Differential adherence patterns were observed on the RAJ-IVOC at the 10⁷ CFU inoculum unlike with RSE cells only

All three test strains, inoculated at a bacteria:cell ratio of 10:1, demonstrated the same diffuse, moderate adherence pattern on RSE cells with no significant quantitative differences between strains (70 – 76% cells having 1–10 bacteria/cell; p = 0.8 -0.7), irrespective of the presence or absence of slp (Fig 1). This pattern was observed on the squamous epithelial cells of the RAJ-IVOC tissue as well, however, a distinct increased adherence to the columnar region of the same tissue was observed that appeared to be linked to the presence/absence of slp (Fig 2).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Quantitative (Panel A) and qualitative (Panel B) data from the RSE cell adherence assay.

The quantitative data from three adherence assays are shown in Panel A. The ‘diffuse, moderate’ adherence patterns of the test strains on the RSE cells are shown in Panel B. The immunofluorescent images were captured at 400x magnification with the 100 µm scale bar. The bacteria (STEC O157) as indicated by arrows, the RSE cells’ cytokeratins, and the nuclei have green, orange-red and blue fluorescence, respectively.

https://doi.org/10.1371/journal.ppat.1013584.g001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Immunofluorescent RAJ-IVOC images and graph depicting STEC O157 adherence and viable counts of recovered bacteria, respectively, post inoculation with 10⁷ CFU test strains.

(A) Immunofluorescent images of tissue sections from RAJ-IVOCs used to set up the adherence assay with 10⁷ CFU inoculum concentration for all the test strains. Post-assay, tissue sections were stained with immunofluorescent antibodies targeting the RAJ cells’ cytokeratin and STEC O157, and images were recorded at 100x magnification. The adherent bacteria (shown with arrows), RAJ cells’ cytokeratin, and the nuclei have green, orange-red and blue fluorescence, respectively. The squamous (S), junction (J) and columnar (C) regions of the RAJ are indicated, along with a 100 µm scale bar. (B) Cumulative graph from four assays showing viable counts (CFU/ml) of test strains recovered from RAJ-IVOC tissues by bacterial culture. The red dotted line on the graph marks the STEC O157 detection limit of 10 CFU/ml for non-enrichment cultures.

https://doi.org/10.1371/journal.ppat.1013584.g002

For the RAJ-IVOC (S1 Fig), two different concentrations of the test strains were used in the inoculum, 10⁷ and 10⁸ CFU in 2 ml media, to determine the optimal inoculation dose (Figs 2 and S2). The lower inoculum concentration allowed for better clarity in bacterial adherence phenotype and hence, the 10⁷ CFU in 2 ml media inoculum was used in all subsequent six RAJ-IVOC assays. Immunofluorescent imaging of tissue sections of the frozen RAJ-IVOCs indicated the differential adherence pattern of the test strains, with EDL932-WT and EDL932 Δslp-p:slp adhering in greater numbers along the columnar epithelia especially along the FAE cells compared to EDL932 Δslp (Fig 2). The adherence phenotype was primarily diffuse along the squamous cells but diffuse with several aggregates on the columnar epithelia of the RAJ-IVOC (Fig 2). Increased focal disruption of epithelia, especially around the junction and the columnar epithelia, could be associated only with the presence of STEC O157 (Figs 2, S2 and S6, and S1 Data).

Post-incubation, culture of both the leftover inoculum and the tissue itself enabled the recovery of the test strains from the respective RAJ-IVOCs. The viable counts from the residual inoculum ranged from 10³ to 10⁷ CFU/ml for each test strain (S1 Table). On the other hand, there was a reduction in the viable counts for the test strains recovered from the tissues, suggesting loss of non-adhered bacteria to the washes prior to processing tissues for culture. On an average, about 2- to 3- log decrease in viable counts was observed with the EDL932-WT and EDL932 Δslp-p:slp strains, and a 4- to 5- log reduction in the EDL932 Δslp viable counts (Figs 2 and S3 and S1 Table) when the 10⁷ CFU inoculum was used. This could reflect the poor adherence by the EDL932 Δslp strain, as observed in the immunofluorescent images, although the differences in counts were not statistically significant (p = 0.5 - 0.9). These differences in the recovery of the tissue adherent-bacteria by culture were not discernable with the 10⁸ CFU inoculum where a reduction by 2- to 3- logs was observed in all viable counts irrespective of the strain (S1 Table and S2 and S3 Figs).

