Transgenic Expression of PRSS1^R122H Sensitizes Mice to Pancreatitis

Abstract

Background & Aims:

Mutations in the trypsinogen gene (PRSS1) cause human hereditary pancreatitis. However, it is not clear how mutant forms of PRSS1 contribute to disease development. We studied the effects of expressing mutant forms of human PRSS1 in mice.

Methods:

We expressed forms of PRSS1 with and without the mutation encoding R122H (PRSS1^R122H) specifically in pancreatic acinar cells under control of a full-length pancreatic elastase gene promoter. Mice that did not express these transgenes were used as controls. Mice were given injections of caerulein to induce acute pancreatitis or injections of lipopolysaccharide (LPS) to induce chronic pancreatitis. Other groups of mice were fed ethanol or placed on a high-fat diet to induce pancreatitis. Pancreata were collected and analyzed by histology, immunoblots, real-time PCR, and immunohistochemistry. Trypsin enzymatic activity and chymotrypsin enzymatic activity were measured in pancreatic homogenates. Blood was collected and serum amylase activity was measured.

Results:

Pancreata from mice expressing transgenes encoding PRSS1 or PRSS1^R122H had focal areas of inflammation; these lesions were more prominent in mice that express PRSS1^R122H. Pancreata from mice that express PRSS1 or PRSS1^R122H had increased levels of HSP70 and NRF2 and reduced levels of chymotrypsin C (CTRC), compared with control mice. Increased expression of PRSS1 or PRSS1^R122H increased focal damage in pancreatic tissues and increased the severity of acute pancreatitis after caerulein injection. Administration of LPS exacerbated inflammation in mice that express PRSS1^R122H compared to mice that express PRSS1 or control mice. Mice that express PRSS1^R122H developed more severe pancreatitis after ethanol feeding or a high-fat diet than mice that express PRSS1 or control mice. Pancreata from mice that express PRSS1^R122H had more DNA damage, apoptosis, and collagen deposition and increased trypsin activity and infiltration by inflammatory cells than mice that express PRSS1 or control mice.

Conclusions:

Expression of a transgene encoding PRSS1^R122H in mice promoted inflammation and increase the severity of pancreatitis, compared with mice that express PRSS1 or control mice. These mice might be used as a model for human hereditary pancreatitis and can be studied to determine mechanisms of induction of pancreatitis by LPS, ethanol, or a high-fat diet.

Lay Summary:

We created mice that express a form of trypsinogen with the same mutation as in patients with inherited pancreatitis. Studies of these mice provide information on how this disease develops in humans.

Graphical Abstract

Introduction:

Pancreatitis, acute (AP) and chronic (CP), are among the most common causes of hospitalization worldwide. Unfortunately, effective and specific methods of treatment do not exist for these diseases. This is primarily due to a lack of understanding of the mechanisms of initiation and progression resulting from a lack of useful animal models. Observations made more than one century ago by Dr. Chiari suggested that pancreatitis was essentially an autodigestive disease caused by prematurely activated digestive enzymes.¹ The identification of gene mutations in the trypsin enzyme activation cascade associated with human hereditary CP supported this concept.² The most prominent mutant genes include cationic trypsinogen (PRSS1) and the pancreatic serine protease inhibitor Kazal type 1 (SPINK1).^{3, 4} However, the role of trypsin activity as an initiator of pancreatitis is not definite, as recent studies using genetic mouse models suggest that intracellular trypsin activity is not sufficient or necessary to initiate pancreatitis.^5–7 Moreover, 20% of patients carrying the most common trypsinogen mutation (PRSS1^R122H, referred to as R122H in this report) do not develop pancreatitis. These observations suggest that there may be an essential role for environmental influences in the development of the disease.^{3, 4} Elucidation of the mechanisms of importance in CP would be significantly facilitated by a tractable mouse model.

Several previous studies have investigated mouse models expressing mutant trypsinogens. Previously, a mouse line expressing mouse PRSS1^R122H was suggested to develop spontaneous CP.⁸ However, it is uncertain if mouse PRSS1 is the ortholog of human PRSS1 and unfortunately that mouse line was lost. In another study, expression of human PRSS1^R122H in mice failed itself to generate a pathological response, although it increased the severity of caerulein-induced pancreatitis.⁹ Recently, it was reported that expression of human PRSS1 and PRSS1^R122H trypsinogens caused the development of spontaneous pancreatitis in up to 10% of mice with no differences between those expressing wild-type and mutant forms.¹⁰ These disparate results have not been able to clarify the role of trypsinogen mutations as an initiator of CP nor have they provided a tractable model for investigation of the human disease.

