IL‐4/STAT6‐signaling Influences Local Inflammation and Regeneration Processes During Acute Pancreatitis and Promotes Fibrosis by a Direct Activation of Pancreatic Fibroblasts During Chronic Pancreatitis
Hager ElSheikh, Annika Salvers, Judith Piesker, Shenja Buchholz, Rabea Lange, Sabine Ameling, Leif Steil, Elke Hammer, Georg Homuth, Uwe Völker, Barbara M. Bröker, Frank U. Weiss, Matthias Sendler, Juliane Glaubitz
TL;DR
This study shows that IL-4/STAT6 signaling reduces inflammation and supports healing in acute pancreatitis but has limited effect on fibrosis in chronic pancreatitis.
Contribution
The study reveals the dual role of IL-4/STAT6 signaling in modulating inflammation and regeneration in acute pancreatitis and its limited impact on fibrosis in chronic pancreatitis.
Findings
IL-4/STAT6 signaling suppresses pro-inflammatory responses and promotes acinar cell regeneration in acute pancreatitis.
STAT6 deletion leads to mixed macrophage phenotypes and prolonged inflammation in pancreatitis.
STAT6 signaling directly activates fibroblasts to produce extracellular matrix components but only moderately reduces fibrosis in chronic pancreatitis.
Abstract
In both acute and chronic pancreatitis, the local pancreatic immune response has a significant influence on the course of the disease. Utilizing a murine model deficient in STAT6, we demonstrate that the IL‐4/IL‐13/STAT6‐signaling pathway, which is a central component of the type 2 immune response, is rapidly activated during acute pancreatitis, thereby suppressing the pro‐inflammatory reaction and dampening the inflammation‐driven disease severity. The deletion of STAT6 in Stat6‐/‐ knock‐out mice surprisingly do not affect the numbers of CD206+ macrophages nor the release of TGF‐β. Notably, Stat6‐/‐ macrophages in CP are characterized by a high expression of the M2 markers Fizz1, Ym1 and Arg1, but also showed pro‐inflammatory properties, indicated by the expression of Nos2, Il1b and Mmp9. This mixed functional phenotype corresponded to a prolonged pro‐inflammatory response in…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5
FIGURE 6
FIGURE 7
FIGURE 8- —Deutsche Forschungsgemeinschaft10.13039/501100001659
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsPancreatitis Pathology and Treatment · IL-33, ST2, and ILC Pathways · Phagocytosis and Immune Regulation
Introduction
1
Acute pancreatitis (AP) is the most frequent non‐malignant disease of the gastrointestinal tract that requires in‐patient hospital admission. In the majority of patients, the disease takes a mild, self‐limiting course, but in approximately 20% of all cases a severe necrotizing form of pancreatitis develops, which is associated with serious complications such as organ failure or bacterial infection of the pancreatic necrosis [1,2]. One third of patients experience recurrent episodes of AP, which can progress to a chronic form of pancreatitis (CP) [3]. During the manifestation of CP, exocrine tissue is successively replaced by fibrotic tissue, which ultimately results in exocrine and endocrine insufficiency. CP is associated with a considerable reduction in quality of life, a significantly reduced life expectancy as well as an increased risk of developing pancreatic cancer [4,5]. To date, no causal therapy for either acute or chronic pancreatitis has been identified.
AP is defined as self‐digestion of the pancreas by its own proteases. Trypsinogen is processed to trypsin by the lysosomal protease cathepsin B and activates a cascade of digestive enzymes within the pancreatic acinar cells [6]. The proteolysis in acinar cells results in the activation of the transcription factor NFκB in acinar cells and in necrotic cell death [7]. The release of damage‐associated molecular patterns (DAMPs) by the necrotic cells [8] then triggers an intense local pro‐inflammatory immune response [9,10]. Pro‐inflammatory classical activated macrophages play a central role in this process and turn the local immune reaction into a systemic immune response that activates both the innate and the adaptive immune system [9,11,12]. The systemic immune reaction defines the severity of the disease and is decisive for systemic complications such as organ failure or the infection of pancreatic necrosis with commensal intestinal bacteria [10,13].