No STEC O157 were detected in the no bacteria controls set up for the RSE cell- and RAJ-IVOC -adherence assays eliminating the possibility of any pre-existing or cross-contaminating STEC O157. Occasionally, a few pinpoint background colonies were obtained on culture plates that were ruled out to be STEC or E. coli based on the phenotype and serological testing. All recovered bacteria produced sorbitol non-fermenting, colorless, MUG-non-utilizing, non-fluorescent colonies typical of STEC O157 and readily agglutinated with the O157 latex agglutination reagent. Additional PCR verified the genotype of the recovered test strains. The distinctive polymorphic amplified typing sequence (PATS) profiles for the STEC O157 strain EDL932 helped confirm the relatedness of the wild type, mutant, and complemented strains used in this study (Table A in S2 Table and S4 Fig). In addition, the PCR results with the vector and slp -gene primers distinguished the EDL932 Δslp and EDL932 Δslp-p:slp strains from the parent EDL932-WT (Table B in S2 Table and S4 Fig).

The RAJ-IVOC tissue integrity was disrupted primarily by the STEC O157 inoculum

The RedDot2 nuclear stain confirmed the viability of cells within the RAJ-IVOC with the absence of nuclear staining in the un-fixed RAJ-IVOC tissues pre- and post- 3 h incubation (S5 Fig). Additional histopathological evaluations linked epithelial disruption in inoculated IVOCs primarily to the presence of STEC O157, as previously reported [42]. The squamous and columnar epithelial cell regions along with the junction were clearly visible in all H&E-stained RAJ-IVOC tissue sections along with lymphoid follicles underlying some of the columnar epithelia (FAE cells); the latter region being referred to as the glandular or mucosal region in some of the pathology reports (S1 Data). Some animals had antemortem anomalies and/or disease contributing to histopathology in parts of the RAJ limiting extensive use of the tissue in setting up IVOCs (S1 Data). However, as shown in the representative histopathological report (S6 Fig), overall, the glandular and squamous regions were normal/unremarkable in the tissue sections used to set up the no bacteria- control RAJ-IVOC (S6 Fig). Additional epithelial disruptions were observed largely in the glandular region of RAJ-IVOC exposed to EDL932-WT, EDL932 Δslp, or EDL932 Δslp-p:slp, respectively (S6 Fig and S1 Data).

pIgR localizes primarily along the columnar epithelial region of the RAJ-IVOC with likely suppression after Slp binding

Immunofluorescent staining of the RAJ-IVOC tissue sections for pIgR indicated a predominant distribution of this receptor in the columnar epithelial region of the RAJ with minimal-to-no pIgR along the squamous epithelia (Fig 3). pIgR could be detected both within and on the luminal surface of the columnar epithelia (Fig 3). Hence, the increased adherence of slp-expressing EDL932-WT and EDL932 Δslp-p:slp to the columnar epithelia (Figs 2 and S2) appeared to also correlate with the increased expression of host cell pIgR (Fig 3) in that region. This biased distribution of pIgR on the RAJ-IVOC tissues was also observed by RNAScope in situ hybridization (ISH); distinct spatial expression of the pIgR mRNA was detected primarily in the cytoplasm of columnar cells in all the RAJ-IVOC tissue sections evaluated (S3 Table and S2 Data).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Representative immunofluorescent images of tissue sections from the RAJ-IVOC adherence assay depicting pIgR distribution.

