNeasy minikit according to the manufacturer's instructions (Qiagen). Samples were treated with RQ1 DNase I (Promega) at 37°C for 10 min, followed by 15 min at 72°C. Reverse transcription-PCR (RT-PCR) was carried out by adding Moloney murine leukemia virus (M-MLV) RT (Promega M170B), RT buffer (Promega), random primers (Promega), RNasin (Promega), deoxynucleoside triphosphates (dNTP) (10 mM), and RNase-free water to the DNase-treated RNA extract and incubating at 37°C for 1 h followed by 10 min at 72°C, and cooled samples were stored at −20°C. CXCL1 (KC) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNAs were amplified with primer pairs mCXCL1-F/mCXCL1-R and mGAPDH-F/mGAPDH-R (Table 1), by RT-quantitative PCR (qRT-PCR) using the 7300 Applied Biosystems instrument under standard cycle conditions for Fast SYBR green master mix. Changes in gene expression levels were analyzed relative to the control levels (PBS samples), with GAPDH as a standard, using the ΔΔC_T method.

Sample collection for flow cytometry.

A 4-cm segment of the terminal colon was cut, opened longitudinally, rinsed in sterile PBS and placed in 4 ml of RPMI 1640 supplemented with 10% fetal bovine serum (FBS), penicillin-streptomycin (P/S), GlutaMAX, DNase (Roche, 10104159001), and liberase (Roche, 540112001) in a C-Mac tube (Miltenyi Biotec), followed by tissue dissociation using a gentleMACS dissociator (Miltenyi Biotec). The tissue was homogenized using the “intestine” setting, followed by incubation at 37°C with 5% CO₂ for 30 min in a shaking incubator, and a final dissociation step was performed using the “lung 2” setting. The digested preparation was disrupted to a single-cell suspension by passage through a 70-μm sieve (BD Labware/Falcon, USA; catalog no. 352350) and suspended in RPMI 1640 supplemented with 10% FBS and P/S at 0.5 × 10⁶ to 1 × 10⁶ cells/ml.

Extracellular antigen analysis.

Cells were stained for surface markers as indicated in PBS containing 1% bovine serum albumin with 0.5% sodium azide (PBA) for 30 min at 4°C and fixed with IC fixation buffer (eBioscience). Prior to primary antibody staining, all cells were blocked for Fc receptors (FcR) using mouse FcR blocking reagent (Miltenyi Biotec) for 10 min at 4°C. Antibodies were purchased from BD Pharmingen or eBioscience. Data were acquired on a BD Fortessa III, and 20,000 lymphocytes or myeloid events were analyzed with the FlowJo (Tress Star) analysis program. Data are shown as a percentage of myeloid or lymphocyte gates. Myeloid and lymphocyte gates were determined by their position on the forward- and side-scatter plots generated by the cytometer. A fluorescence minus one (FMO) control was included for each fluorescent marker, and the expression of a particular marker was calculated by subtracting FMO fluorescence values from fluorescent-antibody levels.

Total live cells were assessed by trypan blue exclusion. Forward- and side-scatter gates on FlowJo software were used to gate myeloid cells, and the percentage of this gate was used to determine total number of myeloid cells. The myeloid cell gate was further analyzed to finalize the percentage of neutrophils (CD11b⁺ Ly6G⁺) and establish the total number of neutrophils in each mouse colon.

mIVOC.

The mouse intestinal in vitro organ culture (mIVOC) model was used to assess A/E lesion formation caused by C. rodentium expressing Tir Y451A/Y471A/Δ478-547 as described by Girard et al. (41). Briefly, segments from the terminal colon were inoculated with 50 μl of the appropriate overnight bacterial culture, corresponding to approximately 10⁷ CFU, and incubated at 37°C in 5% CO₂ atmosphere on a seesaw rocker (18 cycles min⁻¹) for 8 h. Explants were gently rinsed with PBS and fixed in 2.5% glutaraldehyde for electron microscopy analysis.

Electron microscopy.

Additional explants/tissue cultured cell samples were processed for electron microscopy, as previously described (38). Samples for scanning electron microscopy (SEM) were examined at an accelerating voltage of 25 kV using a JEOL JSM-5300 scanning electron microscope [JEOL (UK) Ltd., Herts, United Kingdom]. Samples for transmission electron microscopy (TEM) were observed using a Phillips 201 transmission electron microscope at an accelerating voltage of 60 kV (Philips, United Kingdom).

Statistical analysis.

