The Helicobacter pylori infection alters the intercellular junctions on the pancreas of gerbils (Meriones unguiculatus)

Helicobacter pylori is a common resident in the stomach of at least half of the world’s population and recent evidence suggest its emergence in other organs such as the pancreas. In this organ, the presence of H. pylori DNA has been reported in cats, although the functional implications remain unknown. In this work, we determined distinct features related to the H. pylori manifestation in pancreas in a rodent model, in order to analyse its functional and structural effect. Gerbils inoculated with H. pylori exhibited the presence of this bacterium, as revealed by the expression of some virulence factors, as CagA and OMPs in stomach and pancreas, and confirmed by urease activity, bacterial culture, PCR and immunofluorescence assays. Non-apparent morphological changes were observed in pancreatic tissue of infected animals; however, delocalization of intercellular junction proteins (claudin-1, claudin-4, occludin, ZO-1, E-cadherin, β-catenin, desmoglein-2 and desmoplakin I/II) and rearrangement of the actin-cytoskeleton were exhibited. This structural damage was consistent with alterations in the distribution of insulin and glucagon, and a systemic inflammation, event demonstrated by elevated IL-8 levels. Overall, these findings indicate that H. pylori can reach the pancreas, possibly affecting its function and contributing to the development of pancreatic diseases. Graphical Abstract Supplementary Information The online version contains supplementary material available at 10.1007/s11274-024-04081-0.


Introduction
Helicobacter pylori is a Gram-negative bacterium that infects at least half of the world's population (Malfertheiner et al. 2023).Actually, it is the most common infectious pathogen of the gastroduodenal tract.Different factors, such as geographical region, host age, and human socioeconomic and hygienic aspects influence the prevalence or incidence of H. pylori infection (Hooi et al. 2017).The infection is acquired by oral ingestion of contaminated food, and mainly transmitted within families in early childhood (Suerbaum and Michetti 2002).After being ingested, H. pylori evades the bactericidal activity of the gastric luminal content and enters the mucous layer.There, the bacterial urease hydrolyses urea into carbon dioxide and ammonia, allowing H. pylori to survive in the acidic milieu (Mobley 2001).In addition, motility is essential for colonization, thereby, the presence of flagella contributes to bacterial adaptation to the gastric niche (Lertsethtakarn et al. 2011).Most H. pylori strains possess the cag pathogenicity island (cag-PAI), a 37-kb genomic region encompassing 29 genes.Many of them encode components of the type IV secretion system that translocates the CagA protein into the host cell cytoplasm.When CagA enters the epithelial cell, it is phosphorylated and attaches to the SHP-2 tyrosine phosphatase, which causes the host cell to produce growth factor-like cellular response and cytokines (Suerbaum and Michetti 2002).Other molecules also contribute to pathogenicity, including adhesins (e.g., SabA and BabA), outer membrane proteins (OMPs), serine proteases (HtrA) and the vacuolating cytotoxin A (VacA) (Matsuo et al. 2017;Chauhan et al. 2019;Doohan et al. 2021).
H. pylori produces continuous gastric inflammation in all infected people (Suerbaum and Michetti 2002).Chronic gastritis strongly correlates with the risk of clinical sequelae, such as mucosal atrophy, gastric or duodenal ulcers, or carcinoma and lymphoma in stomach (Suerbaum and Michetti 2002).In addition to colonizing the stomach, evidences suggest the presence of H. pylori in other organs such as the intestine, liver, glad bladder, coronary arteries, and pancreas (Nilsson et al. 2006;Testerman and Morris 2014;Shojaee Tabrizi et al. 2015).The pancreatic diseases associated with H. pylori infection are acute, chronic, and autoimmune pancreatitis, pancreatic cancer, and diabetes mellitus (DM) (Manes et al. 2003;Nilsson et al. 2006;Rieder et al. 2007;Rabelo-Gonçalves et al. 2015).The majority data about H. pylori and its potential participation in the development of pancreatic diseases refers to autoimmune forms of chronic pancreatitis and pancreatic adenocarcinoma.Evidence indicates a modest increased pancreatic cancer risk in infected individuals (Kunovsky et al. 2021).Systematic reviews and meta-analyses have linked H. pylori infection with insulin resistance, DM, neglected glycaemic control and diabetic complications, as well as other components of metabolic syndrome and non-alcoholic fatty liver disease (Polyzos et al. 2021).It has been proposed that the increase of gastrointestinal permeability, caused by H. pylori, contributes to systemic malfunction and the development of diabetes (De Kort et al. 2011).Moreover, Song et al. (2021) evaluated the H. pylori eradication rate between patients with and without DM, demonstrating an approximately double pooled odds ratios (OR) of unsuccessful eradication in the former than the latter (OR: 2.08; 95% CI: 1.56e2.77)(Song et al. 2021).
Although many animal models have been used for the analysis of H. pylori infection, such as mice, rats, cats, pigs, and gerbils (Honda et al. 1998;Mendoza-Elizalde et al. 2016), the bacterial DNA presence has been reported in the pancreas of cats, although the functional implications in this tissue are still unknown (Shojaee Tabrizi et al. 2015).Since gerbil is a robust, efficient, and cost-effective model that recapitulates several features of gastric inflammation and carcinogenesis in humans induced by H. pylori, the objective of this work was to determine the H. pylori manifestation in pancreas, in order to analyse its effect on this epithelial tissue.The H. pylori infection in gerbils, was confirmed by the bacteria presence in gastric tissue and elevated IL-8 levels at serum.This is the first report that proved the H. pylori occurrence in pancreas by different methodology approaches, such as urease activity, bacterial culture, and PCR (glmM and cagA genes) and immunofluorescence (α-Hp, α-CagA and α-OMPs antibodies) assays.In pancreas, this bacterium produced alterations in tight junction, adherens junction and desmosome structures, as observed by the redistribution of claudin-1, claudin-4, occludin, ZO-1, E-cadherin, β-catenin, desmoglein-2 and desmoplakin I/II proteins.Furthermore, the actin-cytoskeleton was also disturbed, as reported in gastric tissue.These structural modifications produced changes in the insulin and glucagon localization, suggesting that H. pylori infection could affect the pancreas functions.