These controversies prompted us to further explore the effects on the expression of human PRSS1 and R122H on the development of pancreatitis. In the current study, the full-length mouse elastase gene was utilized to drive the expression of the human trypsinogens. The mouse elastase promoter is a specific regulator of a prominent mouse digestive enzyme and provides a robust expression system in the pancreas.^{5, 11} We observed that expression of either human WT PRSS1 or its mutant form, R122H, did not cause spontaneous pancreatitis even if levels were highly elevated by breeding to homozygosity. However, caerulein hyperstimulation resulted in acute pancreatitis that was significantly more severe in both WT PRSS1 and R122H expressing mice than control mice. When challenged with environmental insults that were insufficient to produce pancreatitis themselves, chronic inflammatory changes were significantly higher in the R122H mice. These data indicate that the presence of mutant trypsinogen dramatically increases the susceptibility and severity of pancreatitis to environmental insults. Thus, the R122H mice provide a new model of human hereditary pancreatitis that should be useful to define essential regulators of the disease further.

Materials and Methods

Generation of Transgenic Mice:

Human cationic trypsinogen (PRSS1) cDNA was cloned by RT-PCR from archived normal human pancreas total RNA by RT-PCR. To facilitate the detection of its expression, a hemagglutinin (HA) tag was added to its c-terminus. Mutant PRSS1 R122H was generated by site-directed mutagenesis. For conditional expression, a loxp-GFP-stop-loxp cassette was added before PRSS1, and then the entire loxp-GFP-stop-loxp-PRSS1 cassette was targeted to a mouse BAC (clone RP23–359O19, size 222kb) carrying the intact elastase I gene using recombineering as previously described.⁵ Transgenic mice were developed at the MD Anderson Cancer Center Genetically Engineered Mouse Facility by pronuclear microinjection. To specifically target conditional expression of the transgene to pancreatic acinar cells, PRSS1 transgenic mice were bred with elastase promoter regulated CreERT transgenic (BAC) mice and induced with tamoxifen (TM).⁵ All experimental mice, trypsinogens expressing mice (PRSS1 and PRSS1^R122H) and control mice expressing CreERT alone (BAC) were induced with oral gavage of TM (100 mg/kg/day in corn oil) for five consecutive days. For experimental analyses, sex and age-matched animals were sacrificed at specified times. Single transgenic littermates received the same treatment and served as controls. All mouse colonies were maintained in a specific pathogen-free barrier facility at the University of Texas M. D. Anderson Cancer Center. All animal studies conducted conformed to the Institutional Animal Care and Use Committee protocols.

Genotyping:

Mice tails (0.2cm) were snipped at 3 weeks of age. DNA was prepared for PCR by incubating the tail with 250 ul of 50mM sodium hydroxide at 95C for 15 min followed by neutralization with 25ul of 1M Tris (pH 8.0) containing 10mMEDTA. Following primers were used to detect Cre gene: forward - gcctgcattaccggtcga, reverse - tatcctggcagcgatcgc, and PRSS1: forward - ggacttgtagcagtggcctgc and reverse - tggagttcgtgaccgccgccg.

Induction of pancreatitis.

Transgenic trypsinogens mice and control mice were challenged with caerulein (12 intraperitoneal injections, 50 μg/kg every hour of caerulein solubilized in phosphate-buffered saline at a final concentration of 15μg/mL ) to induce acute pancreatitis. Five littermates in each group. The parameters of AP were assessed 12 hours after the last caerulein treatment. Control animals received an equal amount of saline. For induction of chronic pancreatitis, we used the intraperitoneal injection of lipopolysaccharide (LPS) (LPS E.coli 0111:B4, Sigma). Two weeks after induction with tamoxifen for 5 consecutive days, LPS was administered via i.p. injection twice a week (4 mg/kg dissolved in 100 ul sterile saline) for two consecutive weeks.

Diet-induced pancreatitis:

Adult (65–75-day-old) transgenic mice or control littermates were fed with previously described ¹² ethanol diet or control diet (Liquid Lieber-DeCarli’82 Ethanol Diet, Control Diet, Bio-Serv) prepared according to the manufacturer’s instructions. For the first 5-day accommodation period, mice were fed with a mixed diet containing 1/3 of ethanol diet and 2/3 of control diet. Then, mice were given 2/3 of ethanol diet and 1/3 of control diet for 4 weeks. Feeding with a high-fat diet (60% fat, Test Diet DIO 58Y1; Lab Supply, Fort Worth, TX) was performed for 4 weeks. After 4 weeks of treatments, animals were sacrificed and the pancreas was removed for evaluation.

