Caerulein-induced acute pancreatitis in mice that constitutively overexpress Reg/PAP genes
© Norkina et al. 2006
Received: 16 December 2005
Accepted: 15 May 2006
Published: 15 May 2006
Skip to main content
© Norkina et al. 2006
Received: 16 December 2005
Accepted: 15 May 2006
Published: 15 May 2006
The cystic fibrosis (CF) mouse pancreas has constitutively elevated expression of the Reg/PAP cell stress genes (60-fold greater Reg3α, and 10-fold greater PAP/Reg3β and Reg3γ). These genes are suggested to be involved in protection or recovery from pancreatic injury.
To test this idea the supramaximal caerulein model was used to induce acute pancreatitis in wild type and CF mice. Serum amylase, pancreatic water content (as a measure of edema), pancreatic myeloperoxidase activity, and Reg/PAP expression were quantified.
In both wild type and CF mice caerulein induced similar elevations in serum amylase (maximal at 12 h), pancreatic edema (maximal at 7 h), and pancreatic myeloperoxidase activity (MPO, a marker of neutrophil infiltration; maximal at 7 h). By immunohistochemistry, Reg3α was strongly expressed in the untreated CF pancreas but not in wild type. During pancreatitis, Reg3α was intensely expressed in foci of inflamed tissue in both wild type and CF.
These data demonstrate that the severity of caerulein-induced pancreatitis is not ameliorated in the CF mouse even though the Reg/PAP stress genes are already highly upregulated. While Reg/PAP may be protective they may also have a negative effect during pancreatitis due to their anti-apoptotic activity, which has been shown to increase the severity of pancreatitis.
There is a strong association of the Reg/PAP genes with pancreatic stress and injury, especially in response to pancreatitis [1, 2]. The role of these proteins has been investigated under various conditions. PAP appears to have an anti-inflammatory effect in pancreatic injury  as well as in inflammatory bowel disease . In vitro experiments demonstrated that PAP can inhibit TNF-α mediated inflammatory responses of macrophages  and of epithelial and endothelial cells . Experimental evidence also suggests that Reg/PAP are mitogenic and/or anti-apoptotic and enhance cell survival during development and in injured tissues [5–7]. The anti-apoptotic activity of Reg/PAP is of interest to pancreatitis, as experimental evidence shows that reduction of apoptosis can be associated with a worsened severity of pancreatitis [8, 9].
The Reg/PAP proteins are synthesized in a soluble form that, upon tryptic cleavage of an 11 amino acid N-terminal fragment, undergo conversion to fibrils . It has been proposed that the fibrils could form clot-like structures which, intracellularly would help control cell damage, and extracelluarly would preserve the integrity of the ductal epithelium during pancreatitis .
Because there are seven known Reg/PAP genes which span about 75 kb of mouse chromosome 6  plus Reg4 on chromosome 3 , it has not been practical to knockout all these genes together to study their functions. Although one knockout model exists for Reg1/PSP (pancreatic stone protein) , the presence of multiple isoforms of PAP (Reg3) may allow compensation for the loss of a single form and result in the potential absence of a phenotype. Therefore, we took advantage of the CFTR (cystic fibrosis transmembrane conductance regulator) null mouse (CF mouse), which has constitutively elevated expression levels of the Reg3α, PAP/Reg3β and Reg3γ genes . CFTR is a cAMP-activated chloride channel expressed in various epithelia of the body, and is especially important for fluid and bicarbonate ion secretion in the gastrointestinal system to neutralize gastric acid in the small intestine . In humans, the pancreas depends on CFTR for fluid and bicarbonate ion secretion and it is one of the most strongly damaged organs in CF when CFTR is absent or nonfunctional due to mutation . In contrast, the CF mouse pancreas is only mildly affected by loss of CFTR function [9, 17, 18]. A likely reason for this is that the mouse expresses a calcium-regulated chloride channel and the mouse pancreatic duct is not reliant on CFTR for proper function [19, 20]. A secondary effect of CF caused by the most common CFTR mutation, ΔF508, may be due to misfolded CFTR protein and subsequent activation of the endoplasmic reticulum unfolded protein response . This secondary effect of CF is absent in the mouse model as there is no Cftr mRNA or protein expressed in the null mouse .
