IL-17 receptor A functions to help maintain barrier integrity and limit activation of immunopathogenic response to H. pylori infection

ABSTRACT

Activation of Th17 cell responses, including the production of IL-17A and IL-21, contributes to host defense and inflammatory responses by coordinating adaptive and innate immune responses. IL-17A and IL-17F signal through a multimeric receptor, which includes the IL-17 receptor A (IL-17RA) subunit and the IL-17RC subunit. IL-17RA is expressed by many cell types, and data from previous studies suggest that loss of IL-17 receptor is required to limit immunopathology in the Helicobacter pylori model of infection. Here, an Il17ra^-/- mouse was generated on the FVB/n background, and the role of IL-17 signaling in the maintenance of barrier responses to H. pylori was investigated. Generating the Il17ra^-/- on the FVB/n background allowed for the examination of responses in the paragastric lymph node and will allow for future investigation into carcinogenesis. While uninfected Il17ra^-/- mice do not develop spontaneous gastritis following H. pylori infection, Il17ra^-/- mice develop severe gastric inflammation accompanied by lymphoid follicle production and exacerbated production of Th17 cytokines. Increased inflammation in the tissue, increased IgA levels in the lumen, and reduced production of Muc5ac in the corpus correlate with increased H. pylori-induced paragastric lymph node activation. These data suggest that the cross talk between immune cells and epithelial cells regulates mucin production, IgA production, and translocation, impacting the integrity of the gastric mucosa and therefore activating of the adaptive immune response.

INTRODUCTION

The gastric environment, while exposed to numerous environmental factors, food, microorganisms, or viruses, is also considered inhospitable to many invading pathogens. The lumen of the stomach is relatively acidic and contains enzymes including acidic chyme, lingual lipase, pepsin, and rennin, which aid in chemical digestion (1). Furthermore, the stomach is responsible for the production of intrinsic factor, which is vital for the absorption of vitamin B₁₂. These digestive factors could be damaging to the epithelial cells of the stomach, but there is also a layer of bicarbonate-rich mucus that neutralizes the acid and contains several proteins with antimicrobial functions (2, 3). These barrier proteins include immunoglobulin A (IgA), β-defensins, lipocalin, and likely several others (4, 5). Typically, the maintenance of these physical and chemical barriers is facilitated by specialized epithelial cells through tight junctions and the production of many barrier proteins, resulting in healthy gastric tissue (6). When bacterial colonization occurs or the stomach is exposed to toxins, toxicants, drugs, or dietary stress, there can be a disturbance of this homeostasis (7).

Helicobacter pylori has several virulence factors that facilitate its colonization of the gastric mucosa (8). These bacterial factors allow for local neutralization of the acidic environment, penetration of the mucus layer (urease and flagella), activation of inflammation (lipoproteins), disruption of metabolism (vacuolating toxin impacts mitochondria function), and even changes to the tight junctions through beta-catenin activation (Cag type 4 secretion system activity) (9). Epithelial cell turnover is required to return the tissue to homeostasis after infection or damage, but chronic H. pylori colonization and activation of the inflammatory response lead to changes in gastric stem cell populations over time and documented loss of parietal cells (6, 10). The pathological outcomes of this complex long-term pathogen-host interaction can lead to gastric ulcers, atrophic gastritis, dysplasia, intestinal metaplasia, or even gastric adenocarcinoma (11). H. pylori infection remains the number one risk factor for the development of gastric cancer (12, 13).

Inflammation is known to play a role in the development of many cancers including gastric cancer (14, 15). A major component of the chronic inflammatory response to H. pylori infection is T helper 17 cells (Th17) and their cytokines, namely interleukin 21 (IL-21) and interleukin 17 family members (IL-17a, IL-17f, or IL-17a/f) (16–18). Th17 responses are key for the control of extracellular pathogens including extracellular bacteria and pathogenic yeasts (19, 20). In the absence of IL-17 responses, neutrophil recruitment is dampened, and antimicrobial responses are blunted at mucosal sites including the gastrointestinal tract, skin, and lungs. IL-17 responses are also required for neutrophil recruitment in response to H. pylori infection (21). Neutrophil recruitment is reduced when IL-17RA receptor signaling is absent, but it does not always result in a change in bacterial burden. This suggests that other factors beyond the presence of neutrophils may impact H. pylori colonization levels including changes to metabolite availability, mucous composition, antimicrobial production, and/or barrier function (21, 22).

Surprisingly, Il17ra^-/- mice (on the C57Bl/6 background) colonized with H. pylori have increased lymphocytic infiltration compared to wild-type (WT) C57Bl/6 mice by 3 months post-infection (21, 22). The increased lymphocytic infiltration correlates with the increased expression of the Il21 gene and a significant increase in B cells, which organize to form lymphoid follicles in the gastric mucosa. These previously published studies indicate that IL-17RA signaling must be required to control chronic inflammation, but the mechanism by which IL-17 limits chronic activation of Th17 responses and limits formation of lymphoid follicles is not clear.

For this research, we generated the Il17ra^-/- on the FVB/n background to allow for the examination of immune cell activation patterns in the paragastric lymph node (PLN). Herein, evidence is presented, which suggests that in the absence of IL-17 receptor, the integrity of the gastric barrier is disrupted. During chronic H. pylori infection, IL-17RA deficiency led to reduced Pigr and Nox1 expression, reduced mucin-producing cells, increased parietal and chief cell atrophy, and increased Ceacam1 expression. Meanwhile, there is an increased activation of the adaptive immune responses suggesting increased antigen uptake and presentation, which may be due to the loss of barrier integrity.