Recurrent AP and the development of CP are characterized by a local fibro‐inflammatory immune response involving pancreatic macrophages, T cells, innate lymphoid cells (ILCs) as well as pancreatic fibroblasts and pancreatic stellate cells [14, 15, 16]. After the acute episode the local pro‐inflammatory immune reaction shifts towards an anti‐inflammatory type 2 immune response [12]. Pancreatic GATA3^+^ Th2 cells and type 2 innate lymphoid cells (ILC2s) release IL‐4 and IL‐13 which induce the polarization of CD206^+^ alternatively activated macrophages. These macrophages release TGF‐b and trigger the production of extracellular matrix proteins by pancreatitis‐associated fibroblast (PAFs) [14,17]. This progressive fibrosis replaces the exocrine tissue successively after each relapse of AP. The immune response has been shown to have a critical impact on organ remodeling through the activation of PAFs and thus could represent a therapeutic target for an anti‐fibrotic and organ preserving therapy of CP [12,15].
The transcription factor STAT6 belongs to the family of Signal Transducer and Activator of Transcription (STAT) proteins and is activated by the cytokines IL‐4 and IL‐13. The receptor for both cytokines contains the IL‐4Rα subunit and initiates the phosphorylation of STAT6 via Janus kinase signaling [18]. Homodimers of pSTAT6 then translocate into the nucleus and regulate the transcription of a variety of genes, including GATA3. STAT6 drives the differentiation of alternatively activated M2 macrophages as well as ILC2s and Th2 cells [19] and is an essential mediator of the type 2 immune response. As indicated in previous studies, the differentiation of Th2 cells occurs early after induction of AP, in addition to a systemic activation of regulatory T cells [9,11]. Parallel to the local pro‐inflammatory immune response, a systemic counter‐regulation is initiated, which is driven by an anti‐inflammatory type 2 immune response.
In the present study, we investigated the influence of the STAT6‐mediated type 2 immune response in mouse models of acute and chronic pancreatitis, using STAT6 deficient mice. We could show that STAT6‐signaling significantly reduces the inflammation in the model of AP and thus attenuates the severity of the disease. In CP the role of IL‐4/IL‐13 STAT6‐signaling is more complex: STAT6 activation in PAFs has a direct impact on the composition of the extracellular matrix proteins, indicating a mechanism of tissue fibrosis that is independent of alternatively activated macrophages. Similar to its role in AP, STAT6 reduces inflammation and favors acinar cell regeneration in CP.
Results
2
STAT6 is Necessary for IL‐4/IL‐13 Signaling
2.1
In the first experiment, we analyzed target genes of IL‐4/IL‐13 signaling and tested whether IL‐4/IL‐13 activates also STAT6‐independent signaling pathways. Fluorescent co‐labeling of CD206 and IL‐4R showed the expression of the IL‐4 receptor on bone marrow derived macrophages (BMDMs) of wild type mice (Figure 1A). Next, we stimulated BMDMs from wild type and Stat6‐/‐ mice with 10 ng/mL each of IL‐4 and IL‐13 for 24 h (Figure 1B). Western blot analysis confirmed the deficiency of STAT6 in BMDMs of Stat6‐/‐ mice and showed STAT6 phosphorylation after stimulation with IL‐4/IL‐13 in wild type mice (Figure 1C). Transcriptional profiling of BMDMs of wild type and Stat6‐/‐ mice was performed using Affymetrix Clariom S mouse arrays. The volcano‐plot illustrates the number of up‐ and downregulated genes in wild type BMDMs after stimulation with IL‐4/IL‐13 (Figure 1D,E). In contrast to