H. pylori culture
Two H. pylori strains were used in this work: (1) the reference strain 26695 obtained from ATCC®; and (2) a clinical isolate 279-a derived from a 12 years old male patient with gastroesophageal reflux disease and dyspepsia, obtained and characterized in the Infectious Diseases Laboratory of the Federico Gómez Children's Hospital of Mexico.Bacilli of the clinical isolate were highly flagellated and expressed CagA+, VacA s1m1, BabA 2+, and GlmM proteins.Bacteria were grown in Casman base medium (DIBICO) supplemented with 5% defibrinated sheep blood (HEMO-PROVEEDORES) and antibiotics (3 mg/ml vancomycin, 5 mg/ml trimethoprim and 2 mg/ml amphotericin B; Inv-itrogen™), and incubated at 37 °C with 5% CO2 and 10% humidity.

Ethics statement
The Children's Hospital of Mexico Federico Gomez fulfilled the standard of the Mexican Official Norm (NOM-062-ZOO-1999).The Ethics, Biosafety and Scientific Committees at the Health Institute, as the regulatory office for the approval of research protocols involving the use of laboratory animals and, in fulfilment of the NOM, reviewed and approved all animal experiments.This work was developed under the HIM/2011/080 protocol SSA.1005 and HIM/2023/017.SSA.The study is reported in accordance with ARRIVE guidelines.
In addition, all methods were performed in accordance with the relevant guidelines and regulations.

Tissue procurement
The animal euthanasia was carried out 12 months after infection by exsanguination under deep sedation, using ketamine/ xylazine (PISA), to obtain pancreas, stomach and blood.The blood was stored at − 70 °C for later use.The pancreas and stomach were washed with PBS (137 mM NaCl, 10 mM Na 2 HPO 4 , 1.8 mM KH 2 PO 4 , 2.7 mM KCl, pH 7.4) and divided in three parts for following assays.

Bacterial culture from tissue
The pancreatic and gastric tissues were disintegrated in PBS.Later an inoculum was taken to sow in Casman agar culture dishes, supplemented with 5% defibrinated sheep blood and antibiotics as above.Culture dishes were incubated at 37 °C with 5% CO2 and 10% humidity, to allow the bacterial growth during 7-14 days.H. pylori-positive colonies were morphologically identified by their shape as small, bright, and colourless, and then selected by Gram staining.Finally, the colonies were molecularly characterized by PCR assays.

Bacterial urease activity in animal tissue
To evaluate the presence of bacterial urease in stomach and pancreas, tissues were disintegrated in PBS, and samples were inoculated in Christensen's medium (Mendoza-elizalde et al. 2016).Bacteria hydrolyse urea through the enzyme urease, releasing ammonia and carbon dioxide.These products alkalinize the culture medium by turning the phenol red indicator from yellow to red.