MATERIALS AND METHODS

Mice

IL-17RA-deficient mice were generated using CRISPR/Cas9 methodology in FVB/n mice. The Vanderbilt Transgenic Mouse/ES Cell Shared Resource used cytoplasmic zygote microinjections of IL-17RA Exon 3 (MGI 107399) targeted CRISPR/Cas9 ribonucleoproteins to produce an Il17ra deletion frameshift mutation leading to a mutant allele. The 82-bp deletion allele was selected for further analysis. Upon weaning of the N1 generation, the animals were tailed, and the genomic DNA was PCR amplified. The PCR product was sent for Sanger sequencing to confirm expected mutation (see Fig. S1). To genotype the mice in subsequent generations, genomic DNA from the mouse tail was analyzed using standard PCR followed by gel electrophoresis. Il17ra^em1(hmsa) mice were identified by a band located at 500 bp using primer set 5′- CCTTCTCCCCAAACATTCCT-3′ and 5′-CCACTTGCCTTTTTCTTCCTGTG-3′, forward and reverse primers, respectively (and lack of a 582-bp band). FVB/n mice for controls were bred in the same room. The mice were Helicobacter-free prior to infection. Feces from sentinel mice housed in the same rooms are consistently tested negative for pinworms, mouse parvovirus, and several other murine pathogens. Mice were transferred to an ABSL2 facility prior to infections. Mice were euthanized at predetermined time points post-infection (including uninfected mice) and tissues were collected for analysis.

H. pylori infection and colonization

At age 8–10 weeks, mice were orogastrically inoculated with a suspension of H. pylori strain PMSS1 at 1 × 10⁹ CFU in 0.5 mL of Brucella broth. H. pylori was grown from 10 µL of freezer stock on 5% sheep blood tryptic soy agar (TSA) Plates (ThermoFisher Scientific, Waltham, MA, USA). Cultures were passed every 48 hours onto new plates until there was sufficient H. pylori to inoculate liquid cultures (typically two to three passages). Two plates were used to inoculate each 50 mL flask of Brucella broth supplemented with 10% fetal bovine serum (FBS) (R&D Systems, Minneapolis, MN, USA) and 10 µg/mL vancomycin. Liquid cultures were grown for 18 hours in a BD GasPak anaerobic chamber with EZ sachets per the manufacturer’s recommendations (Becton Dickinson, Franklin Lakes, NJ, USA) at 160 rpm on Maxq2000 orbital shaker (ThermoFisher Scientific, Waltham, MA, USA) at 37°C and 5% CO₂. Samples were quantified by OD₆₀₀ on BioTek ELx808 plate reader (BioTek, Winooski, VT, USA) utilizing Gen5 3.10 software (BioTek, Winooski, VT, USA). A blank well of 10% FBS in Brucella broth was used to normalize background measurements. Cultures were normalized to 1 × 10⁹ CFU in Brucella broth and each dose was given twice, 48 hours apart.

Stomachs harvested from infected mice were cut into three or four longitudinal sections. One section (1/3–1/4 of the stomach) was added to a pre-weighed reinforced tube with five (5) 2.4 mm stainless steel beads (Cat no. 15-340-158 ThermoFisher Scientific, Waltham, MA, USA) containing 600 µL of 10% FBS in Brucella broth on ice. The tube was weighed again after the tissues were added, then homogenized with two 30-second cycles containing a 3-second rest in between, on a speed setting of five in the FisherBrand BeadMill 24 (ThermoFisher Scientific, Waltham, MA, USA). Four hundred microliters of Brucella broth in 10% FBS was added to bring the solution up to 1 mL and then diluted with chilled 10% FBS in Brucella broth into 10², 10³, and 10⁴ solutions. A volume of 100 µL of this solution was then plated onto TSA plates containing 5% sheep blood (Hemostat Laboratories, Dixon, CA, USA) and nalidixic acid (10 µg/mL), vancomycin (50 µg/mL), amphotericin (2 µg/mL), and bacitracin (100 µg/mL). Inoculated plates were stored in airtight containers with BD GasPak EZ Sachets (Becton Dickinson, Franklin Lakes, NJ, USA) at 37°C for 7 days. Colony forming units (CFUs) were then quantified and normalized to grams of stomach tissue homogenized. Log transformation of CFU/g was performed, and an unpaired t-test was performed using GraphPad Prism (GraphPad, San Diego, CA, USA) to determine significance.

Histological analysis of gastric tissue

The middle section of the mouse stomach (duodenum-corpus) from the harvest was placed onto Whatman paper, into a cassette, and moved into 10% neutral buffered formalin for a minimum of 4 hours before processing. Cross-sections of 5 µm were cut and stained for hematoxylin and eosin (H&E). A single pathologist scored indices of inflammation and cancer. Several variables were graded on a 0–3 scale (0, none; 1, mild; 2, moderate; 3, severe) in the gastric antrum and corpus: acute inflammation (polymorphonuclear cell infiltration) and chronic inflammation (mononuclear cell infiltration independent of lymphoid follicles); thus, a maximum inflammation score of 12 was possible for each animal. Low-grade dysplasia was defined as irregular, angulated, and occasionally cystically dilated glands with enlarged overlapping hyperchromatic nuclei. Carcinoma was defined as irregular, angulated, cystically dilated glands with occasional cribriform architecture in the submucosa and muscularis propria, spreading laterally to the surface mucosal component. Alcian blue stainings were performed by the VUMC Translational Pathology Shared Resource. Images were taken on an Evos FL Auto 2 (ThermoFisher Scientific, Waltham, MA, USA).

Immunohistochemistry

Five micrometer unstained cross-sections (formalin-fixed, paraffin-embedded gastric specimens) were deparaffinized by routine methods. For antigen retrieval, the sections were heated to 97°C–105°C in decloaking solution as indicated by antigen in Table S1, and then allowed to cool to room temperature. After the retrieval, the tissue sections were quenched with 3% H₂O₂ in sodium azide for 5 minutes at room temperature. Anti-CD45r (B220), anti-CD4, anti-Muc5ac, or IgA antibodies (Table S1) were used at the concentrations in Table S1, and then incubated for 60 minutes with the tissue sections, followed by antibody localization using an HRP-labeled polymer (DAKO or BioCare as indicated in Table S1. Staining was visualized by 5-minute incubation with chromogen diaminobenzidine.