wild type BMDMs, which showed a significant upregulation of 1012 transcripts and a significant downregulation of 1039 transcripts, cells isolated from Stat6‐/‐ animals upregulated only one transcript (Saa3) (Figure 1F). A detailed analysis of the IL‐4/IL‐13‐regulated genes in wild type BMDMs revealed genes which are involved in the differentiation of alternatively activated macrophages such as Retnla, Chil3, Arg1 or Mrc1. Genes like Socs1 and Socs2, which act as suppressors of cytokine signaling, were also induced, which could explain the anti‐inflammatory phenotype of the IL‐4/IL‐13 stimulated BMDMs and the reduced expression of pro‐inflammatory genes such as Il1b, Mmp9 and Il18 (Figure 1G). We performed an Ingenuity Pathway Analysis (IPA) to identify involved upstream regulators (Supplemental Figure S1A–F). The analysis demonstrated that the IL‐4/IL‐13 STAT6 signaling axis was activated, while pro‐inflammatory signaling pathways, like TLR‐MYD88 signaling, as well as transcriptional regulators, including RELA, IRF3, and IRF7, were downregulated in the presence of IL‐4/IL‐13. A pathway analysis based on transcriptome data, demonstrated that the “Macrophage Alternative Activation Signaling Pathway” was significantly induced whereas pro‐inflammatory pathway such as “Interferon alpha/beta signaling”, “Toll‐like Receptor Signaling” or “IL‐1 Signaling” were significantly repressed (Supplemental Figure S2). In vivo a multitude of factors act on macrophages, especially the phenotype of M2 macrophages include several subtypes and a complex signaling network which is involved. Especially IL‐10 but also IL‐6 contributes to the M2 phenotype and enhances the expression of M2 markers like Arg, Mrc1, Ym1 and Fizz. An interesting observation was that Tgfbi, a hallmark factor released by alternatively activated macrophages, was significantly induced by IL‐10 but not IL‐4/IL13 (supplementary Figure S3). These experiments demonstrate that the IL‐4/IL‐13 dependent differentiation of BMDMs into alternatively activated macrophages is dependent on the presence of STAT6. STAT6‐deficient BMDMs did not respond to stimulation with IL‐4/IL‐13. Nevertheless, other factors, such as IL‐10, are also known to contribute to the phenotype of alternatively activated macrophages.
STAT6 is necessary for IL‐4/IL‐13 signaling. (A) Co‐labeling of CD206 and IL‐4R on isolated bone marrow derived macrophages (BMDMs). (B) BMDMs were isolated from wild type and Stat3‐/‐ mice and stimulated with 10 ng/mL each of recombinant murine IL‐4 and IL‐13. (C) Western blot analysis showed the absence of STAT6 in Stat6‐/‐ mice and the phosphorylation of STAT6 in WT cells following stimulation with IL‐4/IL‐13. GAPDH is referred as loading control. Microarray‐based transcriptome analysis with Clariom S mouse arrays was performed using RNA prepared from wild type BMDMs and Stat6‐/‐ BMDMs in the absence and presence of 10 ng/mL each of IL‐4/IL‐13 (n = 3). Volcano plots illustrate the differentially expressed genes in wild type BMDMs (D) and BMDMs from Stat6‐/‐ mice (E) after stimulation with IL‐4/IL‐13. (F) Pie chart illustration of the number of significantly differentially expressed genes in wild type and Stat6‐/‐ macrophages. (G) Heat map displaying effects of IL‐4/IL‐13‐STAT6‐signaling on expression of selected genes. Significance was tested by using analysis of variance (ANOVA) with e‐Bayesian correction. Statistically significant differential gene expression in terms of transcript levels was defined as a fold change of > 1.5‐fold between the compared conditions and an FDR < 0.05.