PCR from animal tissue
To determine the presence of H. pylori DNA in stomach and pancreas, tissues were lysed and DNA extraction was performed using the Wizard® genomic DNA purification kit (PROMEGA), following the manufacturer's instructions.DNA amount and integrity were measured in the EPOCH microplate reader (BioTEK) and assessed by 1% agarose gel electrophoresis, respectively.The PCR assays were performed using specific oligonucleotides for the bacterial glmM (forward 5ʹ-AAG CTT TTA GGG GTG TTA GGG GTT T-3ʹ and reverse 5ʹ-AAG CTT ACT TTC TAA CAC TAA CGC -3ʹ) (Smith 2004) and cagA genes (forward 5ʹ-AAT ACA CCA ACG CCT CCA AG-3ʹ and reverse 5ʹ-TTG TTG CCG CTT TTG CTT CC-3ʹ) (Lage et al. 1995) and the PCR Master Mix polymerase kit (PROMEGA), in a thermocycler T100 Thermal-Cycler (BioRad).PCR products were separated in a 1% agarose gel and stained with Midori Green Direct (NIPPON Genetics EUROPE).Bands were visualized in an iBright FL 1500 imaging system (Invitrogen).DNA from both H. pylori strains (26,695 and 279-a) were used as positive controls.In negative controls the same PCR mix and oligonucleotides were used, but without DNA template.

IL-8 measuring
From the sera previously obtained, the IL-8 measurement was carried out with the help of the IL-8/cxcl15 ELISA kit (MyBioSuource), following the manufacturer's instructions.96 well plates coated with 100 μl of IL-8 standards or test samples (diluted at 1/4 with sample dilution buffer) were incubated at 37 °C for 90 min.Control (blank) wells contained only sample dilution buffer.Plates were washed twice with washing buffer and incubated for 60 min at 37 °C with 100 μl biotin-labelled antibody (1:100).After three washes, 100 μl of HRP-streptavidin conjugate (1:100) was added and plates were incubated 30 min at 37 °C.Five more washes were carried out and 90 μl of TMB substrate were added into the wells, incubating in dark 10-20 min at 37 °C.The stop solution (50 μl) was added, and immediately the plates were measured at λ = 450 nm in the EPOCH microplate reader (BioTEK).All experiments were performed by duplicate.

Statistical analysis
Results displayed in this work represent the mean and standard error.Statistical analyses were performed by Chi-square for trend, ANOVA and t-Student tests, using GraphPad Prism 6.0 software.Statistically significant differences are designated with asterisks in the table and graphs ( * p < 0.05, * * p < 0.01 or ***p < 0.001).

H. pylori is present in the pancreas, but does not produce evident morphological changes
The effect of H. pylori on the gastric epithelium has been widely studied, evidencing clear disturbance of the barrier integrity.The opening of tight junctions and erosion of gastric epithelium could allow the bacterial entrance to blood stream and other tissues (Alzahrani et al. 2014).Recent findings seem to correlate the H. pylori seropositivity with pancreatic diseases (Bulajic et al. 2014); however, there are scarce studies indicating the presence of bacterial molecules in pancreas, and even less evidence exists for the presence of the complete bacteria in this tissue.Here, we developed a rodent model of infection based on gerbils, and to promote the H. pylori invasion to other tissues, we treated experimentation animals with ethanol in order to produce ulcers, as previously demonstrated (Ahmed et al. 2013).Thus, four animal groups were employed: (1) control, (2) EtOH-treated, (3) Hp-inoculated, and (4) EtOH-treated plus Hp-inoculated; and experiments were performed after 12 months of treatment.Firstly, we evaluated whether animals were successfully infected by H. pylori through the urease test, inoculating a portion of gastric or pancreatic tissue in urea-agar for bacterial culture.Gastric tissue of animals from groups 3 and 4 presented around 77% positivity to the urease test (Fig. 1A).Of note, animals not inoculated also presented positivity for this test.Nevertheless, the percentages of urease activity in groups 3 and 4 are higher, in comparison to the control and EtOH-treated animals (50 and 60% ureasepositive, respectively).In pancreatic tissue, groups 3 and 4 were 90% positive to the urease test, relative to groups 1 and 2, which exhibited 50% of positivity (Fig. 1A).
To confirm the infection, a portion of gastric and pancreatic tissues was inoculated in Casman agar supplemented with sheep's blood and selective antibiotics (vancomycin, trimethoprim, and amphotericin B) (Mendoza-Elizalde et al. 2016) for H. pylori growth.Then, an alike H. pylori colony (small, bright, and colourless) was reseeded and morphologically characterized by Gram staining.The colonies with similar morphology from originally inoculated strains (26,695 and 279-a), came from the stomachs of 3 and 4 animals and from the pancreas of 1 and 2 gerbils, of groups 3 and 4, respectively.Later, genomic DNA derived only from these bacterial colonies was obtained.PCR assays were performed and representative genes from H. pylori, as glmM (Espinoza et al. 2011) (encoding for a phosphoglucosamine mutase) and cagA (encoding for the CagA virulence factor)  (Jeyamani et al. 2018) were detected.By using this DNA as template, and specific oligonucleotides, amplicons of glmM (240 bp) and cagA (400 bp) genes were revealed only in groups 3 and 4, but not in control and EtOH-treated animals (Fig. 1B, C).Of note, both genes were present in pancreas of Hp-inoculated animals.DNA derived from bacterial lysates from strains 26,695 and 279-a were used as template in positive controls.The results obtained from urease test, bacterial cultures only with Hp-features and PCR assays were concentrated in Table 1.
Importantly, not all Hp-inoculated animals from groups 3 and 4 were infected, thus, we considered as animals infected, those presenting the glmM and cagA genes in the gastric tissue.In addition, in the pancreas of some of these animals, these genes were also amplified.Then, 3/9 and 4/9 gerbils from groups 3 and 4, respectively, were infected by H. pylori.In following experiments, only the infected animals were employed (at least 3 animals per group).
Next, we analysed the morphological damage produced in gastric and pancreatic tissues by H. pylori, through haematoxylin-eosin staining.In the control and EtOH-treated groups, no significative changes were found in both studied tissues.In stomach of animals from group 2, only few calcifications were detected (Fig. 2A).In pancreas, some ductal plugging and empty zymogen granules in the acini were observed, similar to those found in control animals (Fig. 2B).In the Hp-inoculated animals, the gastric sections revealed lymphoid accumulation, tissue regeneration and  presence of neutrophils (Fig. 2A).Meanwhile, the pancreas images were similar to control and EtOH-treated groups (Fig. 2B).The gastric preparations from EtOH-treated plus Hp-inoculated animals (group 4) showed atrophy, tissue regeneration and superficial gastritis (Fig. 2A); while pancreatic samples look similar to other groups and no alterations were detected (Fig. 2B).All these alterations at the stomach and pancreas were exhibited only by infected animals from the Hp-inoculated groups.These findings suggest that H. pylori reaches the pancreas and maintains its proliferation ability, as exhibited when infected pancreatic tissue was homogenized and bacteria was cultured in agar medium, although evident morphological changes were not revealed.

