Persistent Salmonellosis Causes Pancreatitis in a Murine Model of Infection

Pancreatitis, a known risk factor for the development of pancreatic ductal adenocarcinoma, is a serious, widespread medical condition usually caused by alcohol abuse or gallstone-mediated ductal obstruction. However, many cases of pancreatitis are of an unknown etiology. Pancreatitis has been linked to bacterial infection, but causality has yet to be established. Here, we found that persistent infection of mice with the bacterial pathogen Salmonella enterica serovar Typhimurium (S. Typhimurium) was sufficient to induce pancreatitis reminiscent of the human disease. Specifically, we found that pancreatitis induced by persistent S. Typhimurium infection was characterized by a loss of pancreatic acinar cells, acinar-to-ductal metaplasia, fibrosis and accumulation of inflammatory cells, including CD11b+ F4/80+, CD11b+ Ly6Cint Ly6G+ and CD11b+ Ly6Chi Ly6G− cells. Furthermore, we found that S. Typhimurium colonized and persisted in the pancreas, associated with pancreatic acinar cells in vivo, and could invade cultured pancreatic acinar cells in vitro. Thus, persistent infection of mice with S. Typhimurium may serve as a useful model for the study of pancreatitis as it relates to bacterial infection. Increased knowledge of how pathogenic bacteria can cause pancreatitis will provide a more integrated picture of the etiology of the disease and could lead to the development of new therapeutic approaches for treatment and prevention of pancreatitis and pancreatic ductal adenocarcinoma.


Introduction
Pancreatitis affects over 80,000 Americans every year, and, in its chronic form, is a known risk factor for the development of pancreatic ductal adenocarcinoma (PDA) [1,2]. Acute pancreatitis ranges in severity from mild interstitial pancreatitis to a much more severe condition associated with necrosis and concomitant multi-organ failure [3]. Most patients with acute pancreatitis suffer from mild interstitial pancreatitis, but up to 20% of patients suffer from severe pancreatitis, which is often fatal [4]. Under conditions of persistent or repeated insult, acute pancreatitis can progress to chronic pancreatitis, which is an often asymptomatic condition diagnosed only after the development of complications. The incidence of biliary pancreatitis increased by 32% between 1994 and 2001, likely due to the climbing obesity rate and obesityassociated increase in the development of gallstones [5,6]. Other established risk factors for the development of pancreatitis include excessive alcohol consumption, cigarette smoking, and genetic predisposition [6]. Even though 70% of chronic pancreatitis cases are attributed to alcohol abuse, 95% of alcoholics never develop pancreatitis. The remaining chronic pancreatitis cases are considered idiopathic in nature [1].
Salmonellae are a leading cause of morbidity and mortality in humans worldwide [30]. Infections with Salmonellae range in severity from self-limiting gastroenteritis to typhoid fever and can lead to chronic carriage. Non-typhoidal Salmonellae such as Salmonella enterica serovar Typhimurium (S. Typhimurium) are a leading cause of inflammatory enterocolitis and death due to foodborne illness, and are a significant cause of invasive bacteremia in immunocompromised hosts. Typhoidal Salmonellae such as Salmonella enterica serovar Typhi (S. Typhi) cause systemic infections characterized by bacterial penetration of the intestinal barrier and extraintestinal dissemination to the liver and spleen, where the microorganisms survive and replicate in professional phagocytes [31,32]. Septic shock and death can occur if infections are left untreated [33]. Much of what is known about the pathogenesis of and host response to Salmonellae comes from experimental infection of mice with S. Typhimurium, which has served as a useful model for the human disease caused by S. Typhi [32].
Here, we report that pancreata of mice persistently infected with S. Typhimurium consistently displayed inflammatory, fibrotic and epithelial responses similar to chronic pancreatitis in humans. S. Typhimurium colonized the pancreas throughout the course of infection, associated with pancreatic acinar cells in vivo, and could directly invade pancreatic acinar cells in vitro. Thus, Salmonella infection can cause pancreatitis, a known risk factor for the development of PDA.

Ethics Statement
All procedures and experiments using mice were approved by Institutional Animal Care and Use Committees at Stony Brook University or Mayo Clinic, and were conducted in accordance with the recommendations outlined in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All procedures and experiments using mice were designed to use the fewest number of mice possible but still achieve meaningful results. For intravenous inoculations, mice were anesthetized by inhalation of isoflurane to facilitate the procedure and minimize distress. All mice were given food and water ad libitum and were monitored twice daily. Any mice that appeared moribund (e.g. ruffled fur, hunched posture, lack of activity) were euthanized immediately. In any case, euthanasia was performed by inhalation of carbon dioxide, a method consistent with the recommendations of the Panel on Euthanasia of the American Veterinary Medical Association.