STAT6 Limits Local Inflammation and Organ Damage in Acute Pancreatitis
2.2
In the next step we evaluated STAT6‐signaling in a mouse model, where AP was induced by hourly injections of caerulein in wild type and Stat6‐/‐ mice (Figure 2A). H&E staining of pancreatic tissue showed increased tissue damage in STAT6‐deficient mice 8 and 24 h after the onset of pancreatitis (Figure 2B). Histological scoring confirmed a significantly increased disease severity in Stat6‐/‐ mice compared to wild type animals (Figure 2C). In Stat6‐/‐ mice we observed significantly elevated numbers of necrotic cells and infiltrating leukocytes in the pancreas, whereas the edema formation was similar to wild type mice (Figure 2D). Measurement of amylase and lipase activity in serum of mice gave further evidence that the disease severity is increased in Stat6‐/‐ animals 8 and 24 h after onset of pancreatitis (Figure 2E). It has been established that neutrophil granulocytes exert a significant influence on both the local and systemic severity of disease [20]. To see whether the deficiency of STAT6 would affect neutrophil migration, we investigated the myeloperoxidase (MPO) activity in lung tissue, but observed no differences between wild type and Stat6‐/‐ mice (Figure 2F), and also, the H&E lung histology showed no difference (Figure 2G). Immunofluorescent labeling of MPO was utilized as a marker for neutrophil granulocytes within pancreatic tissue at 8 and 24 h following pancreatitis induction and exhibited no differences between wild‐type and Stat6‐/‐ animals (Figure 2H,I). The lack of STAT6 resulted in a significantly increased disease severity of AP, which suggests a protective, anti‐inflammatory role of IL‐4/IL‐13/STAT6‐signaling in the early acute phase of pancreatitis.
STAT6 limits local inflammation and organ damage in acute pancreatitis. (A) Acute pancreatitis was induced in Stat6‐/‐ and wild type mice by hourly i.p. injections of caerulein. (B) H&E staining of pancreatic tissue shows pancreatic damage. (C) Pancreatic damage evaluated by histology scoring was significantly elevated in Stat6‐/‐ mice. (D) Stat6‐/‐ mice have higher numbers of infiltrating immune cells as well as necrotic acinar cells than WT mice, whereas edema was similar both groups. (E) Measurement of amylase and lipase activity in serum. (F) Myeloperoxidase activity in lung tissue and lung histology visualized by H&E staining (G), both measures of systemic inflammation, were not affected by Stat6 deficiency. (H) Labeling of MPO in pancreas tissue of wild type and Stat6‐/‐ mice during acute pancreatitis. (I) Quantification of MPO+ cells in pancreas of wild type and Stat6‐/‐ mice. All results shown are based on n = 10 animals per group (8 and 24 h). Data in C‐F and I are shown as mean +/‐ SEM. Differences were tested for statistical significance by unpaired students t‐test for independent samples and significance levels of p < 0.05 are marked by an asterisk.
Macrophage Mediated Inflammation During Acute Pancreatitis in Stat6‐/‐ Mice
2.3
To better characterize the observed differences in disease severity, we investigated the local immune response in more detail. It is well established that the local pro‐inflammatory reaction, which is mainly mediated by pancreatic macrophages [8], is responsible for acinar cell damage and induction of necrosis [21,22]. Immunofluorescent labeling of pancreatic tissue showed a significantly increased number of CD68^+^ macrophages in Stat6‐/‐ mice 8 and 24 h after the onset of disease (Figure 3A,B). Additionally, the number of infiltrating leukocytes marked by CCR2 was elevated 8 h after induction of pancreatitis in Stat6‐/‐ mice (Figure 3A,C). Additional functional analyses in sections of 8 and 24 h pancreatitis tissue revealed an increase in the numbers of 206^+^ alternatively activated macrophages, which, however, was significantly decreased or delayed at 8 h in the Stat6‐/‐ mice. (Figure 3D,E). Elevated numbers of infiltrating cells, but reduced numbers of CD206^+^ alternatively activated macrophages indicate a more pronounced acute pro inflammatory response in the STAT6 deficient mice, and IL‐6 in serum was also significantly elevated during the