Helicobacter pylori bacteria and some of their proteins are present in pancreas
To confirm the presence of the H. pylori bacteria in pancreas, immunofluorescence (IF) assays were performed using an anti-Hp antibody.The bacteria were only observed outside of the cells in animals of groups 3 and 4, and no specific fluorescence signal was detected in control and EtOH-treated gerbils (Fig. 3A).The fluorescence intensity was quantified, and it resulted elevated in Hp-inoculated animals, and even higher in group 4, suggesting that the ulcers produced by EtOH promoted the bacterial access beyond the gastric epithelium (Fig. 3B).
In addition, we also investigated whether some bacterial proteins were localized in pancreas.By IF experiments, CagA (Fig. 4A) and OMPs (Fig. 4B) proteins from H. pylori were mainly detected inside of the pancreatic cells of Hpinoculated animals from groups 3 and 4. In control and EtOH-treated gerbils, no specific fluorescence signal was observed (Fig. 4A, B).The quantification of fluorescence intensity confirmed the presence of both bacterial proteins only in Hp-infected animals (Fig. 4C, D).As control, the gastric epithelium was analysed by IF assays for the localization of H. pylori and CagA and OMP proteins, and accordingly, only infected animals showed signals (Figs. S1 and  S2).
These results corroborate that H. pylori is able to reach the pancreas and produce there some virulence factors, which are in some way internalized.

Helicobacter pylori disturbs intercellular junction proteins in pancreas
The functional state of pancreas is regulated by several adhesion structures, including tight and adherens junctions and desmosomes (Jimenez-Caliani et al. 2017;Myo Min et al. 2022), which are essential for homeostasis, and to preserve the normal epithelial structure and regulate secretion processes of this tissue (Sato et al. 2019).Preserving the normal function of junctional molecules and avoiding their abnormal distribution, the pancreas structure is maintained, and the development of diseases is prevented.Therefore, we studied how the H. pylori infection affects the distribution of some adhesion molecules.By IF assays and using specific antibodies, we analysed the localization changes in tight junction proteins as claudin-1, claudin-4, occludin, and ZO-1; in adherens junction molecules as E-cadherin and β-catenin; and in desmosomes proteins as desmoglein-2 and desmoplakin I/II.The localization of claudin-1, mainly distributed in the cellular borders in control animals, changed in EtOH-treated gerbils, diminishing particularly from plasma membrane (Fig. 5A).While, in the Hp-inoculated group, this protein was internalized toward the cytoplasm and disappeared from cellular borders.In animals infected with H. pylori and EtOH-treated, a dramatic reduction of claudin-1 was observed (Fig. 5A).Similar results were displayed for claudin-4, and a decreasing trend was observed for occludin and ZO-1 (Fig. S3).The fluorescence intensity was quantified and significant differences among treatments were displayed (Fig. 6).In gastric tissue, a comparable redistribution of these proteins was observed, having the group 4 the more dramatic effect (Figs. S4-S6).Delocalization and decrease of E-cadherin and desmoglein-2 were observed in pancreas from groups 2, 3 and 4, with respect to the control animals (Fig. 5B,  C).Results were confirmed by fluorescence quantification, and a similar behaviour was revealed for β-catenin and desmoplakin I/II (Figs. 6 and S3).In stomach, analogous patterns were detected (Figs. S4-S6).
Dynamic rearrangements of the actin cytoskeleton constitute a hallmark of H. pylori infected gastric cells that conducts to invasive growth toward other tissues.Therefore, we analysed the actin distribution in pancreas, observing a main localization at cellular border of endothelial cells and rounding acini in control animals (Fig. 5D).In treated gerbils (groups 2-4), an important reduction was detected when actin-fluorescence intensity was quantified (Fig. 6).It suggests that as in stomach (Figs.S4 and S6), this bacterium is also producing actin cytoskeleton remodelling in pancreas, which probably affects the intercellular junctions and functions of this organ, such as disturbing the hormone secretion.

