Helicobacter pylori Infection Impairs the Mucin Production Rate and Turnover in the Murine Gastric Mucosa

Abstract

To protect the surface of the stomach, the epithelial cells secrete a mucus layer, which is mainly comprised of the MUC5AC mucin. Further protection is provided by a thick glycocalyx on the apical surface of the epithelial cell, with the cell surface mucin MUC1 as a major component. Here, we investigate the production rate and turnover of newly synthesized mucin in mice and analyze the effects of early colonization and chronic infection with H. pylori. Metabolic incorporation of an azido GalNAc analog (GalNAz) was used as a nonradioactive method to perform pulse experiments in the whole animal. First, the subcellular movement of newly synthesized mucin and mucin turnover was determined in uninfected mice. Based on the time line for mucin transport and dissemination, 2, 6, and 12 h after GalNAz injection was selected to collect the stomachs from mice infected with H. pylori strain SS1 during early colonization (7 days) and chronic infection (90 days). The results demonstrated that the speed from the start of glycosylation to the final destination is faster for the membrane-bound mucin to reach the glycocalyx (2 h) than for the secretory mucins to become secreted into the mucus layer (5 h). Furthermore, infection with H. pylori reduces the rate of mucin turnover and decreases the levels of Muc1. Since H. pylori colonizes this mucus niche, the decreased turnover rate indicates that H. pylori creates a more stable and favorable environment for itself by impairing the defense mechanism for clearing the mucosal surface of pathogens by mucus flow.

INTRODUCTION

Helicobacter pylori is a Gram-negative bacterium infecting half of the world's population. Despite the fact that most of the individuals infected with this bacterium are asymptomatic, H. pylori can cause gastric and duodenal ulcers and mucosally associated lymphoid tissue lymphoma in a subset of infected individuals (1, 2).

In order to protect the stomach from invading pathogens, the epithelial cells of the mucosal surface constantly secrete a mucus layer, which is mainly comprised of the secreted mucin MUC5AC. MUC5AC is produced by the surface epithelium and another secreted mucin, MUC6, by the glands (3–6). The cell surface mucin MUC1 is a major component of the gastric glycocalyx, which provides a protective barrier on the apical surface of epithelial cells (7) and may initiate a signaling pathway in response to invasion (8). Mucins are heavily glycosylated glycoproteins consisting of a protein backbone with a large number of O-linked glycosaccharides and a few N-glycan chains. Mucins are secreted via both constitutive and regulated pathways. The constitutive pathway causes a continuous secretion of mucin to maintain the mucus layer washing the mucosal surface, whereas the regulated pathway affords a massive discharge as a response to environmental and/or pathophysiological stimuli (9). Mucin release is accompanied by hydration, resulting in approximately a thousand-fold expansion in volume of the secretory granule contents (10).

H. pylori can bind to both MUC1 and MUC5AC via fucosylated and sialylated glycans in a pH-dependent manner (11, 12). In the rhesus monkey model of H. pylori infection, monkeys with mucins that bind to H. pylori more effectively have a lower H. pylori density in their stomachs, indicating that mucin binding to H. pylori aids in removing the bacteria from the gastric niche (13). Binding of H. pylori to MUC1 results in release of the extracellular domain of the mucin from the epithelial surface, thereby acting as a releasable decoy and preventing prolonged adherence (14). Mucin glycosylation changes the ability of H. pylori to adhere to the mucins and is dependent on the time point of infection (13). During the acute phase of H. pylori infection in adult patients, there is an infiltration of neutrophils, followed by all classes of inflammatory cells, predominantly lymphocytes, accompanied by transient gastric hypochlorhydia. However, in chronic infection, while there is also inflammatory cell infiltration (predominantly lymphocytic), this is accompanied by local gastric inflammation (15). The time course of gastric mucin turnover and eventual alteration of this process during acute and chronic phases of infection with H. pylori is currently unknown. In this study, we used in vivo metabolic labeling to evaluate the production rate and turnover of mucins in the stomach of noninfected mice, and we compared it to mice with early colonization or chronic H. pylori infection. We also investigated the differences in the levels of Muc1 and Muc5ac.