(A) RAJ-IVOC were inoculated with EDL932-WT (10⁷ CFU inoculum) and post-assay tissue sections were then stained with immunofluorescent antibodies targeting pIgR, STEC O157/or RAJ cells’ cytokeratins, and images recorded at 200x magnification. The pIgR (red), adherent bacteria (green; shown with arrows) or RAJ cells’ cytokeratins (green), and the nuclei (blue) are shown. The squamous (S), junction (J) and columnar (C) regions of the RAJ are indicated, along with a 100 µm scale bar. A ‘no primary antibody’ control was included to demonstrate the specificity of the antibodies targeting pIgR. (B) This image depicts the regions around the junction with the predominant distribution of pIgR in the columnar epithelium of the RAJ tissue. (C) The additional image depicts possible interaction of STEC O157 with pIgR in the columnar epithelium of the RAJ tissue.

https://doi.org/10.1371/journal.ppat.1013584.g003

In all the RNAScope ISH assays, the positive control probe targeting the cyclophilin mRNA generated robust signals verifying RNA integrity and the negative control dapB mRNA-targeting probe generated no signal indicating lack of any cross-reacting RNA in the background (S2 Data). In the trial run, an unexpected low ‘average red (pIgR) copies per cell’ was obtained for the single columnar epithelial (SCE) region of the RAJ-IVOC tissue exposed to EDL932 WT (#3: 0.538965) compared to the higher scores for tissues exposed to EDL932 Δslp (#4: 4.946465) or not exposed to any inoculum (Pre-assay control T0: 1.281183 and no bacteria control #1: 2.491031) (S3 Table -Trial Run). Similar output was obtained in the test run with a different set of RAJ-IVOC tissues. For the test run, the ‘average red (pIgR) copies per cell’ scores for the SCE regions were: T0: 3.409, #1: 2.374, #3: 1.760, #4: 4.169, #5: 1.853 (S3 Table-Test Results). Such distinct differences were not observed in the stratified squamous epithelial (SSE) regions. Overall, these results provide additional evidence of the RAJ-IVOC tissue viability and most importantly suggest a possible suppression of pIgR in presence of the interacting ligand, Slp. Further studies are needed to explore this observation as it may provide insights into the reduced mucosal (IgA) responses in STEC colonized or STEC-vaccinated cattle [32,43,44].

Protein-protein docking predictions support pIgR-Slp interaction

A total of 4 and 10 possible 3D-docking models were generated in NovaDock, depicting the human pIgR-E. coli Slp and bovine pIgR-E.coli Slp interactions, respectively. Model energy, cluster size and cluster energy was considered when selecting an optimal model (Fig 4 and S3 Data); a low energy and high cluster size is usually indicative of low energy and biologically relevant conformation (https://www.dnastar.com; [45–47]). Details of the selected models for each of the predicted receptor-ligand combinations are in the supplementary document with the additional model information, providing a list of residues involved in intermolecular contacts at the binding interface (S4 Data). The parameters of the best scoring docking model for the human pIgR-E. coli Slp was, model energy: -18.30, cluster size: 2, cluster energy: -12.01 ± 6.29 with 40 residue contacts (S4 Data). Likewise, the parameters for the optimal bovine pIgR-E. coli Slp docking model was, model energy: -34.95, cluster size: 1, cluster energy: -34.951 ± 0 with 41 residue contacts (S4 Data). Although the cluster size was limited, the low energy of the conformations support optimal interactions between ligand and receptor. Interestingly, BLASTp analysis indicated a homology of only 67% between the human and bovine pIgR sequences (S4 Data), which may have contributed to differences in the Slp-docking models and interacting residues. Overall, as expected, fewer residues were involved in pIgR-Slp docking compared to the more traditional pIgR-dimeric sIgA complex (Fig 4 and S3 Data).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Representative computational models for individual and interacting proteins.

(A) Ligand E. coli Slp (AlphaFold), (B) Receptor, Human pIgR (AlphaFold), (C) Receptor, Bovine pIGR (AlphaFold), (D) Human pIgR-dimeric sIgA complex (RCSB), (E) NovaDock predicted model: Human pIgR-E. coli Slp complex, and (F) NovaDock predicted model: Bovine pIgR-E. coli Slp complex. Color style used was: Yellow-Slp; Green-pIgR; White-dimeric sIgA; Red-J chain.

https://doi.org/10.1371/journal.ppat.1013584.g004

Ethics statement

Bovine RAJ tissues for the IVOCs were collected at necropsies of animals, per the NADC Institutional Animal Care and Use Committee guidelines. No live animals were included in this study making it exempt from a full committee review.