H. pylori infection alters hormones distribution at pancreas
The endocrine function of pancreas is to regulate the blood glucose levels through the secretion of two main hormones, insulin and glucagon (Röder et al. 2016).In this work, we studied the effect of H. pylori infection on the distribution of both hormones.The EtOH-treatment produced a significant re-localization of glucagon toward the cytoplasm, whereas the infection in groups 3 and 4, significantly diminished the amount of this protein in pancreatic tissue (Fig. 7A, C).Insulin was also reduced in pancreas by the infection of H. pylori, and only a redistribution was detected by the treatment with ethanol (Fig. 7B, D).

IL-8 levels are increased in gerbils infected with H. pylori
Persistent H. pylori infection confers an increased risk for gastric diseases involving various cytokines modifications related to the inflammatory immune response.The most characterized pro-inflammatory cytokine production by this bacterial infection is IL-8.Here, we evaluated IL-8 levels in sera from different gerbil groups.In control and ulcerated animals, the cytokine concentration remained unchanged; but in Hp-infected gerbils (groups 3 and 4) we observed a significant increased level (more than twice regarding the control group) of IL-8 (Fig. 8).These findings suggest that H. pylori infection promotes a mechanism for the heightened inflammatory response, as reported in human and other animal models (Sharma et al. 1998;Zhao et al. 2015).
In general, our data indicate that gerbils are susceptible to H. pylori infection and, in some cases, it was promoted by ulcers production.This bacterium penetrates the stomach and can reaches the pancreas, where it internalizes some virulence molecules such as CagA and OMPs.Even when no evident morphological damage was produced in this organ, H. pylori is able to impair the intercellular junction proteins from the pancreatic epithelium, possibly affecting the endocrine function, as suggested by alterations in glucagon and insulin distribution in infected animals.The tissue damage