To explain why there were changes in gene expression in the CF mouse pancreas, despite the fact that the mouse ductal system does not rely on CFTR for proper function, it was hypothesized that another CF-affected organ could be involved . It has been demonstrated that the luminal pH of the CF mouse duodenum is abnormally acidic [14, 23], and evidence was presented showing excess cAMP-mediated signaling by the small intestine to the pancreas, in an attempt to stimulate more bicarbonate ion secretion . It is known that cAMP potentiates calcium-mediated signaling in the acinar cell, and a chronically elevated cAMP signal would be expected to increase the secretory activity of the acinar cell. Chronic stimulation of the acinar cell could result in altered gene expression as an adaptive response to this stress on the secretory pathway. This idea was supported by the finding that pharmacological or genetic correction of the duodenal pH in CF mice largely reversed the overexpression of Reg/PAP genes .
For these reasons, the CF mouse pancreas presents a useful system to investigate the effects of Reg/PAP gene expression on the severity of pancreatitis. We postulated that in the CF mouse pancreas the constitutively elevated expression of Reg/PAP genes would lead to a less severe course of experimentally-induced acute pancreatitis. Contrary to the hypothesis, the results show that caerulein-induced pancreatitis is similar in CF mice despite high levels of expression of Reg/PAP genes.
CFTR(+/-) mice (cftr tm1UNC ) were obtained from Jackson Labs (Bar Harbor, ME, USA) and have been backcrossed on the C57Bl/6 background until congenic as previously described . CFTR(+/-) mice were bred to obtain wild type [CFTR(+/+)] and CF [CFTR(-/-)] mice. Mice of both sexes were used between 6–7 weeks of age and no gender-related differences in the measured parameters were observed. To prevent lethal intestinal obstruction, CF mice were fed ad libitum a complete elemental liquid diet (Peptamen; Nestle, Deerfield, IL, USA)  and wild type were fed the same diet to avoid potential dietary effects. All procedures were approved by the University of Kansas Medical Center Institutional Animal Care and Use Committee.
Pancreatitis was induced by intraperitoneal injection of 50 μg/kg caerulein (Sigma, St. Louis, MO, USA) in 0.9% NaCl . Controls were untreated or received equal volumes of 0.9% NaCl injected intraperitoneally. There were no differences between untreated and NaCl injected mice for any parameters measured, so they were grouped together for analysis. Samples were taken from mice 1 h after a single caerulein injection, and from mice at 7 h, 12 h, 3 days, and 7 days after the beginning of 7 caerulein injections at hourly intervals.
Animals were sacrificed with CO2 gas and trunk blood was collected for serum preparation. Serum amylase was determined using 4,6-ethyldiene(glucose)7-p-nitrophenyl-glucose-α, D-maltoheptaside (Raichem, San Diego, CA, USA). The pancreas was removed and a small portion was taken for preparation of total RNA using the Trizol reagent (Invitrogen, Carlsbad, CA, USA) following the supplier's instructions with modifications to enhance the integrity of pancreatic RNA. First, it was determined that immediate homogenization of the tissue in Trizol was required for isolation of high quality RNA, and that freezing of the tissue in liquid nitrogen followed by later processing was not effective at stabilizing the RNA. Second, a number of modifications to the isolation procedure were made. The Trizol was prechilled and homogenization was performed on ice using a 100:1 ratio of Trizol to tissue instead of the 10:1 ratio the instruction manual suggests. Pancreatic RNA was stable in the Trizol homogenates stored at -70°C until processing. To isolate total RNA, the following steps were used: (i) thaw sample on ice and centrifuge 12 k × g × 10 minutes at 4°C; (ii) transfer the supernatant to a fresh tube, add 200 μl chloroform, mix well by inversion for 15 sec, and incubate 2 minutes at room temperature; (iii) centrifuge 12 k × g × 15 min at 4°C; (iii) transfer the upper phase to a fresh tube, add 0.5 ml isopropanol, mix by inversion and vortexing, and incubate at room temperature for 10 min; (iv) centrifuge 10 k × g × 10 min at 4°C; (v) discard the supernatant and resuspend the pellet in 1 ml 75% ethanol; (vi) centrifuge at 7.5 k × g × 5 min at 4°C; (vii) remove supernatant, dry under vacuum, and resuspend pellets in diethylpyrocarbonate-treated water.