Bacterial strains

The following strains from our previous study [25] were evaluated: (i) STEC O157 strain EDL932 (ATCC 43894: stx₁⁺, stx₂⁺, eaeA⁺, hlyA⁺, slp⁺; American Type Culture Collection/ATCC, Manassas, VA) referred to as EDL932-WT, (ii) STEC O157 Δslp (stx₁⁺, stx₂⁺, eaeA⁺, hlyA⁺, slp^-) referred to as EDL932 Δslp and (iii) STEC O157 Δslp + pUC18::slp (stx₁⁺, stx₂⁺, eaeA⁺, hlyA⁺, slp⁺, ampicillin resistant) referred to as EDL932 Δslp-p:slp.

Bacterial inoculum preparation

All bacteria were grown overnight in Dulbecco’s modified Eagle’s medium (DMEM) with low glucose (DMEM-LG; Invitrogen, Carlsbad, CA), with or without ampicillin (100 μg/ml), at 37°C without aeration. As described previously, the overnight cultures were washed and re-suspended in DMEM with no glucose (DMEM-NG; Invitrogen) prior to testing in the adherence assays [14,42,69,70].

RSE adherence assay

The RSE assay was done in three biological replicates with eight technical replicates per strain in each assay, as previously described [14–16,18,19,42,70,82]. Briefly, RSE cells were suspended in DMEM-NG to a final concentration of 10⁵ cells/ml. Each bacterial isolate was mixed with RSE cells at a bacteria:cell ratio of 10:1. The mixture was incubated at 37°C with aeration (110 rpm) for 4 h, pelleted, washed, and reconstituted in 100 µl of double-distilled water (dH2O). Drops of the suspension (2 µl) were placed on Polysine slides (Thermo Scientific/Pierce, Rockford, IL), dried, fixed, and stained with fluorescence-tagged antibodies specific to the O157-antigen and cytokeratins of the RSE cells as before [14–16,18,19,25,70,82]. Adherence patterns on RSE cells were qualitatively recorded as diffuse, aggregative (in clusters), or nonadherent, and quantitatively as percentages of RSE cells with or without adhering bacteria [15]; a strongly adherent phenotype requires more than 50% of RSE cells with 10 adherent bacteria per cell, moderately adherent is when 50% or less of the RSE cells have 5–10 adherent bacteria per cell, and nonadherent when less than 50% of the RSE cells have only 1–5 adherent bacteria. RSE cells with no added bacteria were included as negative controls to confirm the absence of pre-existing STEC O157 bacteria. Quantitative data were evaluated for statistical significance using one-way ANOVA with Dunnett’s test and the unpaired t-test; p < 0.05 was considered significant (GraphPad Prism version 8.0.0, GraphPad Software, San Diego, CA).

RAJ-IVOC adherence assay

Six RAJ-IVOC adherence assays were conducted, each in duplicate when sufficient tissue was available, with the test strains and a ‘no bacteria’ control. Bovine RAJ tissues for the IVOCs were collected at necropsies of animals that were part of other unrelated studies at the National Animal Disease Center (NADC, Ames, IA) or from a local meat locker (S1 Table). Animals included Holstein, Angus, or Aberdeen cows or steers, 1–9 years of age, fed an alfalfa hay-based maintenance or corn-based finishing diet with ad libitum access to water (S1 Table). The tissues were transported in DMEM-NG (Invitrogen) supplemented with 2.5% fetal bovine serum (FBS; Thermo Scientific HyClone, Logan, UT), 100 µg/ml streptomycin, 100 U/ml penicillin (Pen-Strep; Invitrogen), 2.5 mg/L amphotericin B (Sigma), and 50 µg/ml gentamicin (Invitrogen), on ice, and processed in the laboratory to either harvest RSE cells or to set up the RAJ-IVOCs as previously described [14,42].