Discussion
Most of studies related to H. pylori to date, are focused in the gastric epithelium injury produced by this bacterium, nevertheless, its extragastric presence is gaining relevance recently, due to the relationship with several diseases in distinct organs and tissues (Gravina et al. 2020;Kunovsky et al. 2021).Particularly, in pancreas, H. pylori infection has been associated to pancreatitis, diabetes and cancer (Rabelo-Gonçalves et al. 2015).Nevertheless, the mechanisms to achieve this organ and the effect in the pancreatic epithelium have not been elucidated yet.Therefore, in this study, we used an infection model in gerbils to probe the presence of H. pylori in the pancreas.There, the bacterial infection altered proteins of the intercellular junctions and the localization of some hormones, and increased levels of IL-8 in serum.
H. pylori infects humans, but several animal models are currently used to elucidate the mechanisms of infection.However, in experimentation models it is difficult to obtain efficient rates of infection, due to the complex gastric epithelium response that, just in human, could take up to decades to develop symptoms and gastric diseases.Recently, gerbils have been used as models for the study of this infection, since they accurately resemble gastric inflammation and carcinogenesis produced by H. pylori in humans; besides, it is a competent, strong, and low-cost rodent model (Ansari and Yamaoka 2022).In Mongolian gerbils, H. pylori colonizes the gastric mucosa, producing inflammatory infiltrates in the lamina propria with the presence of neutrophils and mononuclear leukocytes (Di et al. 2020); however, alterations in other organs have not been studied yet.In pancreas of cats, H. pylori has been detected by PCR (Shojaee Tabrizi et al. 2015) and in human, this bacterium has been related with pancreatitis and pancreatic cancer (Manes et al. 2003;Rieder et al. 2007).The pathogenesis and evolution of idiopathic forms of pancreatitis, is associated with changes in the secretion of the exocrine pancreas due to H. pylori infection.Furthermore, the altered epithelial gastric barrier allows the colonization of other pathogens, promoting that pancreatitis becomes chronic, and eventually, contributing to cancer development (Nilsson et al. 2006;Rieder et al. 2007;Dore et al. 2008;Bulajic et al. 2014;Rabelo-Gonçalves et al. 2015;Kunovsky et al. 2021).
Other species of Helicobacter, such as Helicobacter hepaticus, cause chronic hepatitis and hepatocellular carcinoma (HCC) in mice; and Helicobacter spp.have been identified in the liver of patients with cholestatic diseases and in HCC derived from non-cirrhotic liver (Rocha et al. 2005).According to some theories, Helicobacter spp.could go from the stomach to the liver via the duodenum and biliary tract or could enter the liver from the bloodstream via the hepatic portal vein (Pellicano et al. 2008).
In this work, we generated a gerbil model for H. pylori infection during 12 months; in addition, to promote the bacterial access to the organism, ulcers were produced by EtOH-treatment.This treatment has already been employed in other rodent models as rats, causing ulcers successfully, as observed by H&E staining (Ahmed et al. 2013).The presence of H. pylori in stomach and pancreas was demonstrated here by urease test, bacterial cultures derived from these tissues and PCR assays.Hp-inoculated animals presented more urease activity than control and EtOH-treated gerbils.In stomach, the enzymatic activity present in the latter groups could be due to the enzymes expressed by proteobacteria of the genus Proteus spp. or others (Heimesaat et al. 2014).Whilst, in pancreas, the positivity of the test is partially produced by the components of this tissue, such as digestive enzymes and sodium bicarbonate, which eventually protects the duodenum by neutralizing the acid that comes from the stomach (Pandol 2017).These components make the urea-agar mildly alkaline, hence, giving a positive result.Therefore, the urease test is not adequate to determine with certainty the H. pylori infection.Then, to specifically determine the presence of H. pylori at pancreas, bacterial cultures were performed by using tissues from stomach and pancreas, and employing antibiotics (vancomycin, trimethoprim and amphotericin B) (Mendoza-Elizalde et al. 2016) for a selective H. pylori growth.Only colonies with similar morphology to initially inoculated strains were selected and came from stomach (3 and 4 animals) and pancreas (1 and 2 gerbils) of groups 3 and 4, respectively.From these colonies, bacterial DNA was isolated, and the glmM and cagA genes were PCR-amplified.Hence, the rate of infection was 33 and 44% for groups 3 (Hp-inoculated) and 4 (Hp-inoculated + EtOH-treatment), respectively.Albeit the EtOHtreatment favoured the bacterial infection, the percentage of infection obtained was still lower than those obtained in previous studies (Fig. 1) (Velazquez-Guadarrama et al. 2007;Cortés-Márquez et al. 2018).The pronounced influence of genetic diversity, particularly, regarding to the immune system in the infection models, could explain the differential response to the bacterial infection.Therefore, it seems necessary to increase the time of infection until 18 months and the number of animals per group.
In order to corroborate if H. pylori or some of its virulence factors reach the pancreas, we used specific antibodies against H. pylori, CagA and OMP's.By immunofluorescence experiments, signals for H. pylori, CagA and OMPs were only detected in infected groups (3 and 4); while in groups 1 and 2, no fluorescence was distinguished.These findings suggest that H. pylori and some bacterial proteins are achieving the pancreas, displaying a pattern localization similar to that described in an infected stomach (Fig. 4).Alternatively, these proteins could arrive to the pancreas through outer membrane vesicles (OMVs) derived from bacteria or by exosomes produced from infected gastric cells (Chen et al. 2018;Jarzab et al. 2020).
Morphologically, infected animals showed lesions in the gastric epithelium, such as regeneration, lymphoid accumulation, and superficial gastritis, that could be attributed to the presence of H. pylori.In other studies, during prolonged infections, severe inflammation results in the loss of parietal and chief cells, usually accompanied by hyperplasia of the mucous cells of the neck (Ohkusa et al. 2003;Ansari and Yamaoka 2022).Nevertheless, in our model, no apparent morphological injuries were observed in the pancreas of infected gerbils.In all groups, empty acini and plugging in some ducts were detected, as reported in the pancreas of normal and diabetic gerbils (Fig. 2) (Li et al. 2016).
Taking into account that histologically there were no changes in the pancreatic tissue, we reviewed the epithelial structure by analysing the localization of proteins from the intercellular junctions.Proteins from the tight junction (claudin-1, claudin-4, occludin and ZO-1), adherens junction (E-cadherin and β-catenin) and desmosomes (desmoglein-2 and desmoplakin I/II) were delocalized from the cellular borders towards the cytoplasm, mainly in the groups 3 and 4, with respect to group 1.These localization changes correlated with a reduction in the amount of these proteins, as revealed by fluorescence quantification, suggesting that the expression of intercellular molecules is affected by H. pylori infection, as occurs in the stomach (Costa et al. 2013) (Fig. 6).