MATERIALS AND METHODS

Animals.

Six- to 8-week-old, specific-pathogen-free, female C57BL/6 mice were purchased from Taconic (England) or Charles River (Germany). The mice were housed in individually ventilated cages at the Laboratory for Experimental Biomedicine (EBM) for the duration of the study. The animals had free access to water and food throughout the experiment and were monitored daily. All experimental procedures were approved by the ethics committee for animal experiments (Gothenburg, Sweden).

Intraperitoneal GalNAz injection and tissue fixation.

A total volume of 0.5 ml per mouse of Click-iT reagent was prepared by dissolving 2.6 mg of the Click-iT reagent GalNAz (Invitrogen) in 50 μl dimethylsulfoxide (DMSO) and was diluted with 6 mg/ml bromodeoxyuridine (Sigma-Aldrich) in phosphate-buffered saline (PBS; 0.15 mol/liter NaCl, 5 mmol/liter sodium phosphate buffer, pH 7.4). The injections were given intraperitoneally during the dark period. For the basal turnover study, a total number of 24 mice (from Taconic, England) were used, and 2 mice were sacrificed every hour for 12 h. Although only two mice were harvested each hour, the location of newly synthesized mucin was consistent between these two mice, and the incremental movement of newly synthesized mucin from one hour to the next was small. Furthermore, the results from the mice in the noninfected control group in the infection study (below) confirmed the consistency and reproducibility of the method. For the infection study, 18 mice (from Charles River, Germany) were sacrificed 2, 6, and 12 h after GalNAz injection. To keep the secreted mucus layer as undisturbed as possible, the entire glandular part of the stomach (containing luminal material), along with a small piece of duodenum, were fixed in Carnoy's methanol (60% dried methanol, 30% chloroform, 10% glacial acetic acid).

Cultivation of H. pylori used for infection.

The mouse-adapted H. pylori SS1 strain, stored at −70°C as aliquots in Luria-Bertani medium containing 20% glycerol, was used as the stock culture. The bacteria were grown for 3 days on Columbia iso-agar (Becton, Dickinson [BD]) plates and further cultured overnight in Brucella broth (BD) under microaerophilic conditions. Before infecting the mice, the optical density (OD) of the bacteria was adjusted to 1.5, corresponding to approximately 1 × 10⁹ viable bacteria/ml.

Infecting the mice.

Mice were infected with 3 × 10⁸ live H. pylori SS1 bacteria in 300 μl Brucella broth, administered orally using a feeding needle under anesthesia, corresponding to approximately 100 times the minimal effective dose of colonization to ensure that all mice were colonized. For the mucin turnover study, a total of 18 mice were divided into three groups: noninfected, early colonization with H. pylori (sacrificed 7 days postinfection), and chronic H. pylori infection (sacrificed 90 days postinfection). As it was necessary to fix the entire stomach without opening it to keep the mucus layer intact, we determined the number of H. pylori organisms in the murine stomachs in a separate set of experiments.

Quantitative culture of H. pylori SS1 from the stomach.

To evaluate bacterial colonization, one-half of each stomach was homogenized in Brucella broth using a tissue homogenizer (Ultra Turrax; IKA Laboratory Technologies, Staufen, Germany). Serial dilutions of the homogenates were plated on BD modified Helicobacter agar (BD). After at least 7 days of incubation at 37°C under microaerophilic conditions, visible colonies with typical H. pylori morphology were counted, and the urease test was performed for any uncertain colonies. Plates with 10 to 100 colonies were used for calculating the number of bacteria per stomach by multiplying by the appropriate dilution factor.

Histological scoring of mouse stomach infected with H. pylori.

