Masato Nose MD, Department of Pathology, Ehime University School of Medicine, Shigenobu-cho, Onsen-gun, Ehime, 791–0295, Japan. E-mail: email@example.com
In this study we established a new animal model for exploring the pathogenesis of autoimmune pancreatitis. We have found previously that MRL/Mp-+/+(MRL/+) mice develop pancreatitis spontaneously by an autoimmune mechanism but only when they are more than 34 weeks old. Because this disease might be a model of multi-factorial diseases controlled by genetic and environmental factors, beginning at 6 weeks old, we injected polyinosinic:polycytidylic acid (poly I:C) into MRL/+ mice and in addtion, into MRL/Mp mice bearing the Fas deletion mutant gene, lpr (MRL/lpr). Poly I:C induced chronic severe pancreatitis in all the MRL/+ mice and to a lesser extent in the MRL/lpr mice by 18 weeks of age. There was no pancreatitis in control mice of both strains at the same age. Other than chronic pancreatitis, no severe autoimmune diseases were observed in MRL/+ mice. Immunohistochemical examinations revealed predominant infiltration of CD4+ T cells and Mac-2+ activated macrophages in the pancreatic lesions. Splenic expression of the mRNAs for TNF-α and IL-10, which is known to suppress the development of pancreatitis, were increased in both strains of mice. These findings suggest that an MRL strain of mice treated with poly I:C might be a good model for developing new approaches to the study of the pathogenesis of autoimmune pancreatitis.
Chronic pancreatitis is characterized by the presence of an inflammatory infiltrate and progressive destruction of acinar cells and their replacement by fibrous tissue. The dominant feature of the inflammation is infiltration of CD4+ and CD8+ lymphocytes and macrophages . Up-regulation of major histocompatibility complex (MHC) class I and aberrant expression of MHC class II have been described in chronic pancreatitis . In addition, the inflammatory cytokines, TNF-α, IL-6, IL-8, IL-1 and IL-10 have been implicated in the disease process [3–7]. These findings suggest that autoimmunity plays an important role in the development of chronic pancreatitis. Although much attention has focused on its pathogenesis, the cellular and molecular mechanisms, and their genetic bases, involved in chronic inflammation of the pancreas are not clearly understood, due partly to the lack of suitable animal models of autoimmune pancreatitis. Commonly used models, such as alcohol administration , duct ligation  or injection of a toxic compound  have produced inconsistent results and their relevance to human disease has been questioned recently . Thus, a suitable animal model would be very useful for elucidating the pathogenesis of chronic pancreatitis and developing new therapeutic approaches.
An MRL/MpJ strain of mice bearing the lymphoproliferation gene, lpr (MRL/lpr), which was established by backcrosses and intercrosses between LG/J, AKR/J, C3H/Di and C57BL/6 J strains of mice, develops severe autoimmunue diseases such as glomerulonephritis, arteritis, sialoadenitis and arthritis spontaneously, associated with autoantibody production and T cell dysfunction early in life [12–14]. MRL/Mp mice not bearing the lpr gene (MRL/+) also develop these diseases, but at a much later stage in life, and with reduced incidence and severity [13,14], suggesting that the MRL strains have an autoimmune disease-prone genetic background . Moreover, our recent studies in genetic analyses revealed that each lupus lesion in MRL/lpr mice is under a control of polygenic inheritance [16–18].
We showed previously that 34–38-week-old MRL/+ female mice develop chronic pancreatitis spontaneously with an incidence of 71%, whereas the male mice develop pancreatitis with similar histopathological manifestations much later in life, 45–50 weeks, and with an incidence of less than 40%. In contrast, MRL/lpr mice develop autoimmune diseases in the first few months of life which are fatal within 20–28 weeks, but only a few were observed with pancreatitis. Although pancreatic inflammatory lesions in aged MRL/+ mice are mediated essentially by an autoimmune mechanism, as shown in our previous studies using adoptive cell transfer or thymus transplantation [19,20], the effect of ageing on the development of the disease in MRL/+ mice remains unclear. On the other hand, mice homozygous for the alymphoplasia (aly) mutation can be a good model for autoimmune pancreatitis since this strain spontaneously develops autoimmune disease of exocrine organs involving salivary glands and pancreas [21,22]. Recently, the aly mutation was clarified to be a point mutation in Nf-kappa b-inducing kinase . However, autoimmune pancreatitis might be one of the multifactorial diseases which are generated by complex combinations of intrinsic or extrinsic environmental factors and particular genetic backgrounds. Thus, we grasp the pancreatitis in MRL/+ mice as a multifactorial disease model, and then attempted to generate pancreatitis in younger MRL/+ mice using exogenous stimulation in this study.