acute phase in these mice (Figure 3F). At 8 h of pancreatitis, IL‐1β, a pro‐inflammatory cytokine released by classical activated macrophages, was also elevated in serum of Stat6‐/‐ mice (Supplementary Figure S4A). Furthermore, chemokines CCL5, CXCL5 and CXCL11, which are responsible for the recruitment of leukocytes to the site of inflammation, were elevated in the serum of Stat6‐/‐ animals whereas STAT6‐inducible chemokines were decreased (Supplementary Figure S4B). Next, we evaluated the effect of the IL‐4/IL‐13 STAT6 axis on Toll‐like receptor signaling in mouse macrophages in vitro. Therefor we isolated BMDMs from wild type and Stat6‐/‐ mice, exposed them to 10 ng/mL each of IL‐4/IL‐13, to 10 ng/mL LPS, or to both. In wild type cells quantitative RT‐PCR showed a significant co‐stimulatory effect of IL‐4/IL‐13 and LPS on the marker gene expression of alternatively activated macrophages, Arg1 and Ym1. In STAT6‐deficient cells neither IL‐4/IL‐13 nor LPS induced Arg1 and Ym1 expression. In contrast, marker genes of a pro‐inflammatory macrophage phenotype, such as Mmp9 and Il1b, were induced by LPS in both wild‐type and STAT6‐deficient cells. IL‐4/IL‐13 treatment completely repressed the transcription of those genes in wild type but not at all in STAT6‐deficient macrophages (Figure 3G). In the next step we analyzed the anti‐inflammatory effect of IL‐4/IL‐13 under pancreatitis conditions. We isolated BMDMs from wild type mice and co‐incubated them with isolated acinar cells from wild type mice that had been stimulated with 0.001 mM CCK. Quantitative RT‐PCR analysis of Arg1, Ym1, Mmp9 and Il1b confirmed the significant anti‐inflammatory effect of the IL‐4/IL‐13 treatment (Figure 3H). Furthermore, the release of pro‐inflammatory cytokines IL‐6 and IL‐12p40 as well as the chemokine CXCL1 was significantly reduced in BMDMs co‐incubated with CCK stimulated acini in the presence of IL‐4/IL‐13. In addition, the release of the anti‐inflammatory cytokine IL‐10 and the chemokine CCL17 was significantly elevated under co‐stimulatory conditions (Figure 3I). Phagocytosis assay showed no significant impact of IL‐4/IL‐13 STAT6 signaling on the phagocytic capacity of BMDMs (supplementary Figure S4C,D). In contrast we observed a slightly increased proliferation of BMDMs in the presence of IL‐4/IL13, which was absent in Stat6‐/‐ BMDMs (supplementary Figure S4E,F) [23]. The matrix metalloproteinase‐9 (MMP9), which is expressed by pro‐inflammatory macrophages and neutrophils, was also suppressed at the protein level by IL‐4/IL‐13 STAT6‐signaling. Corroborating our in vitro observations, we stained MMP9 in AP tissue of wild type and Stat6‐/‐ mice and found a significant increase of MMP9^+^ cells 24 h after the onset of the disease in STAT6‐deficiency, which indicates a prolonged pro‐inflammatory response (Figure 3J,K). To investigate this aspect in more detail, we labeled the pro‐inflammatory transcription factor NFκBp65 in pancreatic tissue and observed at 8 h pancreatitis a nuclear localization of NFκBp65 in infiltrating cells, which declined after 24 h. Notably, STAT6 deficient mice showed a persistent nuclear localization of NFkBp65 even after 24 h (Supplementary Figure S4G). Quantitative RT‐PCR analyses of pancreatic tissue confirmed in STAT6 deficient animals a significantly more prominent increase in the mRNA expression of pro‐inflammatory genes Il1b and Tnfa, as well as of the macrophage activation markers Cd80 and Cd86. The gene expression of Mrc1 which is associated with alternatively activated macrophages was not different 24 h after induction of pancreatitis which is in agreement with the previous CD206^+^ labeling (Figure 3D,E, Supplementary Figure S4H) Interestingly other marker genes such as Ym1 or Arg showed at 24 h of pancreatitis a clearly reduced expression in Stat6‐/‐ mice compared to wild type mice (Supplementary Figure S4H). These results indicate a pro‐longed pro‐inflammatory response and a delayed/defective development of alternatively activated macrophages. Our results from in vivo and in vitro experiments demonstrate an anti‐inflammatory and tissue‐protective role of the IL‐4/IL‐13 STAT6 axis during the early phase of AP.