Hematoxylin and eosin (H&E) staining was performed on fixed (Carnoy's methanol) paraffin-embedded samples containing the full length of longitudinally cut sections of the gastric specimen, starting at the forestomach and ending at the duodenum. A whole longitudinal section for each mouse in all groups was examined for scoring of gastritis. Gastritis scoring for corpus was performed from the limiting ridge of the stomach to the junction between corpus and antrum, whereas scoring for antrum was performed on the segment from the junction between antrum and corpus to the junction between duodenum and antrum. The blinded gastritis scores were obtained based on the level of cellular infiltration, number of abscesses, mucus metaplasia, and atrophy. The cellular infiltration (migration of lymphocytes and neutrophils) was graded from 0 to 6, where 0 is none; 1 is mild multifocal; 2 is mild widespread or moderate multifocal; 3 is mild widespread and moderate multifocal, or severe multifocal; 4 is moderate widespread; 5 is moderate widespread and severe multifocal; and 6 is severe widespread. In control uninfected mice, we could detect a few infiltrating immune cells, which were accepted as the baseline for scoring. The total number of abscesses was counted (including very small borderline ones) in a full-length section in each animal and expressed as a grade from 0 to 3, where 0 is <8 abscesses, 1 is 9 to 15 abscesses, 2 is 16 to 25 abscesses, and 3 is >25 abscesses. Mucus metaplasia and atrophy were graded from 0 to 3, corresponding to absent, mild, moderate, and severe.

Fluorescent detection of GalNAz, Muc5ac, Muc1, and H. pylori in fixed tissue.

Carnoy's fixed paraffin-embedded 5-μm sections were dewaxed, hydrated, and washed in PBS. Tissue sections were incubated with 40 μl of the reaction mix from the tetramethylrhodamine (TAMRA) glycoprotein detection kit (Invitrogen) and incubated at room temperature for 2 h. After washing with PBS, samples were blocked in 5% fetal bovine serum in PBS containing 0.05% Tween 20. Muc5ac was detected using the 45M1 antibody (1/1,000) (Sigma) and visualized by an anti-mouse Alexa Fluor 488 conjugate (1/100) (Invitrogen). Muc1 was detected using the CT2 antibody (1/200) (kind gift from Sandra Gendler, the Mayo Clinic, Scottsdale, AZ) followed by Donkey anti-Armenian hamster (1/200) (Jackson) and anti-donkey Alexa Fluor 488 (1/100) or Alexa Fluor 647 (1/100) (Invitrogen). For H. pylori detection, rabbit anti-H. pylori serum (1/1,000) (kindly provided by Thomas Borén, Umeå University, Sweden) was visualized using anti-rabbit Alexa Fluor 405 (1/100) (Invitrogen). Fetal bovine serum (5%) in PBS was used for dilution of antibodies. Sections were mounted using Prolong antifade mounting media (Invitrogen). A Zeiss LSM 510 META microscope was used for image acquisition and processing. Quantification of integrated fluorescence density was obtained using the Image J program, which is a public domain image-processing program (Wayne Rasband at the Research Services Branch of the National Institute of Mental Health, National Institutes of Health, Bethesda, MD). Briefly, the specimen of each mouse was studied by outlining the surface and neck cells at the beginning of the corpus and antrum. Care was taken to make sure it was always the same area measured, as we detected some variations in mucin turnover depending on tissue localization. The threshold for each stain was adjusted to the same level for all samples. The measurement was then set for area, integrated density (the total intensity of fluorescence in the defined area), and mean gray value.

Immunohistochemistry for Muc5ac.

Rodent decloaker (Biocare Medical) was used for antigen retrieval at 80°C for 2 h. Sections were then treated with 3% (vol/vol) hydrogen peroxide, washed, and blocked by rodent block M (Biocare Medical) for 30 min. The primary antibody (45M1) was diluted in antibody diluent (1:1,000; Dako) and incubated for 1 h. The sections then were incubated with MM HRP-Polymer (Biocare Medical) for 20 min. Bound antibody was visualized with diaminobenzidine, and the sections were counterstained with Harris' hematoxylin.