Measurement of edema, serum amylase, and trypsin activity:

To determine the level of edema, we used the formula: whole Pancreata weight/Bodyweight x 100. Serum amylase activity was measured using the Phadebas test as recommended (Pharmacia Diagnostics, USA). Trypsin enzymatic activity and chymotrypsin enzymatic activity were measured in pancreatic homogenates using the trypsin activity kit (Abcam, ab102531, USA) and the chymotrypsin assay kit (Fluorometric, ab234051, USA), respectively.

Immunohistochemistry:

The pancreatic paraffin sections were first deparaffinized and antigen retrieval (1 x DAKO target retrieval solution, DAKO, Carpentaria, CA) was performed in a steamer for 20 minutes at 98°C. Endogenous peroxidase activity was blocked with H₂O₂ for 15 minutes, followed by washing and blocking for 1 hour with appropriate serum at room temperature. Next, primary antibodies were applied and incubated overnight (4°C). Frozen sections were fixed in cold acetone for 10 min at −20°C and blocked with proper serum for 1 hour at room temperature. The primary antibodies used for IHC are listed in Supplementary Tabl.1 (Supplementary Materials). After overnight incubation slides were washed with PBS and PBS containing 0.05% Tween20 and then incubated with appropriate secondary antibodies (Vectastatin Elite ABC Kit, Vector Laboratories). Positive labeling was detected by exposing the sample to DAB+Substrate system (DAKO). The slides were counterstained with Gill no. 3 hematoxylin (Sigma Aldrich). The quantification of CD45 positive cells and cleaved caspase 3 positive cells per visual field (40x) under a light microscope was performed including at least 3 cases/group (at least 10 images/case). The quantification of the phosphorylated form of histone 2A (γH2AX) was performed using ImmunoRatio software (3 cases/group, the same threshold for all 40x images, results expressed as the percentage of cells with positive nuclei staining/DAB area).

Collagen detection:

Collagen deposits were detected using Picro-Sirius Red Stain Kit (Abcam, ab150681) in paraffin-embedded pancreatic tissue. Briefly, after deparaffinized and hydration in distilled water, the slides were incubated for 1 hour with Picro-Sirius Red Solution and then quickly rinsed in 2 changes of Acetic Acid Solution. After dehydration in 2 changes of absolute alcohol, the slides were mounted.

Western blot Analysis:

Pancreas tissue was homogenized in lysis buffer and immunoblot analysis was performed with antibodies against HA-probe (#SC-7392, Santa Cruz Biotechnology, USA), human PRSS1 (#AP10735c, Abgent, Inc., USA), CTRC (ab35694, Abcam, UK), Bip (#3177, Cell Signaling Technology, USA), CHOP (#2895, Cell Signaling Technology, USA), IRE1-α (#3294, Cell Signaling Technology, USA), BCL-2 (#15071, Cell Signaling Technology, USA), p-JNK (#9255, Cell Signaling Technology, USA), P62 (#39749, Cell Signaling Technology, USA), anti-GAPDH antibody (#G8795, Sigma, USA) and secondary antibodies (anti-rabbit IgG #7074 or anti-mouse IgG #7076, Cell Signaling Technology, USA).

Real-time RT-PCR:

Total RNA was isolated using TRIzol reagent(15596–018, Invitrogen USA). One microgram of total RNA was reverse transcribed using QuantiTect Reverse Transcription kit (205313, Qiagen, USA). Real-time quantitative PCR using iQTM SYBR Green Supermix (170–8882, Bio-Rad, USA) or SensiMIX ™ Sybr No-Rox Kit (Bioline, Taunton, MA) were run on the Bio-Rad iCycler or LightCycler®480 (Roche, Indianapolis, IN). GAPDH gene was used as an internal RNA loading control. Primers used are listed in Supplementary Tabl.2 (Supplementary Materials).