Macrophage mediated inflammation during acute pancreatitis in Stat6‐/‐ mice. (A) Labeling of CD68 and CCR2 in pancreas tissue of wild type and Stat6‐/‐ mice during acute pancreatitis. (B,C) Quantification of CD68+ macrophages and CCR2+ infiltrating cells in pancreas of wild type and Stat6‐/‐ mice. (D and E) Labeling and quantification of CD206+ cells in pancreatic tissue of wild type and Stat6‐/‐ mice. (F) Measurement of serum IL‐6 in wild type and Stat6‐/‐ mice. (G) BMDMs were isolated from wild type and Stat6‐/‐ mice and stimulated with LPS in the presence or absence of IL‐4/IL‐13. Quantitative RT‐PCR analysis of Arg1, Ym1, Mmp9, and Il1b was performed (n = 3). (H) BMDMs were isolated from wild type mice and co‐incubated for 24 h with freshly prepared and CCK stimulated acini in the presence or absence of IL‐4/IL‐13 to mimic pancreatic conditions. Gene expression analysis of Arg1, Ym1, Mmp9, and Il1b was performed (n = 3). (I) BMDMs were isolated from wild type mice and co‐incubated for 24 h with freshly prepared and CCK stimulated acini in the presence or absence of IL‐4/IL‐13. Cytokine release of IL‐6, IL12p40, CXCL1, IL‐10 and CCL17 was measured in supernatant (n = 4). (J) Labeling MMP9 in AP tissue of wild type and Stat6‐/‐ mice and (K) quantification of the MMP9+ cells. All results shown are based on n = 10 animals per group (8 and 24 h). Data in B, C, E‐I and K are shown as mean +/‐ SEM. Differences were tested for statistical significance by unpaired students t‐test for independent samples and significance levels of p < 0.05 are marked by an asterisk. qRT‐PCR results were analyzed by ANOVA followed by Holm‐Šídák test for multiple comparisons.
Fibrogenesis in Chronic Pancreatitis Depends on IL‐4/IL‐13 STAT6‐Signaling
2.4
In previous work we and others have shown that the IL‐4/IL‐13 STAT6‐signaling pathway plays a critical role for the development of fibrosis in chronic pancreatitis [14,15]. Pancreatic Th2 cells and ILC2s release IL‐4 and IL‐13 and induce the polarization of alternatively activated macrophages, which in turn stimulate pancreatic fibroblasts via the release of TGF‐β. In contrast to immune cells, pancreatic acinar cells did not show any expression of STAT6, but several other cells in the chronic inflamed pancreas showed a distinct labeling of STAT6. This expression was absent in the Stat6‐/‐ mice (supplementary Figure S5A). We were interested to find out to which extent the fibrosis can be attenuated by the deletion of STAT6. CP was induced in in wild type and Stat6‐/‐ mice by repetitive injections of caerulein over 4 weeks (Figure 4A). Staining and quantification of fibrotic filaments by azan blue showed a significantly reduced fibrosis in the Stat6‐/‐ mice compared to the wild type animals (Figure 4B). However, there was no reduction of fibroblast growth factor receptor (FGFR) expressing cells in the pancreas of Stat6‐/‐ mice. Labeling of αSMA, a marker of myofibroblast formation, showed decreased expression in CP‐tissue of Stat6‐/‐ mice, whereas the co‐labeling of collagen 1 and α‐amylase was also not altered (Figure 4C). To evaluate the extracellular matrix protein composition we compared the pancreas tissue proteomes between Stat6‐/‐ and wild type CP and found in 6036 proteins only 136 differences (Figure 4D,E). In STAT6‐deficient pancreas the protein amounts of collagens COL3A1, COL6A5, COL1A2, COL4A1 and COL4A2, elastin, fibrillin‐1, fibulin and several laminin subunits were significantly decreased. Interestingly, the abundance of the fibroblast marker proteins vimentin and desmin was unchanged (Figure 4F), so that the reduction of extracellular matrix protein in Stat6‐/‐ apparently is not due to reduced numbers of pancreatic fibroblasts. A specific quantitative RT‐PCR analysis of CP tissue from wild type and Stat6‐/‐ mice provided further evidence that the expression of some