Briefly, the RAJ tissue was cut into multiple 4 cm x 2 cm (length x width) rectangular pieces, with the length encompassing 2 cm of each region on either side of the RAJ. The tissue pieces were carefully placed on top of a stack prepared in a flat-bottomed polystyrene clear tissue culture dish (Corning/Costar, Sigma-Aldrich Corp., St. Louis, Mo.) comprising of a sterile dental wax disc (Polysciences, Inc., Warrington, PA), a 1–2 mm thick sterile sponge soaked in DMEM-HG (Invitrogen) with 10% FBS and a sterile Whatman filter disc (Grade 1, 32 mm; Sigma) as shown in S1 Fig. The cut tissue piece was secured with sterile pins along the edges, with the mucosal/luminal surface facing outwards. Sterile DMEM-NG with 3% agarose was used to seal gaps around the edges of the tissue to prevent spillage of the DMEM-HG media below onto the exposed surface of the tissue and to contain the bacterial inoculum on the exposed luminal surface of the RAJ-IVOC (S1 Fig). Given the estimated total of 10⁴ cells on the exposed surface area of the RAJ-IVOC, bacterial inoculum of 10⁸ CFU bacteria (10,000. bacteria:1 cell ratio) and 10⁷ CFU bacteria (1000 bacteria:1 cell ratio) in DMEM-NG (2 ml total volume) were evaluated in comparative assays to determine the optimal bacterial concentration producing distinct adherence phenotypes [42]. For no bacteria controls, 2 ml DMEM-NG without bacteria was added.

Post-inoculation, all RAJ-IVOC containing dishes were incubated at 39^oC with 5% CO₂ and gentle shaking at 100–110 rpm for 3 hours. Following incubation, the inoculum or media left on the exposed tissue surface was aspirated and plated on sorbitol MacConkey agar (BD Biosciences) containing 4-methylumbelliferyl-β-d-glucuronide (MUG, 100 mg/liter; Sigma) (SMAC-MUG) and MacConkey agar (BD Biosciences) containing MUG (100 mg/liter; Sigma) (MAC-MUG) to isolate STEC O157, and other background bacteria if any, respectively [42]. Each RAJ-IVOC tissue was then gently disengaged from the agarose, rinsed, and one half, weighing about 1–2 g, was frozen in 10 ml LB with 30% glycerol (LB-glycerol) for subsequent bacterial culture. The second half was flash frozen in Optimal Cutting Temperature solution (OCT; Tissue-Tek, Sakura Finetek, Torrance, CA) for sectioning and staining with hematoxylin and eosin (H&E) dyes at the Microscopy Services, NADC for histopathological evaluations, or with antibodies tagged with fluorescent tags to study the adherence of the inoculated bacteria [42]. Adherence patterns were qualitatively recorded as diffuse, aggregative (microcolonies), or nonadherent. The H&E-stained slides were analyzed using the Aperio Digital Pathology system (Leica Biosystems, Deer Park IL) and the corresponding Aperio Image Scope software (Leica).

RAJ-IVOC tissue viability test

The viability of the RAJ-IVOC tissues was tested as described previously [42], using two uninoculated and unfixed tissue samples per assay, one pre-incubation, and the other post-incubation. The RedDot2 nuclear staining dye (RedDot2 Far-Red Nuclear Stain, Biotium, Inc., Fremont, CA) was used per the manufacturer’s instructions. Briefly, each tissue sample was soaked in RedDot2 reagent (Biotium) diluted in DMEM-NG, for 30 min at room temperature, before rinsing and flash freezing in OCT. The stained frozen tissues sections, on Colorfrost slides (Thermo Fisher Scientific, Pittsburgh, PA) were air dried and cover slipped with Prolong Glass anti-fade reagent (Invitrogen) before visualization by fluorescent microscopy.