Immunofluorescence

A small piece of the mouse stomach (corpus) from the harvest was placed into 2% paraformaldehyde for 1.5 hours. During fixation, the tissue was sub-dissected into 2 mm²-sized pieces. The tissue was washed in phosphate-buffered saline (PBS), permeabilized in 0.2% Triton in PBS containing 0.1% Tween 20 (PBST) then blocked overnight at 4°C in 5% normal goat serum (NGS). After blocking, the tissue was washed in PBST. The tissue was then serially incubated as follows, with PBST washes after each staining: Alexa Fluor 594 conjugated wheat germ agglutinin (WGA) 1:200 in NGS for 2 hours at room temperature (RT), Alexa Fluor 647 Phalloidin 1:200 in NGS at RT for 2 hours, and then Hoechst nuclear stain 1:200 in NGS at RT for 1 hour. After a final PBST wash series, tissue was mounted on #1.5 cover glass luminal side down in ProLong Gold Antifade mounting media and allowed to polymerize for 24 hours prior to imaging.

Fluorescence light microscopy and image processing

Confocal microscopy was performed using a Nikon FN1 upright light microscope equipped with an A1 R HD25 laser-scanning head, four excitation lasers (405, 488, 561, and 640 nm), and a 60×/1.4 NA objective. All images used for quantitative comparisons were prepared with equal treatment, acquired with identical parameters (e.g., pinhole diameter and detector gain), and processed in an identical manner. Post-processing to remove noise was completed using the Denoise.ai deep learning algorithm in Nikon Elements software. Of note, images acquired with the resonant scanner followed by Denoise.ai processing yielded similar results to images acquired with the Galvano scanner, so the resonant scanner was used to reduce image acquisition time. Image stacks were viewed as a maximum intensity projection (Fig. 5A, main panel) or as three-dimensional, depth-coded images (Fig. 5A, inset) generated in Nikon Elements. Images were contrast-enhanced and cropped using a combination of Nikon Elements and ImageJ/FIJI software from the National Institutes of Health (NIH).

Image analysis and statistics

Confocal image analysis and signal intensity measurement from image data were performed using ImageJ/FIJI (NIH). To perform intensity analysis (Fig. 5C), linear regions of interest (ROIs) were drawn perpendicular to the tissue surface. The three-value walking average background-subtracted fluorescence intensity (arbitrary units) from raw 16-bit images along each ROI was measured (sample graphs in Fig. 5C), and peak values were compared (Fig. 5D).

RNA isolation and real-time rtPCR

A longitudinal stomach section (approximately 1/3–1/4 of the tissue) was used for RNA extraction. One milliliter of TRIzol Reagent (Invitrogen, Waltham, MA, USA) along with the stomach section was added to the gentleMACS Dissociator (Miltenyi Biotec, Begisch Gladbach, Germany) (m-tubes) and homogenized. Homogenate was transferred to a microcentrifuge tube, and the TRIzol protocol was followed with slight modifications (two chloroform extraction steps). Paragastric lymph node RNA was also extracted via the TRIzol protocol, but the tissue was manually lysed in a 12-well plate with a 3 mL syringe plunger and 500 µL of TRIzol. At the end of the protocol, the pellet was then resuspended in 100 µL nuclease-free water before following the manufacturer’s recommendation for the Qiagen RNeasy Mini kit’s clean up protocol (Qiagen, Hilden, Germany). When possible, RNA was adjusted up to 5,000 ng before making cDNA via applied biosystems by Thermo Fisher Scientific HighCapacity cDNA Reverse Transcription Kit (ThermoFisher Scientific, Waltham, MA, USA) per the manufacturer’s specifications. A quantity of 5 µL of this cDNA (diluted 1:10) was used in real-time assays powered by TaqMan Fast Advanced Master Mix, TaqMan genetic probes, and Applied Biosystems QuantStudio 6 Flex (Applied Biosystems, Waltham, MA, USA) per the manufacturer’s specifications.

For real-time rtPCR, the relative gene expression method was utilized in order to measure the gene expression of IL17a, IL21, Dbef14, Pigr, Nox1, Ceacam1, Itln1, and Ltf in the stomach. Gene expression of IL17a, IL21, Fn1, and Cola1a was measured from paragastric lymph node samples. For both tissues, glyceraldehyde 3-phosphate dehydrogenase (Gapdh) was used as the endogenous control for normalization. All real-time q-rtPCR was performed using an Applied Biosystems QuantStudio 6 Flex instrument using TaqMan Fast Advanced Master Mix (ThermoFisher Scientific, Waltham, MA, USA). The levels of gene transcription are indicated as “relative units” based on a comparison of tissue from H. pylori-infected mice with tissue from uninfected mice. All samples were calibrated to a control sample prepared from uninfected gastric tissue of WT mice. Relative units were calculated using the 2^-ΔΔCt method, where the ΔCt is calculated as the difference in the cycle threshold of the gene of interest from the endogenous control gene (Gapdh) and then the ΔΔCt is the difference between the ΔCt of the sample compared to the control sample (uninfected tissue). Primer and probe sets were purchased as Taqman Gene Expression Assays from Applied Biosystems [mouse primer sets: Ceacam-1 (Mm04204476), Defb14 (Mm00806979_m1), Itln-1 (Mm04243906), Ltf (Mm00434787), Gapdh (Mm99999915_g1), Il17a (Mm00439619_m1), Ifng (Mm99999071_m1), Pigr (Mm00465049_m1), Nox1 (Mm00549170_m1), Il21 (Mm00517640_m1), Col1a1 (Mm00801666_g1), and Fn1 (Mm01256744_m1) (Applied Biosystems, Waltham, MA, USA)].