TUNEL assay

Apoptosis in Tissue was detected by DeadEnd™ Fluorometric TUNEL System (#G3250,Promega, USA). Paraffin-Embedded Tissue was pretreated with removing paraffin by xylene and Rehydration by decreasing concentrations of ethanol. Then immerse slides in 4% formaldehyde and permeabilize the tissue with 100μl of a 20μg/ml Proteinase K solution. Incubate at room temperature for 8–10 minutes. Repeat Fix step and add 100μl Equilibration Buffer. Add 50μl of TdT reaction mix to the tissue on an area. Cover slides with plastic coverslips to ensure even distribution of the mix. Incubate slides for 60 minutes at 37°C in a humidified chamber. Stop Reaction by removing Plastic Coverslips. Immerse slides in 2X SSC for 15 minutes. Detect localized green fluorescence of apoptotic tissue by confocal fluorescence microscopy.

Statistical Analysis:

p-values were analyzed by the Kruskal-Wallis H test using IBM SPSS 20.0 and GraphPad 6 software. p-values less than 0.05 were considered statistically significant.

Results

Robust and efficient expression of human trypsinogens was achieved in mouse pancreatic acinar cells

In the current study, we aimed to assess the effects of elevated expression of human trypsinogen in mouse pancreas using two novel transgenic models. Wild-type human PRSS1 and R122H were tagged with hemagglutinin (HA) and cloned behind a lox-GFP stop-lox cassette (Fig. 1A, referred as PRSS1 and R122H mice in this study). These constructs with an internal ribosome entry site (IRES) were then inserted between translation stop and 3-UTR of the full-length elastase I gene contained in a bacterial artificial chromosome (Fig. 1B). Pancreatic specific expression was accomplished by crossing the mice with a BAC-Elastase-CreERT (BAC) line that expressed CreERT. After induction with tamoxifen (100mg/kg/d for 5d by oral gavage), the expression of human trypsinogen was determined by western blot using an anti-PRSS1 antibody (Fig. 1C). The efficiency of transgene expression was further confirmed with immunohistochemistry using an anti-HA antibody and expression of the transgenes were observed in 90% of the pancreatic acinar cells (Fig. 1D).

Fig. 1. Human cationic trypsinogen was conditionally expressed in transgenic mice.

A. Human cationic trypsinogen (PRSS1) and its R122H mutant form were tagged with HA and cloned behind a loxp-GFP loxp cassette for conditional expression upon Cre-mediated recombination. B. This cassette was targeted between the translation stop and 3-UTR of the mouse elastase gene in a bacterial artificial chromosome by recombineering and transgenic mice were developed by pronuclear injection. C. The floxed trypsinogen transgenic mice were crossed with BAC Elastase-CreERT(BAC) mice and induced with tamoxifen (100mg/kg body weight) for 5 days. Expression of PRSS1 trypsinogen was detected by Western blot using an anti-PRSS1 antibody. A lower band, which may be a cross-reacting mouse PRSS, was also noted. D. Immunohistochemical images showed efficient expression of the HA-tagged human trypsinogens.

Expression of human trypsinogens resulted in gene dosage- and time-dependent focal inflammation as well as alterations of cellular stress pathways

Despite elevated levels of trypsinogen expression, no spontaneous CP developed in any of the animals up to 30 weeks (Supplementary Fig. 1). Acinar cells were filled with zymogen granules and in the standard configuration and ducts and islets were unaffected. However, upon close examination, small focal areas of inflammatory (CD45 positive) cells were observed (Fig. 2A). Quantitation of CD45 cells indicated significantly elevated levels of CD45 positive cells in R122H mice compared to either BAC or PRSS1 mice (Fig.2B).

Fig. 2. Focal inflammation and stress responses occurred in mice expressing mutant R122H trypsinogen.

A. Despite no obvious spontaneous pancreatitis, small focal areas of immune cell infiltrations were more often found in the pancreas of mice expressing mutant R122H trypsinogen than in wild-type trypsinogen (PRSS1) and control mice (BAC). Representative immunohistochemical images for CD45 positive cells in control mice (BAC) and R122H and PRSS1 mice. B. Quantification under light microscope of CD45 positive cells for control (BAC), R122H and PRSS1 mice. C-E. Relative mRNA expression of nuclear factor (erythroid-derived 2)-like 2 (NRF2) (C), heat shock protein 70 (HSP70) (D) and CTRC (E) in R122H, PRSS1 and control mice (BAC) was measured by real-time RT-PCR. (*, p<0.05; **, p<0.01; ns, p>0.05).