RAJ-IVOC tissue culture for bacteria

The ~ 1–2 g of RAJ-IVOC tissue frozen in LB-glycerol was thawed and cultured for STEC O157 as previously described [42,83–85]. Briefly, the tissue was minced and suspended in 25–50 ml Trypticase soy broth (BD Bioscience, San Jose, CA), supplemented with cefixime (50 μg/L; U.S. Pharmacopeia. Washington, D.C.), potassium tellurite (2.5 mg/L; Sigma), and vancomycin (40 mg/L; Alfa Aesar, Haverhill, MA) (TSB-CTV), and the suspension was serially diluted with sterile saline (0.15 M NaCl) both before and after overnight incubation at 37°C with aeration. The pre-incubation dilutions were spread plated on SMAC-MUG (non-enrichment culture) and post-incubation suspension dilutions were plated on SMAC-MUG supplemented with cefixime (50 μg/L), potassium tellurite (2.5 mg/L) and vancomycin (40 mg/L) (SMAC-CTMV; selective-enrichment culture). All plates were read after overnight incubation at 37^oC and colonies that did not ferment sorbitol or utilize MUG (non-fluorescent under UV light) were confirmed to be STEC O157 by latex agglutination tests (E. coli O157 latex, Oxoid Diagnostic Reagents, Oxoid Ltd., Hampshire, UK). The suspensions were additionally plated on MAC-MUG for increased recovery of the lactose-fermenting, MUG-utilizing (fluorescent under UV light) background non-STEC bacteria, if any. Bacteria recovered from the RAJ-IVOC tissue cultures were verified by PCR, as described below. Quantitative data from the comparative RAJ-IVOC tissue cultures were evaluated for statistical significance if any differences in adherence were observed using the unpaired t-test or one-way ANOVA with Dunnett’s test; p < 0.05 was considered significant (GraphPad Prism).

PCR verification of strains

(i) Polymorphic amplified typing sequence (PATS) was used to DNA-fingerprint bacterial isolates, pre- and post-assays, as described previously [86–89] to confirm derivation from the STEC O157 strain EDL932. Specifically, primer pairs targeting 8 polymorphic XbaI-, 7 polymorphic AvrII- restriction enzyme sites, and 4 virulence genes encoding Shiga toxins 1 and 2 (stx₁ and stx₂), intimin-γ (eaeA), and hemolysin-A (hlyA), were used to generate amplicons from colony lysates [19,86,87,89,90]. Purified (QIAquick PCR purification kit, Qiagen, Valencia, CA) PCR reactions amplifying the AvrII- restriction enzyme sites were digested with the AvrII restriction enzyme (New England Biolabs, Beverly, MA) to confirm the presence of the restriction site. Following 3% agarose gel-electrophoresis, the presence or absence of amplicons for XbaI sites and the virulence genes was recorded using “1” and “0”, respectively. For the AvrII site, absence of an amplicon was recorded as “0”, while the presence of the restriction site with a single nucleotide polymorphism was scored as “1”, an intact restriction site as “2”, and a restriction site duplication as “3” [86,87,90]. (ii) Rapid differentiation of the wild-type, mutant and complemented strains was done using primers targeting the vector pUC18, and the slp and stx₂ genes that were designed using the ‘PCR Primer Design’ tool in DNAStar (DNAStar Navigator Version 17.4.1.17; https://www.dnastar.com): (i) pUC18 primers: Forward 5′-TCGCGCGTTTCGGTGATGA-3′ and Reverse 5′-ACGAAAGGGCCTCGTGATACG-3′; amplicon size, 2685 bp, (ii) slp primers: Forward 5′-ATGAACATGACAAAAGGTGC-3′ and Reverse 5′-TTATTTGACCAGCTCAGGTGTTA-3′; amplicon size 567 bp, and (iii) stx2 primers used for PATS profiling [86–89]: Forward 5′-GGCACTGTCTGAAACTGCTCC-3′ and Reverse 5′-TCGCCAGTTATCTGACATTCTG-3′; amplicon size 255 bp.