Polyinosinic:polycytidylic acid (poly I:C) is a synthetic double-stranded polyribonucleotide . Because of its structural resemblance to double-stranded viral RNA, poly I:C can elicit immunomodulation, including activation of macrophages, NK and B cells and the production of cytokines [25–27]. In addition to lymphocyte activation, poly I:C induces rat exocrine pancreatic endothelium to express intercellular adhesion molecule 1, up-regulates both rat aortic and human umbilical endothelial expression of several adhesion molecules and stimulates the release of IL-6 and IL-8 . The endothelial cell activation might be an initiating event in the development of inflammation. Moreover, multiple injections of poly I:C accelerated the production of anti-dsDNA antibody and renal disease in NZB/NZW female mice . These findings indicate that poly I:C might accelerate the development of autoimmune diseases in a particular genetic background.
In this study, we injected poly I:C intraperitoneally into MRL/+ and MRL/lpr mice to see whether it would accelerate the development of autoimmune diseases, including pancreatitis. We found that poly I:C treatments accerelate the development of chronic pancreatitis and also intrahepatic cholangitis, but not other lupus lesions. We describe here a useful animal model in which autoimmune pancreatitis and primary biliary cirrhosis are induced in an MRL strain of mice by poly I:C.
MATERIALS AND METHODS
Six-week-old female MRL/Mp-+/+(MRL/+) and MRL/Mp-lpr/lpr (MRL/lpr) mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) and maintained in our animal centre under specific pathogen-free conditions. Animals were randomly distributed into the different groups as indicated. Poly I:C (Sigma Chemical Co., St Louis, MO, USA) was injected intraperitoneally at a dose of 5 mg/kg once every 3 days, starting when the animals were 6 weeks old and until they were 18 weeks old. The controls received the same volume of carrier solution.
At the age of 18 weeks, both strains of mice were anaesthetized with ether and sacrificed by cervical dislocation. Blood was collected and sera were stored at −20°C. Half of the pancreas, kidneys, liver, lungs, heart, spleen, axillary lymph nodes, submandibular glands and ankle joints were removed from each mouse for histopathological examinations. Weights of spleen and bilateral axillary lymph nodes were measured at autopsy. Tissues were fixed in 10% formalin in 0·01M phosphate buffer (pH 7·2) and embedded in paraffin. After sectioning they were stained with haematoxylin and eosin or with Elastica-Masson.
Histopathological evaluation of pancreatic lesions was performed by light microscopy. The severity of each lesion in at least one section of pancreas was scored on a 0–4+ scale based on the histopathological changes described by Kanno et al.. Briefly, 0: pancreas without mononuclear cell infiltration, indicating it is almost normal; 1+: mononuclear cell aggregation and/or infiltration within the interstitium without any parenchymal destruction; 2+: focal parenchymal destruction with mononuclear cell infiltration; 3+: diffuse parenchymal destruction but some intact parenchymal residue is retained; 4+: almost all pancreatic tissue, except pancreatic islets, destroyed or replaced with fibrosis or adipose tissue. The maximum score was used as the pancreatitis grade of each individual. To estimate the incidence of pancreatitis, mice with pancreatic lesions that scored >2+ were defined as positive. Other lesions, glomerulonephritis, vasculitis in kidneys, arthritis in ankle joints, sialoadenitis in submandibular glands and intrahepatic cholangitis, were evaluated according to the methods described [16–18,30].