Flow cytometry of the paragastric lymph node

The paragastric lymph node digestion protocol was adapted from Nayar et al. (23). Briefly, paragastric lymph nodes were collected and placed in 5 mL of RPMI-1640 without L-glutamine with 2% FBS on ice until the conclusion of the harvest. The samples were transferred to 5 mL polystyrene tubes with flea magnetic stir bars and incubated at 37°C for 1 hour in Collagenase P buffer [0.8 mg/mL Collagenase Dispase (Cat no. 11097113001), 0.2 mg/mL Collagenase P (Cat no. 11213857001), DNase I (Cat no. 10104159001) (Sigma Aldrich, St. Louis, MO, USA)]. The digested lymph nodes were added to RPMI-1640 without L-glutamine with 2% FBS to neutralize the reaction, centrifuged at 1,400 rpm for 5 minutes at 4°C, and resuspended in 20% mouse serum in fluorescence-activated cell sorting (FACS) buffer (0.5% bovine serum albumin in 1x PBS buffer) as an Fc Block and for surface staining with the following antibodies: Anti-CD45/PE-Cy7 (clone 30-F11 Tonbo San Diego, CA), Anti-gp38/PE (clone 8.1.1 eBioscience), Anti-CD31/BV421 (clone 390 eBioscience, San Diego, CA, USA), Anti-CD4/PE-Cy5 (clone H129.19 BD Bioscience, Franklin Lakes, NJ, USA), and Anti-CD45R/APC (clone RA3-6B2 Tonbo, San Diego, CA, USA). The samples were then washed and resuspended in 300 µL FACS buffer plus CountBright Plus Absolute Counting Beads (Cat. no. C36995 Invitrogen, Waltham, MA, USA) for acquisition on a Cytek Aurora 4-laser and analysis on SpectroFlo (Cytek, Fremont, CA, USA).

Fitc-dextran barrier assay

The barrier assessment assay was adapted from Chanez-Paredes et al. (24). In short, mice were deprived of bedding and food for 3 hours before orogastric gavage with 80 mg/mL FITC-4 kDa dextran (60842-46-8 Sigma Aldrich, St. Louis, MO, USA). The control mice received a gavage of ultra-pure water. One hour after gavage, 0.5 mL of blood was collected via cardiac puncture and added to 150 µL anti-coagulant sodium citrate solution. Blood was then centrifuged at 1,500 × g for 10 minutes at 25°C. One hundred microliters of plasma from each experimental sample was then added to a single well of a black costar 96-well assay plate (Cat no. 3904, Corning, Corning, NY, USA) to be read against a standard curve of serially diluted probe stock at Ex/Em 495/525 nm at gains of 50, 60, 70, 80, 90, and 100 on CYTATION3 Imaging Reader and analyzed on Gen5.304 software. Control sample (no FITC-Dextran) relative fluorescence unit values were subtracted from experimental sample values and graphed in GraphPad Prism.

Immunoglobulin A ELISAs

Fecal samples were collected within 24 hours of the experimental end points. Fecal pellets were resuspended by vortexing and/or pipetting in 1× PBS with 0.01% NaN₃ and protease inhibitors 1 mL/0.1 g feces (cOmplete Tablets EASYpack Protease Inhibitor Cocktail Tablets, Roche, Basel, Switzerland). This provided normalization by fecal mass since, in subsequent steps, the same volume of each sample was used for the assay. Debris was centrifuged out of the samples (14,000 rpm for 15 minutes at room temperature, twice). The samples were then run through the Invitrogen Mouse IgA ELISA Kit (88-50450-22 Waltham, MA, USA). First, each sample was diluted 1:50 in assay/sample diluent and then, following the manufacturer’s instructions, the amount of IgA was quantified on the BioTek ELx808 plate reader (BioTek, Winooski, VT, USA). Standards were run in duplicate. Four-parameter logistic regression was applied to fit the unknowns to the standard curve.

Gastric washes were collected at the time of tissue harvest. Excised stomach tissue was opened along the lesser curvature and washed in 1 mL of PBS with 0.01% NaN₃ and protease inhibitors by shaking for 10 seconds. Gastric wash was processed to remove large debris/food by spinning at 13,000 rpm for 10 minutes at 4°C. Total IgA was determined using undiluted gastric wash samples in the ThermoFisher IgA ELISA kit as directed by the manufacturer (Cat no. 88-50450-22) with one modification, the standard was diluted to allow for the highest standard to be at 50 ng/mL.

For the H. pylori-specific ELISA, ELISA plates were coated with H. pylori lysate (PMSS1 strain) at 10 µg/mL overnight and then gastric wash samples were applied undiluted. Following incubation and wash with ELISA wash buffer (Cat no. 00-0400-59, Invitrogen, Waltham, MA, USA), a goat anti-mouse IgA antibody-HRP secondary (1040-05 Southern BioTech, Birmingham, AL, USA) was diluted 1:6,000 in assay buffer and added. Signal was developed with 1× 3,3’,5,5’-tetramethylbenzidine (TMB) substrate (Cat no. 00-4201-56, Invitrogen) followed by 1 M H₂SO₄ stop solution, and absorbances were read as described above. The data are reported as the fold above the absorbance reading of gastric wash from uninfected mice (a mixed pool of three uninfected samples).

RESULTS

IL-17RA deficiency in FVB/n mice has more profound impact on pathology than deficiency in C57Bl/6 mice

Previous publications investigating how IL-17RA deficiency impacted H. pylori-induced gastric inflammation were performed in C57Bl/6 mice (21, 22). To further investigate the mechanisms by which IL-17 controls inflammation during H. pylori infection, an IL-17RA-deficient mouse was generated on the FVB/n background using CRISPR/Cas9 (see Materials and Methods; Fig. S1). Mice on the FVB/n background allow for the examination of responses in the paragastric lymph node, which is difficult to consistently isolate in C57Bl/6; furthermore, research is underway in a carcinogenesis model, which is based on the FVB/n background.