Bacteria
S. Typhimurium strain IR715, which is a spontaneous, nalidixic acid-resistant derivative of S. Typhimurium strain 14028 (American Type Culture Collection), was used as the wild-type strain. Where indicated, isogenic, invA-deficient or enhanced GFPexpressing S. Typhimurium were used. These strains were generated by bacteriophage P22-mediated transduction, moving the invA::cam and rpsM::egfp(kan) mutations described previously [34,35] from isogenic derivatives of S. Typhimurium strain 14028 into S. Typhimurium strain IR715. Bacteria were grown aerobically for 16-18 h at 37uC in 3 ml of Luria-Bertani (LB) broth (250 rpm) or on LB agar using standard microbiological techniques.
Mice C57BL/6J mice (8 to 12 weeks of age), which lack a functional Nramp1/Slc11a1 locus, were purchased from The Jackson Laboratory and used as the wild-type strain of mouse. The Nramp1/ Slc11a1 locus encodes the natural resistance-associated macrophage protein 1 (Nramp1) (Slc11a1) divalent metal transporter, which enhances host resistance to a number of intracellular pathogens, including S. Typhimurium, by limiting essential metal availability within the phagocyte phagosome [36]. C57BL/6J mice are phenotypically Nramp1 2 because of a G169D mutation and have been used as model hosts to study acute salmonellosis. A breeder pair of transgenic C57BL/6J Nramp1 G169 mice [36] was generously provided by Dr. Ferric Fang (University of Washing-ton). These mice, which are phenotypically Nramp1 + , have been used as model hosts to study persistent salmonellosis [37] and were bred at Stony Brook University, Division of Laboratory Animal Resources. C57BL/6J Nramp1 G169 mice (8 to 12 weeks of age) bred at Stony Brook University, Division of Laboratory Animal Resources were used for the experiments described in this study.

Mouse Injections
Mouse injections were performed using naïve, 8-to 12-week-old sex-matched C57BL/6J mice. Briefly, mice were injected intraperitoneally with purified S. Typhimurium lipopolysaccharide (LPS) (Enzo Life Sciences) (5 mg/kg) every other day for 10 days. One day after the last injection, pancreata were harvested and processed for analysis by histopathology or flow cytometry.

Mouse Infections
Mouse infections were performed using naïve, 8-to 12-week-old sex-matched C57BL/6J Nramp1 G169 mice. Briefly, mice were inoculated intravenously with 5610 3 colony forming units (CFU) of S. Typhimurium strain IR715 suspended in 0.1 ml of PBS, unless indicated otherwise. Ten-fold serial dilutions of the inoculum were plated on LB agar to confirm the inoculum titer. At indicated times after inoculation, target organs (i.e. pancreas, liver, and spleen) were harvested and processed for analysis by histopathology, flow cytometry or organ burden assay. Bacterial loads were determined by lysing cells from a single cell suspension with Triton X-100 (0.05%) and plating for CFU on LB agar containing nalidixic acid (50 mg/ml). Mice infected with S. Typhimurium were euthanized when moribund or at the termination of the experiment.

Histological Staining and Quantitation
Pancreatic tissues were fixed overnight in 4% paraformaldehyde, dehydrated and paraffin embedded. Routinely, tissue sections were stained using hematoxylin and eosin (H&E) for overall tissue structure, cytokeratin 19 (Abcam, Cambridge, MA) for metaplasia, and Picrosirius Red Stain Kit (Polysciences) for fibrosis. Immunohistochemistry (IHC) was performed as described previously [38]. Briefly, tissue sections were stained using antimouse antibodies specific for collagen I, F4/80 or Ly6B.2 (all from AbD Serotec, Kidlington, UK). Stained tissue sections were examined and photographed using an Olympus BX41 light microscope (Olympus). Inflammatory cell infiltration was quantified by examining 5 representative slides (eight 20x fields per slide) per mouse (n = 4 per group) using ImageScope v11.1.2.752 software (Aperio, Vista, CA). F4/80 and collagen staining was quantified by determining the percentage of positive pixels per field, whereas Ly6B.2 staining was quantified by using an algorithm that calculated positive cell nuclei.