To understand the lack of evident spontaneous chronic pancreatitis despite focal damage observed in R122H mice, we analyzed the levels of genes involved in cellular stress responses. Real-time RT-PCR indicated significantly elevated levels of nuclear factor (erythroid-derived 2)-like 2 (NRF2) (Fig.2C) and heat shock protein 70 (HSP70) (Fig.2D) in R122H mice compared to those in PRSS1 or BAC mice. Also, mice expressing either transgenic trypsinogen had lower levels chymotrypsin C (CTRC) at both mRNA (Fig. 2E) and protein levels (Supplementary Fig. 2A–B) compared to control mice. However, murine trypsinogen T7 (T7) and chymotrypsin B (CTRB) mRNA expression exhibited no significant difference in R122H, PRSS1 or BAC mice (Supplementary Fig. 2C–D). These adaptive responses could be expected to reduce the influence of increased trypsinogen levels. Indeed, basal trypsin activity was below detectable levels in these mice (Supplementary Fig. 2E).

To determine if increased gene dosage of human trypsinogens would have a more significant impact on the pancreas, we bred the PRSS1 and R122H mice to homozygosity, respectively. We observed an increase in focal acinar cell damage, fatty replacement, and focal fibrosis in the R122H mice and to a lesser extent in the PRSS1 mice (Supplementary Fig. 3). However, the full spectrum of chronic fibrotic pancreatitis did not develop even with increased gene dosage of either trypsinogen.

Elevated expression of either PRSS1 or R122H increased the severity of caerulein-induced acute pancreatitis

To determine whether the presence of the human PRSS1 or PRSS1^R122H would affect pancreatitis induced by experimental insults, we challenged the transgenic mice and control littermates with 12 hourly caerulein injections (50μg/kg/h) (Supplementary Fig. 4A). Pancreatitis was examined 12 hours after the last injection. Mice with either PRSS1 or PRSS1^R122H expression developed more severe acute pancreatitis as indicated by tissue damage, the presence of cleaved caspase 3 (Fig. 3A, Supplementary Fig. 4B) and TUNEL positive cells (Supplementary Fig. 5). These histological changes were accompanied by significantly increased levels of pancreatitis parameters including ER stress signaling pathway activation (Fig. 3B), impaired autophagy (Supplementary Fig. 4C), edema (Supplementary Fig. 4D), serum amylase (Fig. 3C), and histological scoring (Fig. 3D).

Fig. 3. High-dose of caerulein increased the severity of acute pancreatitis in transgenic trypsinogen expressing mice.

A. Representative H&E images showed more severity damages in trypsinogen transgenic (R122H and PRSS1) mice compared to control (BAC) mice after 12 repeated caerulein injections. Magnification 10x, scale bar 200μm. Immunohistochemistry indicated there were more cleaved caspase 3 positive cells in trypsinogens transgenic mice. B. ER stress-related molecules were detected by Western blot. C. Serum amylase levels were higher in trypsinogens transgenic mice after treatment with caerulein (*, p<0.05; **, p<0.01). D. Semiquantitative histological score evaluation including edema, inflammatory cell infiltration, and cell damage indicated elevated damage in trypsinogen transgenic mice (*, p<0.05). E. After caerulein treatment for 1 hour, the level of trypsin activity in PRSS1 and R122H mice demonstrated a significant increase compared to control (BAC) mice. Trypsin activity in R122H mice was significantly higher than that of PRSS1 mice (*, p<0.05).

One hour after caerulein treatment, trypsin activity in PRSS1 and R122H mice demonstrated a significant increase compared to control (BAC) mice (Fig. 3E). Moreover, trypsin activity in R122H mice was significantly higher than that of PRSS1 mice, which is in line with the severe inflammatory response in R122H mice. The level of trypsin activity was still high in both PRSS1 and R122H mice compared to BAC controls 12 hours after caerulein treatment (Supplementary Fig. 4E). Lower chymotrypsin activity in PRSS1 and R122H mice was consistent with the decreased CTRC expression levels in these mice (Supplementary Fig. 4F).

Expression of human R122H trypsinogen in mouse acinar cells sensitized the pancreas to LPS

Next, we assessed if chronic pancreatitis would be observed in the transgenic mice after administration of LPS (4mg/kg twice per week for 2 weeks) (Supplementary Fig. 6A), which mimic bacterial infection. The pancreas weights were significantly reduced only in R122H mice treated with LPS (Supplementary Fig. 6B). Histological examination showed that the development of chronic inflammatory changes was more prominent in R122H mice with comparatively few small lesions observed in PRSS1 mice and nearly no pathological changes in BAC control mice (Fig. 4A).

Fig. 4. Human mutant PRSS1 ^R122H transgenic expression caused extensive chronic inflammatory changes after LPS stimulation.