To determine if the Il17ra^em1(hmsa) mice (FVB/n background) had the same phenotype in response to H. pylori infection as the Il17ra^tm1Koll mice (21) (C57Bl/6 background), mice were infected with H. pylori strain PMSS1, which contains a functional type 4 secretion system, and at 6 weeks and 3 months post-infection, tissues and biological samples were harvested. Colonization efficiency is <100% in mice and especially for mice on the FVB/n background. Il17ra^em1(hmsa) had a higher colonization rate than FVB/n control mice (Fig. S2), but when successfully colonized, there is no significant difference in bacterial burden (Fig. S2). While some Il17ra^-/- mice start to exhibit increased inflammation by 6 weeks post-infection, there are no consistent changes in the pathological outcomes at this acute infection time point (Fig. S3). Consistent with IL-17RA deficiency in C57Bl/6 mice (21, 22), the Il17ra^-/- mice exhibited significantly more inflammation and developed significantly more lymphoid follicles after H. pylori infection by 3 months post-infection. The increased immunopathology was more pronounced in IL17ra^-/- mice on the FVB/n background compared to those on the C57Bl/6 background (Fig. 1). Concomitant with the increased number of lymphoid follicles and gastritis, T cell cytokines, including Il17a and Il21, are expressed at significantly higher levels in Il17ra^-/- compared to the wild-type (FVB/n) mice after H. pylori infection (Fig. 2A). B lymphocytes and CD4⁺ T lymphocytes are commonly found in lymphoid follicles, which are a prominent feature of the immunopathology of H. pylori-infected Il17ra^-/- mice. Immunohistochemistry was performed on mice infected at 3 months post-infection to determine how these cellular infiltrates organize. In Fig. 2B, it is evident that CD4⁺ T lymphocytes are found both in lymphoid follicles and in infiltrating bands of cells and between glands out toward the lumen of the stomach. In the H. pylori-infected Il17ra^-/- mice, B220⁺ B lymphocytes organize within lymphoid follicles, and few B220⁺ cells are observed in the H. pylori-infected FVB/n mice at this same time point (Fig. 2C).

Fig 1.

IL-17RA is required to limit pathology in response to H. pylori infection. Acute and chronic inflammation in stomach tissue at 3 mpi (months post-infection) with strain PMSS1 were scored by a blinded pathologist. Il17ra^-/- mice showed exacerbated inflammation (A) and increased lymphoid follicles (B) compared to respective control mice, FVB/N or C57Bl/6 mice. Statistical analysis was performed using a two-way ANOVA by comparing the impact of the mouse background, as well as, the mouse genotype, and error bars represent mean ± SEM. See Materials and Methods for the scoring system (scale is 0–12). P ≤ 0.05, **P ≤ 0.01, and ****P ≤ 0.0001. Panels C and D are representative H&E stained tissues at 3 months post-infection. Images are of the antrum-corpus transition at 200× magnification illustrating increased total inflammation and increased lymphoid aggregates in WT mice (left) versus Il17ra^-/- mice (right); lymphoid aggregates are circled in the representative H&Es.

Fig 2.

Increased inflammation is associated with aberrant expression of IL-21 and increased B cell responses. (A) The level of T cell cytokine gene expression in RNA extracted from stomach tissue at 3 mpi with H. pylori was assessed by qRT-PCR. Il17a and Il21 expression was significantly higher in Il17ra^-/- mice than in wild-type mice. Relative units are calculated as described in the Materials and Methods, Gapdh was used as endogenous control, RNA from the tissue of uninfected mice was pooled and used as a reference sample, and an unpaired t-test was used to determine significance. Error bars represent ±SEM; *P  ≤  0.05, **P  ≤  0.01, ***P  ≤  0.001, and ****P  ≤  0.0001, compared with WT. (B) Representative staining for CD4⁺ T cells in mice infected with H. pylori for 3 months. (C) B220⁺ B cells in mice infected with H. pylori for 3 months. Genotypes are labeled (400× magnification); lymphoid aggregates are circled with a perforated line, while bands of infiltrating cells are highlighted with arrows and brackets.

IL-17RA is required for intact mucosal barrier response to H. pylori

With an exacerbation of the gastric inflammatory response (even without an increase in bacterial burden), we rationalized that the mucosal barrier of the stomach might be disrupted—either allowing for more antigen to be taken up and/or for more antimicrobials to cross the barrier into the lumen. Since the barrier is a complex environment, we surveyed the expression of several antimicrobials including lactotransferrin (Ltf), NADPH oxidase subunit 1 (Nox1), the polymeric immunoglobulin receptor (Pigr), Beta defensin 14 (Defb14), and Intelectin (Itln). The only difference found in these genes in uninfected mice was in Ltf relative gene expression (Fig. S4). After infection, H. pylori -infected Il17ra^-/- mice had significantly lower expression of Nox1, Defb14, and Pigr compared to WT mice infected with H. pylori. There was no difference in the expression of Ltf or Itln after 3 months of infection (Fig. 3A). To determine if the decrease in Pigr gene expression might impact the translocation of IgA to the gastric mucosa, immunohistological staining for IgA (Fig. 3B) and ELISAs for IgA were performed on both gastric washes and fecal samples from IL17ra^-/- and WT mice. The immunohistochemical stainings for IgA suggest there may be some different patterns of expression, but overall, there appears to be significant IgA levels detected in lymphoid follicles and at the interface between epithelial cells and the lumen, especially in the Il17ra^-/- mice. As previously reported for IL-17-deficient mice in studies focused on intestines (25), there is decreased IgA in fecal matter suggesting that IL-17 regulates Pigr expression (encoding pIgR) in the intestines. In our studies, even though there is a decrease in Pigr expression in the stomachs of H. pylori-infected Il17ra^-/- mice, total IgA levels and H. pylori-specific IgA levels are significantly higher in the gastric washes from H. pylori-infected Il17ra^-/- mice compared to H. pylori-infected FVB/n mice by 3 months post-infection (Fig. 3C). This suggests that B cells are producing high levels of IgA in the Il17ra^-/- mice and, further, that there is either sufficient pIgR to translocate the IgA or that IgA is also moving across the epithelium in a pIgR-independent mechanism (such as the leak pathway).