Fluorescence Microscopy
Pancreatic tissues were fixed for three hours in 4% paraformaldehyde, washed three times (5 minutes per wash) with PBS (0.1 M) and floated overnight in 30% sucrose. Tissues were then incubated for 30 minutes in a 1:1 mixture of 30% sucrose and optimal cutting temperature compound (OCT), embedded in OCT and frozen at 280uC. Tissue sections of 7 mm each were produced, permeabilized with 0.1% Triton X-100 in 10 mM PBS and blocked with 5% normal donkey serum and 1% BSA in 10 mM PBS for 1 hour at room temperature. Tissue sections were then stained with Alexa Fluor 594 phalloidin (Invitrogen) in 10 mM PBS supplemented with 1% BSA and 0.1% Triton X-100 for 1 hour at room temperature, washed three times with 0.1% Triton X-100 in PBS and rinsed with deionized water. Slides were mounted in VECTASHIELD mounting medium with DAPI (Vector Laboratories). Stained tissue sections were examined and photographed using a Zeiss 510LS Meta confocal microscope (Carl Zeiss MicroImaging).

Acinar Cell Culture and Infection Assay
The murine acinar cell line 266-6 (American Type Culture Collection) was maintained in DMEM supplemented with 10% fetal bovine serum (Atlanta Biologicals) and sodium pyruvate (1 mM), and incubated at 37uC in 5% CO 2 . Acinar cell infections were performed using a standard gentamicin protection assay [31,39]. Briefly, acinar cells were suspended in medium lacking antibiotics and seeded at 5610 5 cells per ml per well into a 24-well tissue culture plate. After overnight incubation, the medium was replaced with 0.5 ml of fresh medium and the cells were infected with bacteria at a multiplicity of infection of 50. Upon addition of bacteria, the plate was centrifuged for 5 minutes at 1,000 rpm to facilitate bacterial contact with the acinar cells. After 20 minutes of incubation, the wells were washed three times with 1 ml of PBS to remove non-cell-associated bacteria, and fresh medium supplemented with gentamicin (25 mg/ml) was added to each well to kill all extracellular bacteria. This was referred to as the 0 hour time point. After 1 hour of incubation, the wells were washed three times with 1 ml of PBS and the acinar cells were lysed using 0.5 ml of Triton X-100 (0.1%) to release intracellular bacteria. These bacteria were enumerated by plating onto LB agar. In experiments where the acinar cells were infected with GFP-expressing bacteria, the cells were harvested and analyzed by flow cytometry.

S. Typhimurium LPS Induces Pancreatic Inflammation
Bacterial infection, including that caused by Salmonellae, has been implicated in human cases of pancreatitis [13,[17][18][19][20][21][22][23][24][25][26][27]. Consistent with this notion, Escherichia coli LPS exacerbates pancreatitis in alcohol and caerulein-induced animal models, hypothetically through its direct effects on pancreatic acinar cells [40][41][42]. Yet, Escherichia coli LPS alone fails to induce the desmoplastic reaction or obvious acinar cell stresses that are hallmarks of pancreatitis [43]. It is generally known that administration of LPS induces a systemic inflammatory response that affects many organs. To determine how the pancreas may be affected by S. Typhimurium LPS-induced systemic inflammation, C57BL/6J mice were injected intraperitoneally with S. Typhimurium LPS every other day for 10 days. One day after the last injection, pancreata were harvested and processed for analysis by histopathology. Consistent with previous studies [43], H&E staining revealed little tissue damage or associated edema ( Figure 1A). However, IHC detection of F4/80 and Ly6B.2 revealed a uniform, pancreas-wide distribution of F4/80 + macrophages ( Figures 1A and 1B) with very few Ly6B.2 + neutrophils (data not shown). Flow cytometric analysis confirmed the presence of large numbers of F4/80 + cells in pancreata of LPS-treated mice as compared to mock-treated mice ( Figure 1C). These cells also expressed surface CD11b (Figures 1C and 1D). The CD11b + cells present in pancreata of mice treated with LPS were a heterogeneous population that consisted mostly of CD11b + F4/ 80 + macrophages ( Figures 1C and 1D), but also included CD11b + Ly6C hi Ly6G 2 inflammatory monocytes and CD11b + Ly6C int Ly6G + neutrophilic granulocytes (Figures 1E and 1F). Thus, S. Typhimurium LPS induces pancreatic inflammation, but fails to induce the reactive epithelial and fibrotic responses that are characteristic of pancreatitis.