In the gastric mucosa the OMPs expression helps H. pylori to attach to gastric epithelial cells at the primary stage of infection and rises the virulence of this bacterium.OMPs also cooperate with other virulence factors such as CagA and VacA to increase the release of inflammatory factors, neutrophil infiltration, and helping to the colonization, persistent infection, and severe clinical consequences (Xu et al. 2020).Besides, H. pylori internalizes CagA through its T4SS system, which can be also secreted together with other virulence factors.In the epithelial cells, this oncoprotein induces the delocalization of intercellular junction proteins such as ZO-1, E-cadherin, and β-catenin, which leads to the epithelial barrier impairment (Alzahrani et al. 2014).In epithelial cells monolayers like MDCK, CagA causes polarity defects characterized by alterations of ZO-1 and E-cadherin, in a PAR1b-depending manner (Takahashi-Kanemitsu et al. 2020).Moreover, CagA has the ability to physically interact with the cytoplasmic domain of E-cadherin.The CagA/Ecadherin interaction interferes with and destabilizes the formation of the E-cadherin/ β-catenin complex and abnormally re-localizes the membrane-bound portion of β-catenin to the nucleus, where it activates Wnt-target genes in a β-catenin/ TCF-dependent manner (Takahashi-Kanemitsu et al. 2020).Otherwise, the H. pylori serine protease HtrA, opens cell-tocell junctions through cleavage of occludin, claudin-8, and E-cadherin, thus inducing the disintegration of their epithelial barrier functions (Tegtmeyer et al. 2017).In gastric epithelial cells, such as NCI-N87 and MKN28, HtrA also cuts the desmosomal cadherin, desmoglein-2.Hence, tight junctions, adherent junctions, and desmosomes are targets of this serine protease (Bernegger et al. 2021).Thus, the HtrA activity is necessary for paracellular transmigration of H. pylori across polarized cell monolayers to reach basolateral membranes and the CagA translocation across ɑ5β1 integrin (Tegtmeyer et al. 2017).
In this context, it is possible that H. pylori pathogenicity factors such as urease, OMPs, CagA, VacA and HtrA, could induce host cell signalling involved in altering cellto-cell permeability, to impair the gastric epithelial barrier and then, the bacteria or these factors can be internalized and reach deeper tissues.In pancreas, the effect of these virulence factors over junctions is similar to that described in stomach, as we observed in this work.The breakdown of the pancreatic ductal barrier is known to contribute to the pathophysiology of pancreatitis and the development of pancreatic cancer because tight junctions in the pancreas are crucial regulators of physiologic secretion (Rieder et al. 2007;Kojima et al. 2013).Various inflammatory mediators and carcinogens can trigger tight junction disassembly and disruption of the pancreatic barrier, however, signalling events involved remain poorly understood.Furthermore, in pancreas, the adhesion molecules are crucial for proliferation, cell migration, and signal transduction, as well as in the development and tissue repair.When the cell-cell adhesion between endothelium and/or pancreatic acinar cells weakens, the accompanying interstitial oedema encourages the migration of inflammatory cells and disturbs the integrity of the tissue.In pancreatitis, occludin, claudin-1 and ZO-1 are decreased, but no changes in claudin-4 have been reported; whereas E-cadherin and β-catenin are dissociated from the plasma membrane and condensed in the cytosol of acinar cells (Sato et al. 2019).Moreover, E-cadherin is important for maintaining the architecture and homeostasis of the exocrine part and its absence contributes in the development of pathogenic conditions, such as pancreatitis or pancreatic cancer (Serrill et al. 2018).In the pancreas of the mouse model, the loss of desmoplakin expression resulted in the disruption of desmosomal adhesions, that can promote increased local tumour invasion, independently of the adherens junction status (Chun and Hanahan 2010).Altogether, these findings could explain why in this work we observed that H. pylori infection produced changes in all these proteins, which were re-localized from the plasma membrane towards the cytoplasm, and also significantly diminished (Fig. 6).
Epithelial structure is also maintained by the cytoskeleton and dynamic rearrangements of actin; but in H. pylori infected gastric epithelial cells, important changes in actin lead to the development of aberrant morphological changes, cell migration and invasive growth (Wessler et al. 2011).Translocated CagA during H. pylori infection alters SHP-2 (SH2-containing tyrosine phosphatase 2), Crk (C-terminal Src kinase), Grb2 (growth factor receptor-bound protein 2), and MARK2 (microtubule affinity-regulating kinase 2), which dysregulate key cellular biochemical pathways, apoptosis, and rearrangements of the host actin-cytoskeleton (Tohidpour et al. 2017).Similarly, it has been described actin-rearrangement in the pancreas, during both, pancreatitis and pancreatic cancer (Morris and Machesky 2015).In this work, in the stomach of infected gerbils, we also observed important rearrangements in the cytoskeleton, as well as a reduction in the amount of filamentous actin, which should be confirmed by western blot experiments in the future.Similar results were obtained in the pancreatic tissue, where reorganization of the actin-cytoskeleton was observed in infected gerbils (Figs. 5 and 6).These changes might represent the highly variable actin dynamics, trying to maintain the homeostasis of both, acini and beta cells.
In order to regulate the release of hormones in the Langerhans islet, the endocrine cells interact with one another either through homotypic or heterotypic cell-cell adhesion or in a paracrine manner (Jain and Lammert 2009).Pancreatitis can lead to diabetes mellitus, where loss of functional or structural β cells and the altered insulin secretion produced by harmed junctional proteins have been described (Singh et al. 2022).Furthermore, pancreatic inflammation leads to the decrease of islet cell mass leading to the loss of glucagon, insulin, and pancreatic polypeptide, which difficult the control of diabetes with large variations in blood glucose (Singh et al. 2022).In this work, we observed that the H. pylori infection produced alterations in the localization pattern of insulin and glucagon, and an apparent reduction in their levels at pancreas (Fig. 7).This effect could be a consequence of the modified intercellular junction proteins.For example, claudin 4 is involved in regulating the functional state of islet, and may act as a biomarker of β-cell maturation (Li et al. 2020).Furthermore, E-cadherin plays an important role in islet formation, glucose-stimulated insulin secretion and gap junction communication (Jain and Lammert 2009).However, more experiments should be carried out in order to demonstrate the participation of junctional proteins in the endocrine altered functions of the pancreas during H. pylori infection.
During host infection, IL-8 increases in response to H. pylori, and it is a key chemokine in accumulating neutrophils.Expression of this cytokine is often regulated by NF-κB (transcription factor complex nuclear factor-κB), through κB-binding elements in the enhancer/promoter regions of their genes (Eftang et al. 2012).Here, we confirmed that infected animals presented increased levels of IL-8 in serum (Fig. 8).This finding supports the idea that IL-8 appears paramount in the inflammatory response to H. pylori infection, affecting the whole organism and other organs besides the stomach, where the induced inflammation importantly contributes to the damage in pancreas.