Another piece of tissue was fixed overnight in 4% paraformaldehyde in phosphate-buffered saline for histology; and the remaining small piece was blotted to remove excess buffer and weighed, and rapidly frozen in liquid nitrogen. The frozen tissue was lyophilized to dryness (3 days) and weighed again to determine percent wet weight as a measure of tissue edema. The dried pancreas samples were homogenized and extracted with hexadecyltrimethylammonium-bromide followed by determination of myeloperoxidase (MPO) activity as an estimate of neutrophil infiltration as described .
Total RNA prepared from the pancreas was used to measure gene expression as previously described . Briefly, real time RT-PCR was performed on an iCycler (Bio-Rad, Hercules, CA, USA) with a one-step RT-PCR kit (Qiagen, Valencia, CA, USA) that employs the double stranded DNA-binding dye SYBR Green. Serial dilutions of cloned plasmid DNAs for the genes of interest were used to generate standard curves. The genes measured were: Reg3α (GenBank accession NM_011259), PAP/Reg3β (NM_011036), and Reg3γ (NM_011260). Levels of 18S rRNA measured by quantitative RT-PCR and total RNA measured by OD260 nm were used for normalization. Because there were changes in levels of 18S rRNA (see Results), data are expressed as copies of specific mRNAs per μg total RNA.
For the preparation of cDNA coding for mouse Reg3α(GenBank accession NM_011259), mRNA was extracted and reverse transcribed from mouse pancreas as described . The cDNA was amplified by PCR using a forward primer (5'-GAA GAA GGG GTA TCT CTC GAG AAA AGA CAA GGT GAA GAC TTC CAG AAG G-3') that includes an XhoI restriction site followed by a sequence coding for the Kex2 cleavage site and a Reg3α specific sequence at the 3'-end (shown in bold). The reverse primer (5'-CTA CTG CTT GAA CTT GCA GAC-3) annealed to the 3'-untranslated region of the cDNA. The cDNA was subcloned into pcR2.1-TOPO (Invitrogen) and sequenced to verify the identity of the gene. The cDNA was then excised from the plasmid by restriction digestion with XhoI and NotI and subcloned into pPIC9. The plasmid was then transformed into Pichia pastoris strain KM71 and recombinant protein was generated and purified as described . The recombinant protein consisted of amino acid 25 through the end of Reg3α and was used to raise rabbit antisera. An ELISA for Reg3α was then established using antibodies generated in rabbits and guinea pigs as described previously [28, 29]. Paraffin sections (5 μm) of pancreas were processed for immunohistochemistry of Reg3α using the Vectastain ABC technique (Vector Labs, Burlingame, CA, USA).
Comparisons were made with Systat software (Systat, Chicago, IL, USA). Initial analysis showed that the values were not normally distributed, so the nonparametric Kruskal-Wallis one-way analysis of variance was used. P values of less than 0.05 were considered significant. Values are plotted as means and sem.
Pancreatitis is accompanied by tissue edema which was measured by determining the water content of the tissue. In the wild type mouse pancreatic edema was maximal at 7 h (Fig. 1B). In the control CF pancreas the tissue water content was already slightly greater than wild type (Fig. 1B). The water content of the CF pancreas was greater after a single caerulein injection and at 7 d, after resolution of pancreatitis, compared to the corresponding wild type tissue (Fig. 1B).
Injury to acinar cells in pancreatitis causes release of cytokines and recruits neutrophils which can be measured as an increase in tissue myeloperoxidase (MPO) activity . As shown in Fig. 1C, control wild type and control CF mice both have low pancreatic MPO levels. One hour after a single caerulein injection the MPO level in the CF tissue was slightly increased over control while the wild type level was not yet changed (Fig. 1C). Between 7–12 h after caerulein injections, there were significant increases in MPO activity in both wild type and CF pancreas (Fig. 1C). The elevated MPO levels persisted longer in the CF mice, remaining significantly increased at 3d after caerulein administration (Fig. 1C). In both wild type and CF mice, there was a return to control levels by 7d after caerulein injection.
In the CF pancreas, the Reg/PAP genes also exhibited decreased expression after induction of pancreatitis, and the minima were at 7–12 h for all three genes (Fig. 3). Only Reg3γ was significantly overexpressed relative to CF control levels, and this occurred at 3d after caerulein administration (Fig. 3C).