A. The representative histological images of pancreas of R122H, PRSS1, and BAC mice examined by H&E staining. The treatment with LPS led to development of prominent chronic inflammatory changes in R122H mice. In contrast, smaller chronic lesions were observed in PRSS1 mice and only small focal changes in BAC controls. The representative immunohistochemical images of pancreas indicating an increased infiltration of leukocytes (CD45 positive inflammatory cells), macrophages (F4/80 positive cells), and phosphorylated form of histone 2A (γH2AX) (reflecting DNA damage) that were prominent in R122H mice treated with LPS. B-C. Quantification under light microscope (40x) for control (BAC), R122H and PRSS1 mice treated with LPS of CD45 positive cells, F4/80 positive cells(B), and γH2AX positive cells(C) (**, p<0.01; ***, p<0.001; ns, p>0.05). Western blot of Bip, which is induced by ER stress, indicated that LPS treatment increased ER stress in R122H compared to PRSS1 mice (D).

Immunohistochemical analysis revealed increased infiltration of leukocytes (CD45 positive inflammatory cells) and macrophages (F4/80 positive cells) in R122H mice treated with LPS compared to control or PRSS1 mice (Fig. 4A–B). LPS treatment significantly increased DNA damage (nuclear expression of γH2AX) in the areas of chronic inflammation in R122H mice treated with LPS to a greater extent than that in BAC control or WT PRSS1 mice (Fig. 4A, C). LPS treatment also upregulated the expression of the proinflammatory cytokine interleukin-1 (IL-1β) (Supplementary Fig. 6C). Fibrosis markers of α-SMA and Sirus Red staining of collagen were significantly higher in R122H mice than in BAC control or PRSS1 mice (Supplementary Fig. 6D–E).

In LPS treated mice, trypsin activity was higher in PRSS1 and R122H mice than BAC controls (Supplementary Fig. 6F). There was a trend that trypsin activity was higher in R122H mice than PRSS1 mice, but the difference did not reach statistical significance. LPS treatment also led to increased expression of pancreatic ER stress and oxidative stress response-related genes including Bip (Fig. 4D), activating transcription factor 4 (ATF4) and CHOP (Supplementary Fig. 7A–B). In areas of tissue damage, increased acinar apoptosis (cleaved caspase 3 positive cells) was prominent (Supplementary Fig. 7C).

Expression of human R122H trypsinogen in pancreatic acinar cells sensitized the pancreas to ethanol and high-fat diet

To determine the interaction between expression of human trypsinogens and clinically relevant environmental insults, we assessed if administration of ethanol or a high-fat diet initiated CP in the trypsinogen transgenic mice. First, we determined if increased human trypsinogen expression could potentiate the effect of ethanol, a significant risk factor for CP.^13,14 We fed BAC and trypsinogen transgenic animals with a liquid Lieber-DeCarli diet for 4 weeks (Fig. 5A). Ethanol feeding did not cause apparent histological changes in the pancreas of BAC control mice (Fig. 5B). In contrast, ethanol feeding led to the development of chronic inflammatory changes including acinar atrophy and immune cell infiltrations in 74% (14/19) of R122H mice. Modest changes were observed in about 45% (10/22) of WT PRSS1 mice (Fig. 5B). Ethanol feeding also induced DNA damage (γ-H2AX) (Fig. 5B–C), elevated trypsin activity (Fig. 5D), and collagen deposition (Supplementary Fig. 8).

Fig. 5. Feeding with ethanol or high fat diet induced chronic inflammatory changes in pancreas of transgenic trypsinogens expressing mice.

A. The scheme of treatment. The trypsinogen expressing and control mice were fed with a Liber – De-Carli Diet for 4 weeks. B. The representative H&E images of pancreas of control (BAC) and trypsinogens (R122H and PRSS1) expressing mice fed with ethanol diet for 4 weeks. Corresponding representative immunohistochemical images showing focally localized DNA damage (nuclear staining for γH2AX) (magnification: 40 x, error bar: 100μm). C. Quantification of γH2AX positive cells under light microscope (40x) for control (BAC), R122H and PRSS1 mice fed with ethanol diet. Trypsin activity in control (BAC), R122H and PRSS1 mice (D) (*, p<0.05; **, p<0.01; ns, p>0.05).