Statistical analysis.

All statistical analyses were performed using one-way analysis of variance (ANOVA) with Dunnett's post hoc test in the GraphPad Prism program (GraphPad Software).

RESULTS

Production rate and turnover of mucin in the noninfected murine stomach.

To obtain the basal production rate and turnover of mucin in the murine stomach, the azido GalNAc analog GalNAz was intraperitoneally injected, and samples were collected every hour for 12 h. Previous experiments have demonstrated efficient incorporation of GalNAz into mucin O-glycans in cell culture and in the mouse intestine during biosynthesis (16, 17). Incorporation of GalNAz was much more prominent in the surface epithelium than in the glands, therefore we focused the analysis on the surface epithelium. The images from gastric specimens gathered in our experiment demonstrated metabolic incorporation of GalNAz into newly synthesized glycoproteins from the first hour of injection. At this time point, the newly synthesized mucins were localized at the supranuclear part of the surface epithelial cells (Fig. 1A), in the area where the Golgi apparatus previously has been shown to be located. During the second and third hours, the mucins moved through the cells toward the cell surface (Fig. 1B). Five hours after injection, some newly synthesized mucins were detected on the cell surface, whereas the majority remained inside the cell (Fig. 1C). After 6 h, a few cells had secreted their newly synthesized mucins completely into the mucus layer, although the mucins could still be detected inside the majority of the cells. During the time points 6 to 12 h postinjection, the proportion of cells having released their newly synthesized mucins into the mucus layer increased with time (Fig. 1D, E, F, and G). However, some cells still had not released their newly synthesized mucins after 12 h.

Fig 1.

Confocal images of stomach sections from uninfected mice injected intraperitoneally with GalNAz. (A to F) The sections were collected every hour during a 12-h time course. A TAMRA-conjugated reagent was used for detecting GalNAz incorporated into newly synthesized mucin (visualized in the red channel). The 45M1 antibody was used to stain for Muc5ac (visualized in the green channel) and 4′,6-diamidino-2-phenylindole (DAPI) for nuclear DNA (visualized in the blue channel). Colocalization of Muc5ac and incorporated GalNAz (green and red) appears yellow in the images. (G) Schematic picture of the cellular localization and transfer of newly synthesized mucin. The rate of mucin transfer through the cells to the mucus layer varied slightly between cells, and in the schematic picture the average surface epithelial cell of noninfected mice is depicted.

Furthermore, our results demonstrate that after glycosylation, the membrane-bound mucin reached the glycocalyx (i.e., final destination) in less time than that required for the secretory mucins to become secreted into the mucus layer (Fig. 2A and B). Incorporation of GalNAz into the glycocalyx was first detected at 2 h (Fig. 2A and B), whereas the first weak traces of GalNAz in the secreted mucus layer appear after 5 h (Fig. 1C), and full colocalization of newly synthesized mucin with secreted Muc5ac was obvious at 10 h after GalNAz injection (Fig. 1E and G). In the fluorescent images, the colocalization of Muc5ac and newly synthesized mucin inside the cells is not as obvious in all pictures as the colocalization in the mucus layer is. This can be explained by the higher intensity of the fluorescence signal in the newly synthesized mucin, as well as a lower signal intensity from intracellular Muc5ac than that from Muc5ac in the mucus layer. Together, this makes it more difficult to simultaneously visualize Muc5ac and GalNAz inside the cells (Fig. 2E). However, comparing the staining in the red and green channels separately confirms colocalization of intracellular Muc5ac with newly synthesized mucin (Fig. 2C and D).

Fig 2.