In contrast to the wild type control, the untreated CF pancreas was strongly immunoreactive for Reg3α with labeling of most acinar cells but not islets (Fig. 5F). Intense labeling was often observed in the dilated acinar luminal spaces (Fig. 5F, M) which are characteristic of this mouse model of CF . During pancreatitis in CF mice, Reg3α staining was variable and there were some strongly stained regions adjacent to weakly stained areas (Fig. 5G; 1 h) or more uniformly labeled tissue (Fig. 5H; 7 h). At higher magnification, the intense labeling can be observed in the apical portion of acinar cells as well as in a ductal lumen (Fig. 5N; arrow, CF 1 h). Similar to wild type mice, intense immunoreactivity in CF mice with pancreatitis was frequently observed in inflamed areas (Fig. 5I; CF, 3d). By 7d after caerulein administration, the CF tissue again had a fairly homogeneous labeling of acinar cells but not islets (Fig. 5J). The antibody staining was specific as shown by omission of the primary antibody (Fig. 5O; CF, 3d); this control sample was from a section adjacent to that shown in Fig. 5I, which was intensely stained when the specific antibody was included.
It is widely believed that overexpression of Reg/PAP genes constitutes a protective mechanism in response to pancreatic injury or stress [1, 33, 34]. While it is clear that the Reg/PAP genes are upregulated in response to pancreatic damage in pancreatitis, it is still uncertain what they do in pancreatitis. The most direct evidence comes from an antisense approach which indicates that they can limit the severity of experimental pancreatitis . Using the sodium taurocholate pancreatic ductal infusion model of pancreatitis in rats and an antisense oligonucleotide that targets PAP1, PAP2, and PAP3 mRNAs, it was shown that the severity of pancreatitis was increased. In the antisense treated rats with pancreatitis there was 3.5-fold greater serum amylase, increased tissue edema and histological inflammatory scores, and elevated levels of IL-1α, IL-1β, and IL-4 mRNAs in peripheral blood monocytes compared to rats treated with a sense oligonucleotide . The antisense oligonucleotide, but not control sense oligonucleotide, inhibited the induction of the PAP genes during pancreatitis by 40–60%. Thus, a modest inhibition of PAP gene upregulation during taurocholate-induced pancreatitis increased the severity of the disease response demonstrating the beneficial effect of these genes during pancreatic stress.
These genes are also known to be expressed in tissues other than the pancreas, and in many cases they were shown to be mitogenic or anti-apoptotic, and they promote cell survival in normal development as well as in tissue injury [5, 6, 7, 36, 37]. Their role in inhibiting apoptosis is especially interesting in relation to pancreatitis, as evidence indicates that apoptotic cell death is related to less severe inflammation during pancreatitis as compared to necrotic cell death [8, 9].
We previously showed that the CF mouse pancreas overexpresses Reg3α, PAP/Reg3β, and Reg3γ . In general the Reg/PAP genes are coordinately upregulated by stress to the pancreas  but this does not always hold as shown by microarray analysis of the CF mouse pancreas. Two of the five members of the Reg/PAP family genes represented on the microarray, Reg1/PSP and Reg2/PTP (pancreatic thread protein), were not overexpressed .
In a recent study, it was shown that several cytokines and inflammatory markers are upregulated in the CF mouse pancreas, including ICAM-1, TNF-α, IL-6, and IL-1β . PAP expression in AR42J cells can be activated by TNF-α  and it has been shown that the promoter of the PAP gene can drive expression of a reporter construct through IL-6 stimulation . The elevated cytokine expression may explain the constitutive overexpression of the Reg/PAP genes in the CF mouse pancreas.
We postulated that the CF mouse with already elevated Reg/PAP expression would have less severe caerulein-induced pancreatitis than wild type mice. Even so, the time course of pancreatitis was similar comparing CF and wild type mice. It should be noted, however, that the CF pancreatic tissue content of amylase and other zymogens is about 50–75% that of wild type [9, 24, 40]. Accordingly, if the serum amylase levels were normalized to tissue content, the control CF mice would already have elevated serum amylase compared to wild type, and would also have greater pancreatitis-associated serum amylase levels. In a recent report, it was shown that supramaximal caerulein induced a greater degree of acute pancreatitis in CF mice compared to wild type, based on histological scores, edema, and pancreatic MPO levels . They used 12 hourly injections of caerulein which is more harsh than the 7 injections we used in our study and this might explain the pronounced difference in severity of pancreatitis compared to our results.