S. Typhimurium Infection Induces Pancreatitis
As S. Typhimurium LPS alone was insufficient to induce pancreatitis, we sought to determine the ability of S. Typhimurium infection to induce a pancreatitis phenotype more similar to the human disease condition. Therefore, C57BL/6J Nramp1 G169 mice were inoculated intravenously with 5610 3 CFU of S. Typhimurium. After 10 days of infection, an early time point used to assess establishment of salmonellosis, histopathologic analysis showed evidence of mild pancreatic edema, but no cytokeratin 19-positive epithelial metaplasia or significant fibrosis (Figure 2A). In addition, IHC for F4/80 and Ly6B.2 revealed the presence of significant numbers of focally pooled macrophages and neutrophils, respectively, in pancreata of mice infected with S. Typhimurium as compared to mice left uninfected ( Figures 2B  and 2C). Consistent with these results, we found significantly more CD11b + F4/80 + macrophages ( Figures 2D and 2E), CD11b + Ly6C hi Ly6G 2 inflammatory monocytes and CD11b + Ly6C int Ly6G + neutrophilic granulocytes ( Figures 2F and 2G) in pancreata of mice infected with S. Typhimurium than in pancreata of mice left uninfected. Thus, S. Typhimurium infection induces significant pancreatic inflammation without ductal metaplasia or fibrosis, a phenotype that is similar to acute pancreatitis in humans.

S. Typhimurium Colonizes and Persists in the Pancreas, and can Invade Pancreatic Acinar Cells
A recent study reported on the detection of Salmonella Enteritidis genomic DNA in pancreata of mice infected with Salmonella Enteritidis [44], suggesting that Salmonellae may colonize the pancreas. To characterize the ability of S. Typhimurium to colonize and persist in pancreas, we determined bacterial loads in pancreata of C57BL/6 Nramp1 G169 mice inoculated intravenously with 5610 3 CFU of S. Typhimurium. After 10 days of infection, we recovered substantial numbers of S. Typhimurium from pancreata of infected mice ( Figure 3A). These numbers persisted over a period of 60 days ( Figure 3A) and were similar to the numbers of S. Typhimurium recovered from liver and spleen ( Figures 3B and 3C). To visualize S. Typhimurium in the pancreas, we inoculated C57BL/6 Nramp1 G169 mice intravenously with 5610 3 CFU of S. Typhimurium expressing GFP. After 10 days of infection, we found S. Typhimurium associated with acinar  To characterize the ability of S. Typhimurium to invade pancreatic acinar cells, we infected cultured murine pancreatic acinar cells (line 266-6) with wild-type or invA-deficient S. Typhimurium at a multiplicity of infection of 50. The invA gene encodes an essential structural component of the Salmonella Pathogenicity Island (SPI)-1-encoded Type Three Secretion System (TTSS), which is required for invasion of non-phagocytic cells [45]. After 1 hour of infection, we recovered substantial numbers of intracellular bacteria ( Figure 3E), indicating that S. Typhimurium had invaded the acinar cells. We recovered significantly fewer intracellular invA-deficient S. Typhimurium than wild-type S. Typhimurium, indicating that efficient invasion of the acinar cells was dependent on the SPI-1-encoded TTSS ( Figure 3E). Similar results were obtained when we analyzed by flow cytometry GFP fluorescence of acinar cells infected with wildtype or invA-deficient S. Typhimurium expressing GFP (Figures 3F and 3G). Collectively, these results indicate that S. Typhimurium colonize and persist in pancreas, and that the bacterial burden in the pancreas may be due, at least in part, to direct infection of acinar cells.

Pancreatitis Progresses with Persistent S. Typhimurium Infection
Given that S. Typhimurium induced a pancreatitis-like phenotype during early stages of infection ( Figure 2) and colonized and persisted in the pancreas over a period of 60 days (Figure 3), we next examined pancreata of C57BL/6J Nramp1 G169 mice persistently infected with S. Typhimurium. Unlike the mild response found after 10 days of S. Typhimurium infection (Figure 2), we found significant pancreatic damage after 60 days of infection, as indicated by large areas of acinar cell loss made evident by H&E staining ( Figure 4A). These areas were marked by cytokeratin 19positive acinar to ductal metaplasia (ADM) and a dramatic desmoplastic response highlighted by substantial collagen deposition ( Figures 4A and 4B). In addition, the relatively mild inflammatory response induced during early stages of S. Typhimurium infection had been replaced by large swaths of F4/80 + macrophages and Ly6B.2 + neutrophils occupying the damaged areas of the pancreas (Figures 4C and 4D). Consistent with these results, we found significantly more CD11b + cells, including CD11b + F4/80 + cells (Figures 4E and 4F) and CD11b + Ly6C int Ly6G + cells ( Figures 4G and H, right panel) in pancreata of mice persistently infected with S. Typhimurium than in pancreata of mice left uninfected. A similar trend was observed for CD11b + Ly6C hi Ly6G 2 cells (Figures 4G and 4H,  left panel), but differences did not reach statistical significance. Similar results were obtained when we examined pancreata of 12961/SvJ mice persistently infected with S. Typhimurium administered intragastrically (data not shown). Collectively, the robust inflammatory response and reactive epithelial and fibrotic responses in the pancreas indicate that persistent S. Typhimurium infection induces a progressive condition that is highly similar to chronic pancreatitis in humans.