GalNAz is incorporated into the glycocalyx 2 h after GalNAz injection. (A) Close-up confocal image of the surface mucus cells of a section from an uninfected mouse; red arrows denote GalNAz in the glycocalyx, which is visualized as red. White arrows point at the black holes where the nucleus is present (DAPI was not used to show the nucleus here, as the broad spectra from DAPI poses a risk for spillover of signal into the red channel). (B) Schematic picture of newly synthesized mucin in the glycocalyx and supranuclear area of gastric mucus-producing epithelium in noninfected mice. At this time point there was a clear gap between the GalNAz in the lower cytoplasmic area and the GalNAz in the glycocalyx. (C) Close-up confocal image of Muc5ac staining in a gastric section from an uninfected mouse, visualized as green. (D) Close-up confocal image of newly synthesized mucin in the same gastric section as that shown in panel C, visualized in red. (E) Immunostaining of Muc5ac (brown) in the antrum of a noninfected mouse. The line indicates secreted Muc5ac, and arrows denote intracellular Muc5ac. (F) Confocal images of H. pylori in an antral crypt from a mouse 7 days postinfection. H. pylori is visualized as white, Muc5ac as green. Arrows denote H. pylori.

H. pylori localization and inflammation in the murine stomach differs between early colonization and chronic infection.

Early H. pylori colonization (7 days postinfection) was associated with low gastritis scores, mild to moderate immune cell infiltration, and few abscesses. In contrast, during chronic infection (90 days postinfection), all mice showed moderate widespread and more severe multifocal infiltration, with a higher number of abscesses along with mild mucus metaplasia and atrophy (Fig. 3A and B). Immunofluorescence staining for H. pylori demonstrated the presence of this bacterium in the gastric lumen both during the early colonization and chronic infection of all mice and in crypts of the antrum of some mice during acute infection (Fig. 2F). There was no statistical difference in the number of H. pylori bacteria present in the mouse stomachs during early colonization compared to that during chronic infection (Fig. 3C).

Fig 3.

Gastritis scores and CFU counts from mouse stomachs infected with H. pylori. Gastric scores of antrum (A) and corpus (B) of mice during early colonization and chronic H. pylori infection (***, P < 0.001 compared to control by ANOVA with Dunnett's post hoc test). (C) H. pylori CFU counts from stomachs of infected mice (no significant difference; two-tailed t test).

Changes in mucin production during early colonization and chronic infection with H. pylori.

Based on the time line for mucin transport and dissemination obtained from the time course study of uninfected mice, we selected the time points 2, 6, and 12 h after GalNAz injection to compare the rates and turnover of gastric mucin production in noninfected mice to those of mice during early colonization or chronic H. pylori infection. In the antrum, the newly produced mucins displayed a slower movement from synthesis to secretion at the cell surface during infection. This effect was most pronounced during early colonization (Fig. 4 and 5); i.e., a high level of GalNAz was detected in the supranuclear region of the surface epithelial cells in noninfected and chronically infected mice at 2 h after GalNAz injection. In contrast, the amount of incorporated GalNAz was very low at this time point in mice during early colonization (Fig. 4A, D, and G). At the 6-h time point, the newly synthesized mucin was present inside the cells in noninfected mice (Fig. 4B) and already shed from the mucus layer at the 12-h time point (Fig. 4C). In contrast, mice with both early colonization and chronic infection had a lower intensity of the stained, incorporated GalNAz at 6 h postinjection (Fig. 4E and H), and plenty of newly synthesized mucin was still present inside the cells after 12 h, although some had been released (Fig. 4F and I). Thus, at all three time points the newly synthesized mucin had progressed further toward the mucus layer in uninfected mice than that in mice in early colonization. In addition, infection with H. pylori resulted in a reduction of the total amount of Muc1 (Fig. 6A). Similarly, there also seemed to be a trend toward a decrease in the total amount of Muc5ac, although this was not significant (Fig. 6C).

Fig 4.