The lack of protection from pancreatitis by constitutive overexpression of Reg/PAP genes in the CF mouse pancreas is conceivably due to multiple factors. First, there is already a mild inflammation of the tissue [9, 14] which may predispose to a more severe pancreatitis and could counteract the protective effect of Reg/PAP. Second, the fact that not all the Reg/PAP genes were constitutively elevated in the CF tissue may indicate that the additional overexpression of Reg1/PSP and Reg2/PTP is needed. It has been shown that caerulein-induced pancreatitis protects against death in animals subsequently treated to induce another experimental model of pancreatitis, infusion of the pancreatic duct with taurocholate . Because PAP expression was increased by the caerulein treatment, it was suggested that PAP could mediate the protective effect observed. It is not clear from that study if all the Reg/PAP genes were upregulated and if overexpression of a subset of the genes such as occurs in the CF mouse would be protective.
A third possibility is that the time course of Reg/PAP gene expression may be of importance in their ability to protect. There is some confusion in the literature on the time course of Reg/PAP gene expression in pancreatitis. A previous study showed a transient downregulation of PAP/Reg3β  and our current results confirm this result and extend it to Reg3α and Reg3γ. There is a significant decrease in mRNAs for these genes at very early times after caerulein injection in both wild type and CF mice, followed by recovery and overexpression at later times as the tissue recovers from pancreatitis. These data are in apparent contradiction with published work on Reg/PAP gene expression in pancreatitis. Although the Reg/PAP genes have been referred to as 'acute phase' genes, the majority of studies have measured their expression levels at 12–24 hours or later after induction of acute pancreatitis . The transient downregulation of the Reg/PAP genes in mice with caerulein-induced pancreatitis occurs during the first hours, followed by a rebound and later overexpression. Thus, although these genes are overexpressed in caerulein-induced pancreatitis, they are first transiently downregulated at short times after the insult. It should be noted that the short times have been analyzed only using mice and not rats, and it is conceivable that there will be species differences. The disparity between our results which show a transient decrease in Reg/PAP gene expression in the mouse, and those in the literature showing only the later increased expression during pancreatitis in rats, is likely only an apparent one. The differences are probably due to the different time points evaluated and maybe also reflect species variations.
A fourth possibility for failure to protect the pancreas may lie in the idea that apoptotic cell death in pancreatitis results in less severe damage compared to necrotic cell death  and that the Reg/PAP genes have been demonstrated to be anti-apoptotic [5–7]. It may be that these gene products inhibit programmed cell death of damaged acinar cells in the CF pancreas and result in greater necrotic death during pancreatitis and greater damage. This idea is supported by the recent report of DiMagno et al., who showed decreased apoptosis and more severe inflammation in caerulein-induced pancreatitis in CF mice .
Despite the finding that Reg3α was highly expressed in the control CF mouse pancreas and that protein levels did not change dramatically during pancreatitis, there were differences in the expression pattern by immunohistochemistry. In the control CF pancreas areas of intense immunoreactivity were associated with areas containing dilated acinar lumina which are characteristic of this mouse model of CF . In addition, similar to wild type mice, foci of intense Reg3α immunoreactivity in the CF mouse pancreas occurred in regions of inflammation and leukocyte infiltration. So, although the level of protein expression did not change dramatically, the localization of intense expression was associated with the more affected areas of the tissue. This finding is consistent with the connection of Reg/PAP expression and damage to cells of the pancreas. Whether focal increases in expression have a positive or negative effect on limiting damage or the rate of recovery remains an open question.
In summary, the CF mouse exhibits a similar severity of caerulein-induced pancreatitis compared to wild type despite the constitutive high expression of three of the Reg/PAP genes. The constitutive upregulation of Reg/PAP expression in the CF mouse pancreas may result in balanced protective (anti-inflammatory) and aggressive (anti-apoptotic) factors which could account for the lack of effect on the severity of pancreatitis.
cystic fibrosis transmembrane conductance regulator
enzyme-linked immunosorbent assay
intercellular adhesion molecule 1
pancreatitis associated protein
pancreatic stone protein
pancreatic thread protein
tumor necrosis factor α
We thank Larysa Stroganova for diligent maintenance of the CF mouse colony and accurate genotyping, Rosetta Barkley for histological preparations, Eileen Roach for assistance with figure preparation, and Martha Bain for ELISA determinations. This work was supported by National Institutes of Health grant DK56791 to RCD and by a Swiss National Science Foundation grant to RG.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.