Half of the pancreas was embedded in OCT compound (Miles, Elkhart, IN, USA) and stored at −70°C before cutting. Immunohistochemical examinations were performed by the TSA-Indirect method, according to the manufacturer's instructions (NEN Life Science Products, Boston, MA, USA) with minor modification. The antibodies used for the identification were as follows: rat anti-L3T4 (clone GK1·5) for CD4+ T cells; rat anti-Lyt-2 (clone H35) for CD8+ T cells; rat anti-Mac-2 (clone M3/38; Boehringer Mannheim, Germany) for activated macrophages; rabbit anti-Fas IgG, polyclonal (WAKO, Osaka, Japan); and hamster anti-Fas ligand (FasL) (clone MFL1; BD Biosciences, San Jose, CA, USA).
Measurements of cytokines
Serum levels of TNF-α and IL-10 were determined using commercially available ELISA kits (R&D Systems, Inc., Minneapolis, MN, USA), according to the manufacturer's instructions.
RT-PCR for cytokines was performed as described by Graziosi et al. with minor modification. Briefly, total RNA was isolated from spleen using TRIzol RNA isolation reagent (GIBCO BRL, Life Technologies, New York, NY, USA). Two μg of total RNA was reverse transcribed to cDNA. PCR was performed using specific primers. The forward primer for IL-10  was 5′-CGGGAAGACAATAACTG-3′, and the reverse primer was 5′-CATTTCCGATAAGGCTTGG-3′; the forward primer for TNF-α was 5′-AGCCCACGTCGTAGCAAACCACCAA-3′, and the reverse primer was 5′- ACACCCATTCCCTTCACAGA GCAAT-3′. For IFN-γ, 5′-AACGCTACACACTGC ATCTTGG-3′ as the forward primer and 5′-GACTTCAAAGAGTCT GAGG-3′ as the reverse primer were used. Amplified products were subjected to 2% agarose gel electrophoresis and SYBR green staining. The gels were analysed with a FluorImagerSI (Amersham Bioscience Corp., Piscataway, USA). All cytokine values were normalized individually to the corresponding values for the housekeeping gene hypoxanthine-guanine phosphoribosyltransferase (HPRT).
Data are expressed as means ± s.e. Group means were analysed by the Student's t-test. Incidences of diseases were compared by chi-square analysis. P < 0·05 was considered significant.
Poly I:C accelerated the development of chronic pancreatitis in MRL/+ and MRL/lpr mice
Striking differences in both the time of onset and the incidences of chronic pancreatitis were observed in poly I:C-treated compared to controls. By 18 weeks of age, poly I:C-treated mice developed chronic pancreatitis with an incidence of 100% in MRL/+ and an incidence of 93·3% in MRL/lpr mice, but controls did not (P < 0·01 each) (Table 1). In addition, poly I:C-treated MRL/+ and MRL/lpr mice began to lose weight from the second and fourth weeks after the first injection, respectively, and showed consistently less weight gain than their controls (Fig. 1).
Table 1. Incidence of lupus in poly I:C-treated micea
Pancreatic histopathology and immunohistochemistry
Initial pancreatic lesions were observed in the periductal and perivascular regions, characteristic of the infiltration of mononuclear cells, in both strains of poly I:C-treated mice at 18 weeks of age, and these lesions extended into the interstitium (Grade 2+) (Fig. 2a). Following these lesions, cell infiltrates became significant in the pancreatic parenchyma. The acini seemed to be diluted, and the destruction of acini and replacement of parenchyma with fibrous or adipose tissues was seen, but it was not associated with necrotic or haemorrhagic lesions. However, pancreatic islets were unaffected and remained intact throughout the progression of disease (Grade 3+, 4+) (Fig. 2b,c). The pancreatic lesions in MRL/+ mice were more severe than those in MRL/lpr mice. The incidence of Grade 4+ was 93% in MRL/+ mice, but only 47% in MRL/lpr mice (P < 0·01) (Fig. 3).
Immunohistochemistry revealed that CD4+ T cells were predominant, but a few CD8+ T cells and Mac-2+ macrophages were observed in the pancreatic interstitium and parenchyma. FasL was present mainly in the ductal cells and some acinar cells in both strains of poly I:C-treated mice, while Fas was observed only in MRL/+ mice, remarkably in lymphoid cells infiltrating into the periductal regions, but slightly in the ductal cells (Fig. 4a–e).