Confocal images of antrum from uninfected and H. pylori-infected mice. Uninfected mice (A to C), mice during early infection (D to F), and chronic H. pylori infection (G to I) 2, 6, and 12 h, respectively, after intraperitoneal GalNAz injection. Newly synthesized mucin (incorporated GalNAz) is visualized as red, Muc5ac as green, and Muc1 as blue. Colocalization of green and blue results in gray, whereas colocalization of green and red results in yellow.

Fig 5.

Schematic picture of cellular localization and transfer of newly synthesized mucin during infection. Mucin transfer is depicted on an average surface epithelial cell of noninfected mice and compared to those of mice with early and chronic infection in antrum and corpus.

Fig 6.

Quantification of Muc1 and Muc5ac during early and chronic H. pylori infection. Comparison of the integrated density of fluorescence as a measure of Muc1 in antrum (A) and corpus (B), as well as Muc5ac in antrum (C) and corpus (D). Data were compared to the control (n = 6) by ANOVA with Dunnett's post hoc test (**, P < 0.01; NS, not significant).

We also studied the changes in corpus and duodenum of these mice. The results in the noninfected mice demonstrated that the synthesis and turnover of newly synthesized mucins are slower in the corpus than in the antrum (compare Fig. 7A to C to 4A to C). In the corpus, the rate of production and transfer of newly synthesized mucin from the perinuclear region toward the mucus layer was similar in noninfected mice and in mice during early colonization with H. pylori (Fig. 7A to F). This is in contrast to the antrum, where the mucin production rate and turnover were visibly decreased by early H. pylori colonization compared to that in noninfected mice (see representative images in Fig. 4 and the schematic picture in Fig. 5).

Fig 7.

Confocal images of corpus from uninfected and H. pylori-infected mice. Uninfected mice (A to C) and mice during early H. pylori colonization (D to F) and chronic H. pylori infection (G to I) 2, 6, and 12 h, respectively, after intraperitoneal GalNAz injection. Newly synthesized mucin (incorporated GalNAz) is visualized as red, Muc5ac as green, and Muc1 as blue. Colocalization of green and blue results in gray, whereas colocalization of green and red results in yellow.

Similar to the antrum, the total amount of Muc1 was reduced during both early colonization and chronic infection with H. pylori in corpus (Fig. 6B), whereas Muc5ac changes were less obvious between uninfected and infected groups (Fig. 6D).

We did not detect any changes in mucin production in the mouse duodenum during different time points of infection with H. pylori (data not shown).

DISCUSSION

In the present study, we followed in real time the process of mucin production in the healthy murine stomach, as well as in mice infected with H. pylori during early colonization and chronic infection, using metabolic labeling in a pulse experiment. In line with previous studies of the distal colon (18), our results from the stomach demonstrated a continuous production and secretion of mucins. GalNAz was already incorporated into mucins 1 h after injection and then moved through the cell to its final destination in the glycocalyx or mucus layer. We found that the rate of production of newly synthesized mucin was faster in the antrum of noninfected mice than in the corpus. Furthermore, the speed from the start of glycosylation for the membrane-bound mucin to reach the glycocalyx was faster than that for the secretory mucins to get secreted into the mucus layer. During H. pylori infection, mucin production, production rate, and turnover were hampered.