Immunofluorescent staining of RAJ-IVOC tissue sections

(i) For test bacteria and host cells. As previously described [14,42,91], the tissue sections were ethanol-fixed, blocked with 5% normal goat serum, and incubated with primary and secondary antibodies, each at RT for 1 h, targeting the bacteria or host cell cytokeratin. Primary antibodies included the mouse anti-(PAN) cytokeratins (AbD Serotec, Raleigh, NC) targeting the RAJ cell cytokeratins. Secondary antibodies included the Alexa Fluor 594 (red)–labelled goat anti-mouse IgG (H + L; F(ab^’)2fragment) (Invitrogen) targeting the anti-cytokeratins primary antibody and the fluorescein isothiocyanate (FITC; green)–labelled goat anti-O157 (KPL, Gaithersburg, MD) antibodies targeting O157. Air-dried slides were then cover-slipped with Prolong Gold anti-fade reagent containing the DNA stain 4^’,6-diamidino-2-phenylindole (DAPI (blue); Invitrogen). Immunofluorescent images from the stained slides were captured using the Nikon Eclipse E800 fluorescence microscope (Nikon Instruments Inc., Elgin, IL) [14,91]. Control slides with sections from uninoculated RAJ-IVOC were stained similarly to rule out nonspecific binding [14,42,91]. Additional slides were stained with only secondary antibodies to verify the specificity of the primary antibody and/or with FITC-tagged antibodies targeting unrelated Salmonella bacteria to demonstrate specificity of antibodies used, as needed. (ii) For pIgR, EDL932-WT and host cells: The tissue sections were fixed and blocked as before and stained with different combinations of primary and secondary antibodies. The rabbit anti-pIgR IgG (Thermo Scientific Pierce) targeting pIgR was the primary antibody. Secondary antibodies included the Alexa Fluor 594 (red)–labelled goat anti-rabbit IgG (H + L; F(ab^’)2fragment) (Invitrogen) targeting the anti-pIgR primary antibody and the fluorescein isothiocyanate (FITC; green)–labelled goat anti-O157 (KPL, Gaithersburg, MD) antibody targeting STEC O157. In addition, host cell cytokeratins were stained, when needed, with mouse anti-(PAN) cytokeratins (AbD Serotec, Raleigh, NC) primary antibody targeting the RAJ cell cytokeratins and the Alexa Fluor 488 (green) labelled goat anti-mouse IgG (H + L; F(ab^’)2fragment) (Invitrogen) targeting the anti-cytokeratins primary antibody. Control slides were processed, and all stained slides were air-dried, cover-slipped and imaged as described above.