Intrahepatic cholangitis and other lesions
To determine if the immune cell infiltration and tissue destruction were specific to the pancreas or if a generalized immune response was occurring against other organs, especially other exocrine tissues, we examined the submandibular glands, lungs, heart, liver and kidneys. Interestingly, Poly I:C treatment induced intrahepatic cholangitis in both strains (Table 1). This lesion was characterized by lymphocytic infiltration at the periductal regions, followed by destruction of bile ducts, resembling the early stage of primary bile cirrhosis (PBC) (Fig. 2d). Immunohistochemistry showed that CD4+ T cells and Mac-2+ macrophages were predominant, but a few CD8+ T cells were observed in the periductal regions (Fig. 4f–h).
Both strains of control mice developed sialoadenitis at 18 weeks of age as did the poly I:C-treated mice. Only MRL/ lpr mice developed glomerulonephritis, vasculitis and arthritis (Table 1). Compared to controls the weight of spleen increased in both strains of poly I:C-treated mice at this age, whereas there was no significant difference in weights of axillary lymph nodes between mice injected with poly I:C and controls (data not shown).
Poly I:C increased the levels of cytokines in serum and cytokine mRNAs in spleen
At 18 weeks of age, serum levels of TNF-α and IL-10 were increased markedly in both strains of mice subjected to the poly I:C regimen. Concomitantly, the mRNAs of these cytokines and IFN-γ in spleen were increased compared with corresponding controls (Table 2). In addition, in the control groups serum levels of TNF-α and IL-10 in MRL/lpr mice were higher than those in MRL/+ mice (P < 0·0084, P < 0·024, respectively).
Table 2. Effect of poly I:C on serum levels of TNF-α and IL-10 and expression of their mRNAs in spleen
In the present studies, histopathological examinations clearly showed that MRL mice treated with poly I:C developed chronic pancreatic lesions with an incidence of 100% in MRL/+ and 93% in MRL/lpr mice at the age of 18 weeks (Table 1), suggesting that poly I:C accelerated the development of pancreatitis by at least 16 weeks compared to the spontaneous occurrence of chronic pancreatitis in MRL/+ mice at 34–38 weeks of age . The pancreatitis accelerated by poly I:C in both MRL/+ and MRL/lpr mice was characterized by inflammatory cell infiltrates with destruction of acinar cells and replacement by fibrous tissue, and was selective for the exocrine pancreas since it did not involve pancreatic islet cells. Moreover, immunohistochemical studies revealed that the pancreatic inflammatory lesions were mediated by CD4+ T cells and Mac-2+ macrophages. Additionally, in our preliminary studies we observed that the vascular endothelial cells in pancreatitic lesions highly expressed adhesion molecules such as ICAM and VCAM. The histopathology of the pancreatitis and cytokine alterations that we found (Table 2) suggest that the animal model might simulate human chronic pancreatitis.
It has been demonstrated that poly I:C activates macrophages by a process that involves at least two steps. In the first priming step, macrophages are committed by exposure to IFN-α/β induced by poly I:C. The second triggering step involves a direct effect of poly I:C on IFN-primed macrophages , resulting in the production of IFN-α/β, TNF-α and IL-12 [27,34,35], and activation of CD4+ T cells . These activated macrophages may induce destruction of pancreatic parenchyma directly or via antibody-dependent cellular cytotoxicity (ADCC) through the reaction of autoantibodies with acinar cells. In this study, we found infiltration of CD4+ cells and Mac-2+ cells in the pancreatic inflammatory sites (Fig. 4) where they might have been activated by poly I:C. Therefore, macrophages in an MRL strain of mice may play important roles in the progression of pancreatic inflammatory lesions in a particular autoimmune background. In addition, IFN-γ expression in spleen was significantly increased in the treated mice with poly I:C. These results were consistent with previous reports showing that IFN-γ expression was increased in acute pancreatitis in human  and chronic pancreatitis in a rat model . Moreover, anti-IFN-γ administration reveals the therapeutic activity in a mouse pancreatitis model . These findings indicate that IFN-γ expression also may be related to the development of pancreatitis.