The results of our histological scoring showed that H. pylori produced only a mild inflammation in mice, although the severity of gastritis during chronic infection was slightly higher than that during early colonization with H. pylori. This finding was comparable those of to previous studies on the effect of this bacterium on the murine stomach (19). Despite the mild gastritis caused by this bacterium, we detected a reduction in the rate of mucin production and turnover and a decrease in the level of Muc1 but no significant changes in the Muc5ac level. Together, these results show that there are no major changes in the total amount of mucin in the gastric tissue and the mucus layer, as the amount of Muc5ac is vastly higher than the amount of Muc1. However, the production and secretion rate of the mucin is decreased, and the rate at which the secreted mucus washes harmful agents away from the epithelial surface is decreased. H. pylori-infected rhesus monkeys and children (aged 3 to 18 years) secreting mucins with weaker H. pylori binding capacity develop infections with higher H. pylori density and more severe gastritis (13, 20). Patients with primary Sjogren's syndrome are known to produce fewer mucins and are also reported to present a more severe H. pylori-associated pathology (21). These results indicate the ability of secreted mucins to bind to and transport away H. pylori as a mechanism for protecting the gastric epithelium. Thus, the impaired speed of the washing of the mucosal surface caused by the slower mucin secretion rate is likely to provide H. pylori with a more stable environment. The presence of H. pylori inside the crypts of mice during acute infection coincides with the time when we see the most severe slowdown of mucin turnover. Thus, slowing down the mucus flow may enable close association of H. pylori with the epithelial cells and aid in the colonization process. Later in the infection, the mucus flow appears to be less hampered, and then we detect H. pylori mainly in the lumen.

The reduced production of Muc1 by the host is also likely to be beneficial to the bacterium, as mice lacking Muc1 are more susceptible to infection by H. pylori (22). H. pylori binding to gastric epithelial cells is inhibited by MUC1, irrespective of whether or not they bind to this mucin. Strains that bind to MUC1 are removed from the epithelial cell surface as a result of the protein acting as a releasable decoy (13). Strains that would otherwise bind to alternative cell surface ligands are inhibited by steric hindrance caused by MUC1.

Our results showed no significant changes in the amount of Muc5ac between infected and noninfected mice. Similarly, most immunohistochemical studies on human biopsy specimens from H. pylori-infected individuals detect no difference in MUC5AC expression (23). One study, however, demonstrated a reduction of MUC5AC-producing cells in H. pylori-infected individuals (24).

A previous study using the gastric cancer cell line KATO-III infected with H. pylori in vitro found that mucin synthesis is decreased upon infection, supporting our in vivo results. However, in their study they found no change in mucin secretion (25). This difference in results relating to the secreted mucins compared to ours can be explained by many experimental factors. The in vitro experiments were performed with at least a 1,000-fold higher bacterial/mammalian cell ratio than our in vivo experiments. Furthermore, the in vitro experiment was performed under conditions that are incompatible with H. pylori survival. However, the factor that most likely has the most importance with regard to measurement of secreted mucins is that the in vitro experiment was short term (22 h). The detected decrease in mucin synthesis may have resulted in a decrease in mucin secretion at later time points, similar to our in vivo experiments. Byrd et al. also found a transient decrease in MUC1 that was recovered by the 22-h time point. However, the bacteria in their study were shown to be nonviable. Therefore, the reduction in MUC1 may depend on the viability of H. pylori. We have reported in a previous study that MUC1 is shed and coats H. pylori, which leads to a decrease in MUC1 levels at a similar time point using viable bacteria (14).

In patients with chronic gastritis, depletion of the MUC1 extracellular subunit, but not the transmembrane subunit, has been reported (26), which is compatible with MUC1 acting as a releasable decoy (14). Thus, the current study is the first to show downregulation of the cytoplasmic domain of Muc1 in vivo. It is interesting that the in vivo model of H. pylori inhibition of mucin synthesis (used in this study) has far greater sensitivity than the in vitro model (24). The effects on mucin production and secretion achieved in vitro are similar to (or smaller than) the ones we detect in vivo, although the in vitro model uses at least a 1,000-fold higher concentration of H. pylori. As the effect on mucin synthesis in vitro is dependent on the concentration of H. pylori (25), factors such as viability of the bacteria and/or the host immune system are likely to be an important part of H. pylori-dependent mucin inhibition in vivo.

In conclusion, H. pylori colonization in the mucus niche of the murine stomach in vivo leads to decreased mucin production and secretion rate and decreased levels of Muc1 in the mucosa. This indicates disruption in the mucus defense mechanism of clearing the mucosal surface from pathogens by impairing the mucus flow.