Fig 3.

IL-17RA regulates Pigr in the gastric tissue but does not result in IgA deficiency in the gastric mucosa. (A) Expression of genes associated with epithelial cell responses including Lft, Itln, Nox1, Defb14, and Pigr was assessed in stomach tissue at 3 mpi by qRT-PCR. Relative units are calculated as described in Materials and Methods, Gapdh was used as endogenous control, RNA from uninfected mice was pooled and used as a reference sample, and an unpaired t-test was used to determine significance. Error bars represent ±SEM; *P  ≤  0.05 and **P  ≤  0.01, compared with FVB/N. (B) Immunohistochemistry staining for IgA in FVB/N (top) versus Il17ra^-/- mice (bottom). All photomicrographs were taken at a total magnification of 400×. (C) ELISAs were used to quantify total IgA in feces, and to quantify total and relative Hp-specific IgA in gastric washes. Total IgA was calculated using a standard curve, while Hp-specific IgA was quantified as the fold increase above the signal from a pooled group of gastric washes from uninfected mice. (D) The serum concentration of Fitc-Dextran was determined as a measure of tissue permeability in H. pylori-infected FVB/N mice and Il17ra-/- mice. Significance was determined using an unpaired t-test. Error bars represent ±SEM; **P  ≤  0.01 and ***P  ≤  0.001, compared with FVB/N.

The barrier of the gastrointestinal tract was assessed using the FITC-Dextran barrier assessment. At 6 weeks post-infection, mice were fasted and then gavaged with FITC-Dextran. At 1 hour post-inoculation, serum was collected to quantify FITC-Dextran in the blood as a measure of the integrity of the epithelium. H. pylori-infected Il17ra^-/- mice had increased FITC-Dextran in their serum compared to H. pylori-infected FVB/n mice (Fig. 3D). These data suggest that the barrier integrity after H. pylori infection is altered in the absence of IL17RA.

T-cell activation in the paragastric lymph node response is amplified with IL-17RA deficiency

The strong T-cell and B-cell response in the gastric mucosa led us to ask if the site of priming and activation of the adaptive immune response may also be impacted by the IL-17 receptor deficiency. The paragastric lymph nodes, which are located on the lesser curvature of the stomach adjacent to the esophagus, were isolated from H. pylori-infected Il17ra^-/- mice and H. pylori-infected WT mice. Flow cytometry was performed to quantify cell populations in these PLN using CountBright Plus Absolute Counting Beads to quantify the total numbers of different cell types (gating scheme, Fig. S5). H. pylori-infected Il17ra^-/- mice had increased numbers of CD4⁺ and B220⁺ lymphocytes compared to H. pylori-infected FVB/n mice (Fig. 4A). To investigate T-cell activation in the PLN, especially the expression of Th17 cytokines, real-time rtPCR was used to measure Il17a and Il21 expression. Both Il17a and Il21 were significantly increased in H. pylori-infected Il17ra^-/- mice compared to H. pylori-infected FVB/N mice (Fig. 4B). Cytokine transcripts were rarely detected in PLN isolated from uninfected mice (data not shown).

Fig 4.

Lymph node cellularity and gene expression are altered in Il17ra^-/- mice. (A) Flow cytometry including counting beads was performed on PLN of H. pylori-infected mice to quantify CD4⁺CD45⁺ and B220⁺CD45⁺ cells. The data represent one of three similar experiments (gating scheme is presented in Fig. S5). (B) qRT-PCR assays on T cell cytokines genes (Il17a and Il21), and (C) fibroblastic reticular cell (FRCs)-associated genes (Cola1a and Fn1) were performed on RNA extracted from PLNs at 3 mpi. Significant increases were observed in T cell cytokines in Il17ra^-/- mice. Relative expression of Fn1 is lower in Il17ra^-/- mice compared to WT mice, while no significant difference was detected in Cola1a. Relative units are calculated as described in Materials and Methods, Gapdh was used as endogenous control, PLN RNA from uninfected mice was pooled and used as a reference sample, and an unpaired t-test was used to determine significance. Error bars represent ±SEM; *P  ≤  0.05, **P  ≤  0.01, ***P  ≤  0.001, and ****P  ≤  0.0001, compared with WT. (D) Flow cytometry including counting beads was performed on PLN of H. pylori-infected mice to quantify FRC (CD45-gp38 + CD31-). The data represent one of three similar experiments (gating scheme is presented in Fig. S5).

It was previously reported that a subset of stromal cells, fibroblastic reticular cells (FRC), which are located in the T-cell zone of the lymph node’s cortex and help organize its structure and function, rely on IL-17RA signaling to activate metabolic processes to survive (26). In the EAE model, Il17ra^-/- mice (on the C57Bl/6 background) do not have FRC proliferation, have impaired germinal center formation, and reduced antigen-specific antibody production (26). This is counter to our model where we observe chronic activation, germinal center formation in tissues, and high antigen-specific antibody responses. We postulated that the lymphoid follicle development in the gastric mucosa may be compensating for a lack of responses in the paragastric lymph node. To investigate the presence of FRC in the paragastric lymph nodes of Il17ra^-/- mice and FVB/N mice after infection, real-time rtPCR was performed. Genes associated with FRCs, Cola1a (collagen, type I, and alpha 1) and Fn1 (fibronectin), were not significantly different in uninfected samples (Fig. S4B) but have a lower relative expression in PLNs from H. pylori-infected Il17ra^-/- mice compared to PLN H. pylori-infected WT mice (Fig. 4C), but the total number of FRC is not different between these genotypes. This suggests that there is an expansion of T cells and B cells in the absence of IL17RA, but that FRC does not expand proportionally.