Obese patients are more prone to develop pancreatitis and hyperlipidemia is one of the toxic risk factors of CP.^13,14 Therefore, we tested if transgenic expression of PRSS1 or R122H would exacerbate pancreatic damage with a high-fat diet (60% of fat) (Fig. 6A). After 4 weeks feeding of high fat diet, we observed focal chronic inflammatory changes as well as the presence of adipocytes in mice expressing human trypsinogens (Fig. 6B). Feeding a high-fat diet also induced DNA damage (Fig. 6B–C), increased cleaved caspase-3 positive cells (Fig. 6D), and deposition of collagen (Supplementary Fig. 8) with the most significant changes observed in the R122H expressing mice.

Fig. 6.

A. The scheme of treatment. The trypsinogen expressing and control mice were fed with a high fat diet (60% fat) for 4 weeks. B. The representative H&E images of pancreas of control (BAC) and trypsinogens (R122H and PRSS1) expressing mice fed with high fat diet for 4 weeks. Corresponding representative immunohistochemical images showing focally localized DNA damage (nuclear staining for γH2AX). C-D. Quantification of yH2AX positive cells (C) and cleaved caspase 3 (D) under light microscope (40x) for control (BAC), R122H and PRSS1 mice fed with high fat diet (*, p<0.05; **, p<0.01; ns, p>0.05).

Discussion

The current study investigated the influence of expressing human PRSS1 or the most common genetic mutation associated with the development of human HP, mutant PRSS1^R122H in a mouse model. Previous efforts to model human hereditary pancreatitis have had mixed results.^8–10 The major difference between previous models and the one described here was the genetic element used to drive transgene expression. All previous efforts utilized a short elastase promoter/enhancer with excellent specificity for pancreatic acinar cells, but with relatively low efficiency (10%–50%) that often results in mosaic expression.¹⁵ Moreover, this truncated promoter is much weaker than endogenous promoters.¹⁶ In the current study, a full-length elastase gene was utilized to achieve highly efficient and robust expression of the human transgenes. These new models showed clear differences between PRSS1 and R122H that were not seen in previous models. These differences were not due to different levels of expression, as the same genetic strategy was utilized for both and western blotting indicated equal levels of both isoforms. We also found synergy between the effects of expressing R122H mutant trypsinogen and clinically relevant stresses that were themselves insufficient to induce pancreatitis. Therefore, this new model mimics some of the fundamental characteristics of the human disease and should provide the opportunity to gain new insights into human hereditary pancreatitis.

Mice expressing R122H, while not developing spontaneous chronic pancreatitis, developed focal areas of injury identified by increased infiltration of a variety of inflammatory cells and apoptosis compared to PRSS1 or control mice. Adaptive responses including protective increases in HSP70 and NRF2, as well as a decrease in CTRC expression were observed which might reduce the influence of increased trypsinogen levels. However, when treated with low-dose LPS, R122H mice demonstrated exacerbated chronic inflammatory changes and increased expression of ER and oxidative stress response genes. Furthermore, when the mice were challenged with clinically relevant stresses, including ethanol or a high-fat diet, which alone had little or no effect on control animals, the animals expressing R122H responded with much more prominent pancreatic damage than that observed in mice expressing PRSS1. Whether the synergistic responses between R122H expression and the other stressors were due to the preexisting inflammatory microenvironment, or to the preexisting activity of stress pathways in acinar cells, or both, is unclear at this time. Considering that both of alcohol and high-fat diet were two significant high-risk factors for CP,^17–20 this mouse model may be a valuable model to investigate roles of environmental insults involved in the development of hereditary pancreatitis.

In contrast to the differences observed to clinically relevant insults, there was little difference in the responses observed in PRSS1 versus R122H mice when treated with caerulein hyperstimulation. Mice expressing PRSS1 or R122H developed more severe acute pancreatitis than controls, suggesting that elevated levels of trypsinogen contributed to the acute disease, presumably by providing higher amounts of precursor that could be activated and increase trypsin activity. In any case, responses to a strong, damaging, injury were different than those of less severe stimuli where clear differences existed between PRSS1 and R122H expressing mice.

The use of a mouse model to mimic human hereditary pancreatitis is complicated by differences in the structures of mouse and human trypsinogens that affect the actions of chymotrypsin C (CTRC).²¹ CTRC promotes both trypsinogen autoactivation and degradation depending on the site of cleavage. In human HP, the PRSS1 mutations interfere with CTRC-mediated cleavages leading to trypsinogen resistance to degradation and increased autoactivation with markedly elevated levels of trypsin activity.^3,21 However, the regulation of mouse trypsinogens autoactivation by mouse CTRC is different,²¹ and the effects of mouse CTRC on human PRSS1 remain uncertain. The observation that the R122H mutant caused more significant changes than those of PRSS1 under basal, elevated gene dosage, and in combination with environmental stressors, suggests that differences in autoactivation or degradation exist between these molecules in mice. Thus, this new mouse model mimics some crucial aspects of the deleterious results of this mutation observed in humans.