Discussion
The etiology of pancreatitis, a known risk factor for the development of PDA, is most commonly associated with alcohol abuse and gallstone-mediated ductal obstruction [1], though frequently the specific cause is unknown. Bacterial infection has frequently been associated with pancreatitis, but is usually considered a consequence of the disease rather than a contributor to the disease. In fact, infection of the pancreas is the primary cause of pancreatitis-associated death [46]. Here, we found that persistent salmonellosis could cause pancreatitis in a murine model of infection. Specifically, we found that pancreatitis induced by persistent S. Typhimurium infection was characterized by a loss of pancreatic acinar cells, acinar to ductal metaplasia, fibrosis, and accumulation of inflammatory cells (Figures 2 and 4). Furthermore, we found that S. Typhimurium colonized and persisted in the pancreas, associated with pancreatic acinar cells in vivo, and could invade cultured pancreatic acinar cells in vitro (Figure 3). An immediate implication of our results is that persistent, chronic or repeated infections with Salmonellae could lead to the development of pancreatitis.
In humans, S. Typhi is the major cause of persistent or chronic salmonellosis. It is estimated that 3-5% of patients infected with S. Typhi become chronic carriers. Chronic infections with S. Typhi can persist for decades and are often asymptomatic, which makes the identification of chronic carriers difficult. The chronic carrier state has been associated with pre-existing hepatobiliary disease such as the presence of gallstones [47]. Chronic carriers have an increased risk of developing hepatobiliary and pancreatic carcinomas [48][49][50][51][52][53], which links persistent salmonellosis to gastrointestinal cancer. It is generally known that Salmonellae can colonize the liver and form biofilms on gallstones in the gallbladder, indicating that direct colonization of these sites may be a mechanism for chronic inflammation, tissue disturbance, and the promotion of a pro-tumorigenic microenvironment [54]. Similarly, our results indicate that direct colonization of the pancreas by Salmonellae can cause pancreatic inflammation and tissue injury that is characterized by, but not limited to, metaplasia, a known precursor for neoplastic transformation [55].
To the best of our knowledge, this report is the first to document that an important long-term consequence of persistent salmonellosis may be induction of pancreatitis, a known risk factor for the development of PDA. The similarities between said long-term consequence of persistent salmonellosis and the long-term consequences of chronic infections with Helicobacter pylori, a bacterial pathogen that has been identified as one of the primary instigators of intestinal metaplasia in the stomach, are striking. Gastric mucosal tissue injury caused by chronic H. pylori infection leads to intestinal metaplasia, which is believed to result from the differentiation of gastric stem cells towards cells of an intestinal phenotype [56]. This intestinal metaplasia carries a significantly increased risk of developing gastric cancer, the second most common cancer globally [57,58]. Analogously, pancreatic metaplasia due to persistent salmonellosis likely carries an increased risk of developing PDA.
Central to the development of PDA are activating oncogenic Kras mutations [59]. Although K-ras mutations alone may not cause PDA, K-ras mutations in conjunction with pancreatitis have been shown to induce progression of pancreatic cancer [60]. In addition to tissue injury, we found that pancreatitis induced by S. Typhimurium was characterized by a robust inflammatory response. This response consisted of an influx of macrophages, CD11b + Ly6C hi Ly6G 2 inflammatory monocytes and CD11b + Ly6C int Ly6G + neutrophilic granulocytes (Figures 2 and 4). Recent studies have shown that macrophages, through the secretion of cytokines, can induce pancreatic metaplasia [61]. Furthermore, the CD11b + Ly6C hi Ly6G 2 and CD11b + Ly6C int Ly6G + cells that accumulate and persist in tissues of mice infected with S. Typhimurium resemble myeloid-derived suppressor cells, which have been associated with immunosuppression in cancer and, more recently, infection [62,63]. We propose a model where persistent salmonellosis induces pancreatic inflammation and tissue injury that may promote the development of metaplasia, which, in conjunction with an activating K-ras mutation, may lead to the development of PDA.
In conclusion, we have shown that persistent salmonellosis causes pancreatitis in a murine model of infection. This model recapitulates the complexity of the human disease and, therefore, may be useful for the study of pancreatitis as it relates to bacterial infection. Increased knowledge of how pathogenic bacteria can cause pancreatitis will provide a more integrated picture of the etiology of the disease and could lead to the development of new therapeutic approaches for treatment and prevention of pancreatitis and PDA.