Loss of mucin production and shifts in epithelial cell populations in H. pylori-infected IL-17RA-deficient mice

In addition to the evidence that IgA is abundant in the gastric lumen despite reduced expression of Pigr, we examined other components of the barrier in the stomach. Alcian blue stainings were performed to investigate the patterns of mucin production. The stainings of H. pylori-infected tissues revealed that, in addition to some changes in gland structure, mucin-producing cells at the top of the glands are not as abundant in Il17ra^-/- mice compared to FVB/n mice by 3 months post-infection (Fig. S6). Due to the localization of the mucin observed in the staining, confocal immunofluorescence microscopy was performed with WGA to localize glycoproteins in the corpus tissue and nuclear and F-actin stains to highlight the tissue structure at 4 months post-infection. Maximum intensity projection of the surface epithelium and three-dimensional reconstruction of the mucus layer were suggestive of decreased mucus thickness (Fig. 5A). Single confocal planes were used for quantification (Fig. 5), which confirmed a decrease in WGA signal at the tissue surface, consistent with a thinner mucus layer and suggestive of reduced mucin production in the Il17ra^-/- compared to FVB/n (Fig. 5B through D). In the corpus, the epithelial cells at the top of the glands are mucous pit cells, which express Muc5ac, therefore, Muc5ac expression and localization were determined with immunohistochemistry on tissues at 3 months post-infection (Fig. 5E) and at the transcript level in whole tissue (Fig. S6). Together, these data suggest that Muc5ac expression is noticeably reduced in the corpus of Il17ra^-/- mice, especially in areas where the tissue architecture is disrupted compared to FVB/n mice at this time point.

Fig 5.

The mucus layer and mucus production, specifically Muc5ac, are reduced in the absence of IL-17RA. Confocal microscopy immunofluorescence of whole mount stomach (corpus) at 4 months post-infection (A–C). Hoechst (DNA), WGA—stains glycoproteins, and phalloidin (F-actin). (A) Maximum intensity projection of the surface epithelium and mucus (~20–30 µm deep). Main panel scale bars: 100 µm. Insets highlight the mucus layer, 19.8 µm depth-coded, three-dimensional rendering. Inset xy scale bars: 100 µm. Depth-code scale to the right, with cooler colors representing closer layers and warmer colors representing deeper layers. (B) Cropped representative single confocal planes selected for areas optically sectioned orthogonal to the tissue surface. Dashed white boxes indicate areas highlighted in (i–iv). Scale bars: 100 µm. (C) Enlarged single confocal plane of surface WGA staining. Blue and orange arrows indicate sample intensity regions of interest (ROIs) for FVB/n and Il17ra^-/- tissues, respectively. A 10-µm line was drawn perpendicular to the tissue surface from intracellular to extracellular. Corresponding plots of background-subtracted intensity data are graphed at right. Intensity is in arbitrary units (a.u.) measured from RAW 16-bit images. (D) Quantification of peak surface WGA intensity. Fifty ROIs per condition distributed across at least 10 confocal planes. Data visualized as a violin plot; median (central dotted line) and quartiles (outer dotted lines) are indicated; P values were calculated using Mann-Whitney U Test (****P ≤ 0.0001). (E) Immunohistochemistry was performed to localize Muc5ac expression, a gastric Pit cell marker, in stomach tissues both antrum and corpus. Representative staining at 3 months post-infection, total magnification 400×.

IL-17RA deficiency impacts epithelial cell populations during H. pylori infection

To understand the pathology further, dysplasia, loss of parietal cells, loss of chief cells (oxyntic atrophy), mucous neck cell hyperplasia, and foveolar hyperplasia were scored. Two Il17ra^em1(hmsa) mice in a cohort of seven mice developed pathological changes consistent with low-grade dysplasia at 3 months post-infection (Fig. 6A through I), while no WT FVB/n exhibits this pathological outcome. The dysplasia phenotype has not been observed in Il17ra^tm1Koll mice or C57Bl/6 mice over several years of experimentation. Neither genotype of H. pylori-infected mice exhibited mucous neck cell hyperplasia by 3 months post-infection (data not shown). The data on the expression of Itln1 support this since Itln1 has been identified as a mucous neck cell marker in mice. However, loss of parietal cells and chief cells was reported more frequently in the Il17ra^-/- mice compared to FVB/N mice (Fig. 6A), and an increase in mild foveolar hyperplasia was observed in the Il17ra^-/- mice compared to FVB/N mice at the 3 months post-infection time point (Fig. 6B).

Fig 6.