Conclusion
We conclude that H. pylori is capable to reach the pancreas, where it produces an important damage in paracellular proteins and actin-cytoskeleton, affecting the insulin and glucagon distribution.Nevertheless, the mechanisms underlying the arrival of this bacterium or its virulence factors (such as CagA or OMPs) to this secretory epithelium, should be extensively studied in future works, to better understand its relationship with pancreatic diseases.

Fig. 1
Fig. 1 Presence of bacterial genes at pancreas and percentage of gerbils infected with H. pylori.A Urease test in homogenized pancreas and stomach.Data from groups Hp-inoculated and Et-OH treated + Hp-inoculated were statistically compared with Et-OH treated animals, by Chi-square test for trend using the GraphPad Prism software.*p < 0.05.B, C PCR assays of glmM (B) or cagA (C)

Fig. 2
Fig. 2 Histopathology of stomach and pancreas tissues from gerbils.A, B Sections of gerbil's stomach (A) and pancreas (B) tissues were processed for H&E staining and observed under the light microscope.Representative images from one animal are shown for each group

Fig. 3
Fig. 3 Presence of H. pylori at gerbil pancreas.A Pancreas sections were processed for immunofluorescence assays using ⍺-Hp antibody (green).Nuclei (blue) were stained with DAPI.Merge: fluorescence images overlapped with the phase-contrast bright field.Arrows: Hp localization outside the cells.Bar = 50 μm.Representative images

Fig. 4
Fig. 4 Presence of H. pylori proteins at gerbil pancreas.A, B Pancreas sections were processed for immunofluorescence assays using ⍺-CagA (A) or ⍺-OMPs (B) antibodies (red).Nuclei (blue) were stained with DAPI.Arrows: localization of bacterial proteins inside the cells.Merge: fluorescence images overlapped with the phase-contrast bright field.Bar = 50 μm.Representative images from one ani-

Table 1
Number of gerbils positive to urease test, Hp culture, and glmM and cagA gene amplification by PCR assays among different groups