Fas-mediated apoptosis might be involved in the cell-mediated cytotoxicity and blocking the apoptotic process may alleviate the cytotoxic lesions [40,41]. Recentlly, several reports show the expression of Fas and Fas ligand in the duct epithelium of human chronic pancreatitis, and suggests that Fas-mediated apoptosis play an important role on the development of chronic pancreatitis [42,43]. These findings are consistent with ours that pancreatic lesions in poly I:C-treated MRL/+ mice were more severe than those in MRL/lpr mice bearing a Fas deletion mutation (Fig. 3) and Fas ligand was strongly expressed on the ductal cells of poly I:C-treated MRL/+ mice (Fig. 4d). On the other hand, Kaiser and Bhatia showed that apoptosis is a favourable response to acinar cell injury in experimental acute pancreatitis, because apoptotic acinar cells release less digestive enzymes and/or chemotactic factors than necrotic acinar cells [44,45]. Thus, mice with an MRL background treated with poly I:C might provide a new model to further investigate the relationship between pancreatic lesions and Fas-mediated apoptosis.
Recently, considerable attention has been directed to cytokine regulation and responses in acute and chronic pancreatitis. It has been demonstrated that TNF-α levels are increased in the serum of patients with pancreatitis as well as in animal models of pancreatitis [4,46]. Inhibition of TNF-α seems to be associated with improved survival in experimental acute pancreatitis . Macrophage/monocytes are considered to be a major source of TNF-α production  and TNF-α was also found to be secreted by both normal and inflammatory rat pancreatic acinar cells . In addition, a high level of TNF-α mRNA expression was found not only in pancreatic tissue but also in the spleen, liver and lungs in a model of pancreatitis . As shown in Table 2, the production of TNF-α and the expression of its mRNA in spleen were increased significantly in response to poly I:C in both MRL strains of mice, whereas poly I:C did not induce these increases in BALB/c mice , suggesting that TNF-α might be involved in the progression of pancreatitis and the poly I:C acceleration of the development of chronic pancreatitis in mice with an MRL autoimmune genetic background.
Alternatively, IL-10 has been reported to have multiple suppressive effects on the immune response . Administration of recombinant IL-10 prevents necrosis and reduces the severity of experimental acute pancreatitis if given before or after the induction of pancreatitis . Survival from lethal pancreatitis seems to be improved by IL-10 administration . Exogenous IL-10 probably acts, in part, through inhibition of the local release of TNF-α from activated monocytes/macrophages [50,53]. We found that the secretion of IL-10 and the splenic expression of its mRNA were increased with the rise of TNF-α in poly I:C-treated mice, indicating that the elevation of IL-10 might be an adaptive protective response to the rise in TNF-α.
Although we concentrated on pancreatitis in this study, the liver pathology showing remarkable intrahepatic cholangitis suggests that MRL mice treated with poly I:C might be a new model of primary biliary cirrhosis, considering the fact that MRL/lpr mice develop non-suppurative cholangitis spontaneously associated with the production of antimitochondrial autoantibodies .
In conclusion, we demonstrated that poly I:C treatment accelerates the onset and increases the incidence of pancreatitis in MRL mice. Activated T cells and macrophages, and high levels of cytokines suggest that the progression of chronic pancreatitis might be associated with an autoimmune mechanism in poly I:C-treated mice. MRL/+ mice treated with poly I:C developed chronic pancreatitis but not other severe lupus diseases, indicating that MRL/+ mice treated with poly I:C might be a suitable model of certain aspects of human chronic pancreatitis. The model should facilitate elucidation of the immune mechanisms and pathology of chronic pancreatitis, and contribute to the development of effective new therapies.
We wish to thank Dr H. Schulman for reviewing the manuscript and Ms M. Aibara for its preparation. This work was supported in part by a Grant-in Aid for Scientific Research of the Ministry of Education, Science and Culture of Japan.