Histological findings support the loss of secretory cells and increase in hyperplasia and dysplasia. Formalin-fixed and paraffin-embedded H&E tissues are scored by a blinded pathologist at 3 mpi. (A–F) Dysplasia observations in some Il17ra^-/- mice. At 3 mpi, two out of nine Il17ra^-/- mice had mild dysplasia. Panels A–C correspond to an Il17ra^-/- mouse tissue at increasing magnification from left to right (total magnifications 40×, 100×, and 200×). This histology exhibits some distorted, dilated glands with mild dysplastic changes in the distal corpus and transitional mucosa. There are some focal changes suspicious of intramucosal carcinoma. Panels D–I correspond to a different Il17ra^-/- mouse. Panels E and F focus on the corpus and H and I focus on the antrum again increasing in magnification from left to right (40×, 100×, and 200×). This histology exhibits low-grade dysplasia and focal suspicion of invasive adenocarcinoma in the proximal antrum. Band-like inflammatory infiltrate in the antral and corporal submucosa was observed. (J) Tissues were scored for parietal and chief cell atrophy (oxyntic atrophy), and (K) the occurrence of foveolar hyperplasia. Mann-Whitney test was performed to test statistical significance. Error bars represent ±SEM; *P  ≤  0.05 and **P  ≤  0.01 compared with FVB/N. (L) Gene expression of the tight junction regulating protein Ceacam1 was measured at 6 wpi and 3 mpi by qRT-PCR. Relative units are calculated as described in Materials and Methods, Gapdh was used as an endogenous control, RNA from uninfected mice was pooled and used as a reference sample, and an unpaired t-test was used to determine significance. Error bars represent ±SEM; **P  ≤  0.01, compared with FVB/N.

Ceacam1 has been reported to be expressed in transforming epithelial cells and functions as a tight junction barrier protein. To determine if the expression of the Ceacam1 gene was changed with IL-17RA deficiency with increasing pathological changes, real-time rtPCR was performed to measure the expression of Ceacam1 at 6 weeks and 3 months post-infection. Interestingly, a significant difference was not observed at 6 wpi, but at 3 mpi there was a significant increase in the expression of Ceacam1 in Il17ra^-/- mice compared to FVB/N mice (Fig. 6C). This supports the notion that there may be multiple changes in the gastric epithelial cell barrier in the absence of IL-17RA.

DISCUSSION

IL-17 responses have long been associated with the control of neutrophil infiltration, and therefore IL-17 has been associated with protective antimicrobial immune responses of extracellular pathogens. Furthermore, many studies both in the lungs and in the intestines have supported the contribution of IL-17 responses to the production of antimicrobial factors. Our studies in Il17ra^-/- mice, IL17a^-/-, Il17f^-/-, and Il17a/f^-/- mice infected with H. pylori have also shown that IL-17 signaling contributes to antimicrobial responses including recruitment of neutrophils, production of polymeric immunoglobulin receptor (pIgR), beta defensins 14, lipocalin 2, and others (21, 22, 27). A remaining gap in our research is understanding how these deficiencies in antimicrobial responses are linked to the exacerbated chronic inflammatory response observed in Il17ra^-/- mice after H. pylori infection. Much less is understood about these mucosal barriers in the stomach tissue. We have hypothesized that loss of barrier function leads to increased antigen uptake and increased priming and activation of responses in the paragastric lymph nodes.

Working with Il17ra^-/- mice created in the FVB/n background provided an opportunity for us to investigate the role of IL-17 in the activation and priming of the response in the stomach draining lymph node. Our findings suggest that while there may be reduced expression in the paragastric lymph node of Cola1a and Fn1 (collagen a1a and fibronectin), markers of FRC, this reduction does not lead to a deficiency in T-cell activation or B-cell activation. In fact, the paragastric lymph nodes are more enlarged in IL-17RA-deficient mice than in WT mice. This hyperactivation of Th17 responses in the lymph nodes may be driven by more antigens crossing the barrier of the stomach. Unfortunately, this is hard to directly measure. In this manuscript, we provide evidence of several changes to barrier and barrier function molecules including changes in Pigr expression, Nox1 expression, and mucin production by mucous pit cells.

The hyperinflammatory response in the H. pylori-infected Il17ra^-/- mice is not restricted to the lymphatics, and in fact, includes the development of lymphoid follicles in the gastric mucosa. The development of the lymphoid follicles correlated with increased total IgA and H. pylori-specific IgA. Furthermore, there are other changes observed in Il17ra^-/-, which suggest an impacted barrier including loss of mucin production by the mucous pit cells. Mucous pit cells are known to produce Muc5ac. It has been demonstrated that surface foveolar Muc5ac expression is reduced in mice with areas of severe glandular mucous metaplasia by 32 weeks post-H. pylori infection compared to mice, which have mild metaplasia or are not infected (28). Interestingly, Muc5ac-deficient mice develop dysplasia with age, and infection with H. pylori reduced the incidence of dysplasia in the antrum-pyloric region. In the corpus, H. pylori infection of Muc5ac^-/- leads to reduced mucous metaplasia compared to H. pylori infection of WT mice at 32 weeks post-infection. In the Il17ra^-/- in this study, we cannot tie the reduced expression of mucins in the pit cells to exacerbated inflammation or a barrier defect, but together these data do suggest that Muc5ac does impact mucosal barrier homeostasis. While there are differences between populations of secretory cells of the stomach and the intestines, it was demonstrated that IL-17RA signaling in ATOH1 transcription factor positive intestinal epithelial cells was required to regenerate secretory cells following injury in the mouse model of dextran sulfate sodium (DSS)-induced colitis (29). Furthermore, IL-17A acted on Lgr5⁺ intestinal stem cells to promote secretory cell lineage commitment for Paneth, tuft, goblet, and enteroendocrine cells. Together, these data suggest and support our suggestion that IL-17 signaling may be required for the maintenance and integrity of secretory cells in the gastric mucosa.

ETHICS STATEMENT

The Institutional Animal Care and Use Committee (IACUC) of Vanderbilt University Medical Center and the Research and Development Committee of the Veterans Affairs Tennessee Valley Healthcare System approved all animal procedures in this study under protocols V1800070 and V2000068. Animal experiments were performed in accordance with AAALAC guidelines, the AVMA Guidelines on Euthanasia, NIH regulations (Guide for the Care and Use of Laboratory Animals), and the United States Animal Welfare Act. Mice were housed in an accredited research animal facility that is fully staffed with trained personnel.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/iai.00292-23.

Supplemental material. iai.00292-23-s0001.docx.

Figures S1 to S6 and Table S1.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.