Results are presented as a line plot (colonization) with the mean and its standard deviation. The nonparametric Mann-Whitney test and the nonparametric Kruskal-Wallis test with Bonferroni's corrected a posteriori comparisons were used to conduct pairwise and global statistical analyses, respectively, using commercially available GraphPad InStat v3.06 software (GraphPad Software, San Diego, CA, USA). The Mann-Whitney test with comparison to PBS controls (or as indicated in the figure legends) was used for data obtained by flow cytometry using GraphPad Prism software. A P value of 0.05 was considered significant (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001).

Construction of the 3′ tir chromosomal deletion mutants.

Previously, using a method that allows expression of tir mutants from the C. rodentium chromosome, we have introduced point mutations which have shown that A/E lesion formation in vivo is independent of Tir residues Y451 and Y471 (14). In this study, we used this technique to introduce deletions at the 3′ end of tir (Fig. 1). While inserting a kanamycin cassette coupled to the mutated tir into the tir-cesT intergenic region, we deleted Tir residues 382 to 462 (TirΔ382-462), removing the 80 amino acids downstream of the distal TM helix, and Tir residues 478 to 547 (TirΔ478-547), removing the entire segment downstream of Y471 phosphorylation site (Fig. 1). TirΔ382-462 was made in the context of either Y471 or Y471A (TirΔ382-462_Y471 and TirΔ382-462_Y471A), while TirΔ478-547 was made in the context of either Y451/Y471 or Y451A/Y471A (TirΔ478-547_Y451/Y471 and TirΔ478-547_Y451A/Y471A). As a positive control, we used a mock mutant tir-cesT (TirC-ctrl) in which the kanamycin cassettes was introduced into the intergenic region, which did not affect the tir coding sequence, and as a negative control, a nonsense codon was introduced at Tir position 33 (Tir1-33stop) (14). Growth curves in minimal and rich media confirmed that the mutants and parental wild-type strains had identical growth rates (data not shown).

Testing the carboxy-terminal Tir deletions during infection of cultured cells.

We characterized the behavior of the Tir derivatives in vitro, following infection of Swiss 3T3 fibroblast cells with the C. rodentium mutants as described before (14). This revealed that, as expected, C. rodentium expressing either TirΔ382-462_Y471A or TirΔ478-547_Y451A/Y471A failed to induce actin polymerization (data not shown). In contrast, infection of Swiss 3T3 cells with C. rodentium expressing TirΔ382-462_Y471 or TirΔ478-547_Y451/Y471 revealed robust actin polymerization under all the adherent bacteria that show Tir translocation (Fig. 2A). Importantly, while TirΔ382-462 was translocated as efficiently as the wild-type control Tir, TirΔ478-547 translocated with low efficiency, with Tir staining seen within 50% of the adherent bacteria (Fig. 2B). Interestingly, C. rodentium expressing TirΔ478-547_Y451/Y471 produced longer pedestals. These results suggest that Tir residues 382 to 462 are dispensable for Tir translocation and actin polymerization, while the carboxy terminus of Tir plays a role in translocation and hence indirectly in the efficiency (in terms of both frequency and length) of actin pedestals formed in vitro.

Testing the carboxy-terminal Tir deletions during mouse infection.

We next investigated the impact of the C-terminal Tir deletions on colonization of C. rodentium in vivo by enumerating CFU per gram of stools (CFU g⁻¹) collected daily following oral inoculation of C57BL/6 mice for 8 days. This showed that C. rodentium expressing either TirΔ382-462_Y471 or TirΔ382-462_Y471A colonized the mice similarly to the control strain expressing TirC-ctrl (Fig. 3A) (the variations between groups seen on day 2 postinfection are common during C. rodentium infection and do not have biological relevance; the differences seen on day 3 postinfection are not significant). Transmission electron microscopy (TEM) revealed typical A/E lesions in colons infected with C. rodentium expressing either TirΔ382-462_Y471 or TirΔ382-462_Y471A (Fig. 3B). In contrast, C. rodentium expressing Tir1-33stop was rapidly cleared and failed to initiate an infection, reaching the background level as soon as day 3 postinfection (Fig. 3A). These results show that the Tir segment from residue 382 to 462 is dispensable for colonization and that C. rodentium expressing TirΔ382-462_Y471 or TirΔ382-462_Y471A is capable of forming A/E lesions in vivo.