RNAscope in situ hybridization (ISH) screening for pIgR-mRNA

(i) Tissue samples. The RNAscope ISH assay [92] was set up to screen the RAJ-IVOC tissues for the distribution and concentration of pIgR-mRNA. After a trial run with primarily the columnar epithelial region of the RAJ, RAJ-IVOC assays were set up with the following treatments: #T0: uninoculated, unincubated, pre-assay, #1: uninoculated, incubated, no bacteria control, or #3: with EDL932-WT, or #4: EDL932 Δslp, or #5: EDL932 Δslp-p:slp. All the RAJ-IVOC tissues were incubated for 3h, as described above, except for the #T0 IVOC which represented the pre-assay tissue. Each IVOC tissue sample, including #T0, was fixed overnight in neutral buffered formalin (NBF) before moving the formalin-fixed tissues into 70%ethanol. The formalin-fixed tissues were then embedded in paraffin for sectioning and use in the RNAscope ISH assay. (ii)Assay. RNAscope ISH assay was performed using the RNAscope 2.5 HD Reagent Kit RED (Cat. No. 322750, Advanced Cell Diagnostics (ACD), Newark, CA) with the Leica Biosystems BOND RX automated IHC/ISH slide staining system (Leica), according to the manufacturer’s instructions. The formalin-fixed paraffin-embedded tissue sections (4–5 μm) were mounted on Superfrost Plus slides, baked at 60°C for 1 h, and subjected to automated deparaffinization with BOND Dewax Solution and graded alcohol rinses, followed by rehydration in BOND Wash Solution. Heat-induced epitope retrieval was carried out at 88°C using BOND ER Solution 2, after which sections were treated with RNAscope 2.5 LSx Protease and blocked with RNAscope 2.5 LSx H₂O₂. Hybridization was performed at 42 °C for ~2 h with target-specific RNAscope 2.5 LS Probes including: (i) Test, Bt-pIgR-C1 (test; Cat No. 1268218-C1: 20 pairs complimentary to 354–1269 bp of the target pIgR mRNA-accession number NM_174143.1), (ii) Negative Control, dapB (Cat No. 312038: 10 pairs complimentary to 414–862 bp of the non-specific Bacillus subtilis strain SMY dapB mRNA-accession number, EF191515) and (iii) Positive control, Bt-PPIB (Cat No. 319458: 13 pairs complimentary to 24–788 bp of the host cell specific, Bos taurus cyclophilin B mRNA-accession number, NM_174152.2). Signal amplification was achieved through the sequential application of RNAscope AMP 1–6 RED reagents with intervening BOND Wash Solution rinses, and chromogenic detection was performed using the Mixed Red Refine reagent. Slides were counterstained with RNAscope Hematoxylin, blued with RNAscope Bluing reagent, rinsed in deionized water, and hydrated prior to coverslip mounting. All reagents were automatically dispensed in 150 µL volumes, and incubation times followed the ACD RNAscope 2.5 LSx standard protocol, as verified in the Leica BOND RX run setup log. (iii) Digital image analysis. Histologic sections were scanned at 400X magnification using an Aperio Versa 200 scanner. Digital image analysis was performed using HALO Image Analysis Platform version v3.6 and the in-situ hybridization module, ISH v3.4.3.0 (Indica Labs, Inc., Albuquerque, NM). To select the ROI for analysis, the RAJ was located followed by the selection of the SCE to one side and the SSE to the other side. The annotation was performed to ensure that both SCE and SSE had approximately the same length. For each sample, the ISH algorithm was first trained with representative nuclear and pIgR-ISH staining and then applied to the complete selected area. The following variables were obtained for SCE and SSE: total red (pIgR) copies, average red (pIgR) copies per cell, and percentage of pIgR positive cells. The ACD semi-quantitative scoring was also used to characterize expression levels, which ranges from 0 (no expression) to 4 (high expression, > 15 dots/cell) (S3 Table). The H-score reflects both the number of dots per cell and the percentage of cells displaying each specific dot count (S3 Table).

Protein modeling

The AlphaFold protein structure database (https://alphafold.ebi.ac.uk; [93,94] and the Research Collaboratory for Structural Bioinformatics (RCSB) protein databank (https://www.rcsb.org; [95]) were used to access existing predicted structures or interactive models as needed. In addition, de novo protein-protein docking analysis was done using NovaDock in DNASTAR Protean 3D-Version 17.4.3 (2) (DNASTAR, Inc. Madison WI; https://www.dnastar.com). NovaDock employs the “SwarmDock” algorithm for computational modeling and makes docking predictions based on energy calculations and protein flexibility [45–47], generating 3D high-resolution models of the proposed receptor-ligand complexes. Protein structures for docking were selected from the AlphaFold database and included: (i) the ligand, E. coli Slp (UniProt: P37194; PDB ID: AF-P37194-F1-v4; DOI: https://alphafold.ebi.ac.uk/entry/P37194), (ii) the receptor, bovine pIgR (UniProt: P81265; PDB ID: AF-P81265-F1-v4; DOI: https://alphafold.ebi.ac.uk/entry/P81265), and (iii) the receptor, human pIgR (UniProt: P01833; PDB ID: AF-P01833-F1-v4; DOI: https://alphafold.ebi.ac.uk/entry/P01833) (S3 Data). The human pIgR-dimeric sIgA complex (PDB ID: 6UE7; DOI citation: https://doi.org/10.2210/pdb6UE7/pdb; [96]) was acquired from the RCSB protein database for comparison with the de novo docking models developed for human pIgR and E. coli Slp, and bovine pIgR and E. coli Slp proteins, using NovaDock. In addition, protein-protein alignments were done using BLASTp (https://blast.ncbi.nlm.nih.gov/) to verify sequence overlaps, if any.