An important observation in the current study was that transgenic elevation of trypsinogens led to adaptive alterations in a variety of protective mechanisms. Interestingly, a dramatic and equivalent decrease in the level of CTRC occurred in mice expressing either PRSS1 or R122H. This decrease in CTRC suggests that mouse CTRC may be able to activate human trypsinogen, such that down-regulation of CTRC would be protective. Interestingly, a model with constitutively active trypsin (PACE-trypsin) previously developed in our lab did not show a decrease in CTRC (data not shown). In contrast to the equivalent effects of PRSS1 and R122H on CTRC levels, R122H but not PRSS1 expression increased the levels of HSP70 and NRF2. HSP70 has protective effects in studies of acute pancreatitis.²² NRF2 is an important regulator of the expression of several molecules involved in protection from ER and oxidative stresses.²³ It seems likely that these adaptive responses were sufficient to prevent spontaneous pancreatitis from developing due to the expression of R122H. It is also reasonable to propose that adaptive responses occurred in the previous transgenic models described by others, which may explain the general lack of damaging responses. ^8–10 The mechanisms involved these adaptive responses to the expression of the human trypsinogens are currently uncertain. However, increased levels of ER stress may well be involved. The expression of the human trypsinogens increased ER stress when treated with LPS, as indicated by elevated expression of the pro-apoptotic molecule CHOP and increased levels of ATF4. These effects may have been caused either by ER overload or possibly due to increased trypsin activity.⁵

The presented mouse models faithfully recapitulate much of the irregularity of human CP and help explain clinical observations. Low-grade inflammation observed in cells expressing human trypsinogens sensitized them to injury from clinically relevant stressors that separately do not cause pancreatitis in control animals. Taken together with other recent investigations of the role of trypsin in pancreatitis this study provides a more thorough picture. The disease is initiated by acinar cell injury that can be provoked by many different mechanisms.²⁴ Independent of the source of the injury, there are cellular responses that are protective and which tend to restore homeostasis. However, when the stresses overwhelm these protective mechanisms, additional pathways are activated that drive acinar cell damage and engage the inflammatory signaling. The elevation of inflammatory signaling changes the tissue microenvironment by increasing the presence or activity of various cells in the stroma. Thus, R122H expression, even while insufficient to initiate full pancreatitis alone, was able to remodel the pancreas in a manner that prepared it for further injury. This new model supplements the sparse and conflicting data on the effects of human mutant trypsinogen expression and provides support for a renewed focus on external environmental risk factors necessary for the development of pancreatitis under clinically relevant circumstances.

WHAT YOU NEED TO KNOW:

BACKGROUND AND CONTEXT:

Mutations in the trypsinogen gene (PRSS1) cause human hereditary pancreatitis. We studied the effects of expressing mutant forms of human PRSS1 in mice.

NEW FINDINGS:

Expression of a transgene encoding PRSS1^R122H in mice promoted inflammation and increase the severity of pancreatitis, compared with mice that express PRSS1 or control mice.

LIMITATIONS:

These studies were performed in mice. Findings from mice do not always replicate mechanisms in humans.

IMPACT:

These mice might be used as a model for human hereditary pancreatitis and can be studied to determine mechanisms of induction of pancreatitis by LPS, ethanol, or a high-fat diet.

Abbreviations List:

ATF4

Activating Transcription Factor 4

AP

acute pancreatitis

αSMA

α-smooth muscle actin

BAC

bacterial artificial chromosome

BCL-2

B-cell lymphoma-2

Bip

immunoglobulin-binding protein

CHOP

CCAAT-enhancer-binding protein homologous protein

CP

chronic pancreatitis

CTRB

chymotrypsin B

CTRC

chymotrypsin C

HA

hemagglutinin

H&E

hematoxylin and eosin

HP

hereditary pancreatitis

HSP70

Heat shock protein 70

IL-1

interleukin 1

IRE1-α

inositol-requiring enzyme 1-α

LPS

lipopolysaccharide

NRF2

Nuclear factor (erythroid-derived 2)-like 2

p-JNK

phosphorylated c-jun N-terminal kinase

PRSS1

human trypsinogen

R122H

human mutant trypsinogen

γH2AX

phosphorylated form of histone 2A

T7

trypsinogen 7

WT

wild-type