We then tested the phenotype of C. rodentium expressing TirΔ478-547. Consistent with the in vitro data, this mutant exhibited an intermediate phenotype, reaching a colonization level of 10⁶ CFU/g of stool on day 3 postinfection, which persisted at this level until day 8 (Fig. 3A). The level of colonization was 100-fold lower than that seen in mice infected with C. rodentium expressing TirC-ctrl or TirΔ382-462 but 10⁴-fold higher than that in mice infected with Tir1-33stop. We observed no difference in colonization between mice infected with TirΔ478-547_Y451/Y471 or TirΔ478-547_Y451A/Y471A (Fig. 3A). However, as colonization was below the detection level of TEM, we performed a mouse IVOC infection using C. rodentium expressing TirΔ478-547_Y451A-Y471A. Examination of the samples by SEM revealed A/E lesions, similar to those formed by the wild-type strains, suggesting that the Tir segment from residue 478 to 547 is dispensable for A/E lesion formation on mucosal surfaces but does play a role in colonization (Fig. 3B), probably due to its role in Tir translocation.

Tir induces actin polymerization on enterocytes in vivo.

As the results thus far showed that C. rodentium expressing Tir deletions (residues 382 to 462 and 478 to 547) and substitutions (Y451 and Y471) was able to form A/E lesions, we tested whether Tir induced actin polymerization on enterocytes in vivo. For this, mice were infected with wild-type C. rodentium or C. rodentium expressing Tir_Y451A/Y471A (Fig. 1) and enterocytes were isolated at the peak of colonization at day 7 postinfection. Enterocytes isolated from naive mice as a control exhibited a typical “crown” staining pattern of the BB microvilli (Fig. 4). Enterocytes isolated from infected mice showed a good level of C. rodentium adhesion, with multiple bacteria on individual enterocytes. Tir was detected underneath both attached wild-type C. rodentium and C. rodentium expressing Tir_Y451A/Y471A (Fig. 4). However, while intense actin staining was seen at the site of wild-type C. rodentium infection, enterocytes infected with C. rodentium expressing Tir_Y451A/Y471A exhibited mainly weak actin polymerization at the site of bacterial attachment. This suggests that Tir induces actin polymerization on enterocytes in a process involving the tyrosine residues.

Recruitment of immune cells to the C. rodentium infection site.

As actin remodeling during infection can trigger immune responses (42), we next investigated whether Tir Y451 and Y471 modulate host immune responses. For this, groups of 4 to 6 mice were infected with wild-type and mutant C. rodentium, and recruitment of immune cells was analyzed by flow cytometry of homogenized colons at days 7 (peak of colonization) and day 14 (peak of pathology) postinfection. All the tested strains were shed at equivalent levels (Fig. 5A). The flow cytometry analysis showed that in comparison to the mock-infected (PBS) control mice (baseline readout), infection with either the wild type or mutant C. rodentium resulted in equivalent recruitment of macrophages, CD4⁺ T cells, and B cells on days 7 (data not shown) and 14 (Fig. 5B to D) postinfection. In contrast, significantly fewer neutrophils were recruited to the colon following infection with C. rodentium expressing Tir_Y451A/Y471A compared with the positive-control mice infected with C. rodentium expressing wild-type Tir or mice infected with the single Tir tyrosine mutants Y451A or Y471A (Fig. 5E and F). Similarly, although shedding was at equivalent levels (Fig. 5A), neutrophil recruitment was observed following infection with C. rodentium expressing TirΔ382-462_Y471 but not after infection with C. rodentium expressing TirΔ382-462_Y471A (Fig. 5G and H).

As C. rodentium and Tir interact mainly with enterocytes, we next determined whether the Tir tyrosine residues play a role in expression of proinflammatory chemokines. For this we isolated colonic enterocytes from mice infected with the different C. rodentium strains. The purity of the enterocyte preparation was confirmed by flow cytometry analysis following staining with the leukocyte marker CD45 and epithelial cell marker CD326 (EpCAM), revealing a low level of contamination (Fig. 6A). Using the purified enterocytes in qRT-PCR for the chemokines CXCL1 (KC) and CXCL2 (MIP2-alpha) revealed reduced expression in mice infected with C. rodentium expressing TirY451A-Y471A (Fig. 6B). The attenuated inflammatory responses triggered by C. rodentium expressing Tir_Y451A/Y471A were mirrored by a significantly reduced colonic hyperplasia (Fig. 7A and B) as well as neutrophil staining with Ly-6G antibodies (Fig. 7C). These results reveal a novel in vivo role for the Tir tyrosine residues TirY451 and Y471, which are also implicated in actin polymerization during C. rodentium infection (Fig. 4).