collagen genes such as Col3a1 and Col6a5 was affected by the deletion of STAT6, whereas that of others (Col5a2 or Col1a1) was not reduced. Transcript levels of the gene encoding the filament protein Vimentin (Vim) were also not different between wild type and Stat6‐/‐ mice (Figure 4G). We next evaluated the proteome data by Ingenuity Pathway Analysis (IPA) software to identify involved regulatory pathways. In addition to the expected involvement of STAT6 in IL‐4 and IL‐13‐signaling, the analysis showed an impact of STAT6 deletion on the transcription factor MAFB [24], which plays a pivotal role in macrophage differentiation and on transcription factors FOXA2, TP63 and MYC which are involved in pancreatic carcinogenesis. The Voronoi treemap plot illustrates the 50 upstream regulators with the highest p‐value (Figure 4H). The calculated z‐score indicates whether a signaling pathway is activated (positive) or inhibited (negative), demonstrating activation of the macrophage differentiation related transcription factors MAFB and the anti‐inflammatory transcription factor IRF2BP2, whereas PTEN, which is known to counteract acinar cell dedifferentiation, is inhibited (Figure 4I). The pathway analysis by IPA identified several disease related functions and canonical pathways which were significantly affected by the deletion of STAT6 including “Inflammatory response”, “Leukocyte migration”, “Epithelial neoplasm”, “Elastic fiber formation”, “Collagen degradation” and “Extracellular matrix organization” (Supplemental Figure S5B,C, supplementary Tables S1–S4).
Fibrogenesis in chronic pancreatitis depends on IL‐4/STAT6‐signaling in mice. (A) Chronic pancreatitis (CP) was induced by repetitive caerulein injections over 4 weeks in wild type and Stat6‐/‐ mice. (B) Azan blue staining area quantification illustrates pancreas fibrosis in CP mice. (C) Immunofluorescent labeling of FGFR, αSMA, α‐amylase and collagen 1 in CP tissue, and quantification of FGFR+ cells in the pancreas of CP mice. Differences were tested for statistical significance by unpaired students t‐test for independent samples and significance levels of p < 0.05 are marked by an asterisk (n = 10). (D,E) Comparative analysis of pancreas proteome data of wild type and Stat6‐/‐ mice after induction of CP (n = 5 each). A volcano plot (D) and a pie chart (E) illustrate the elevated and decreased protein levels detected in wild type mice compared to Stat6‐/‐ (significance was tested using ROPECA). (F) Heat map illustrating different levels of extracellular matrix proteins between wild type and Stat6‐/‐ mice. While the number in the box represents the change compared to the untreated wild type control mice the amounts of all red labeled proteins were significantly lower in Stat6‐/‐ animals compared with wild type mice after onset of CP. (G) Quantitative RT‐PCR to compare the transcript levels of Col3a1, Col1a1, Col6a5, Col5a2 and Vim in CP tissue of wild type and Stat6‐/‐ mice (n = 5). Data in B, C and G are shown as mean +/‐ SEM. Differences were tested for statistical significance by unpaired students t‐test or Mann–Whitney for non‐normally distributed samples for independent samples, significance levels of p < 0.05 are marked by an asterisk. (H) Voronoi tree map shows the 50 upstream regulators with the smallest significance value (the area in the map reflects the ‐log10 of p‐value whereas the color specifies the group of upstream regulators. (I) A Volcano plot illustrating the calculated z‐score of the upstream regulators (<2 is marked as activated, whereas >−2 is marked as inhibited in Stat6‐/‐ CP animals).
CD206 M2‐Like Macrophages Differentiate in the Absence of STAT6 but Retain Proinflammatory Properties
2.5
As IL‐4/IL‐13‐stimulated macrophages contribute to pancreatic fibrosis by the release of TGF‐β, we next investigated macrophage polarization in the pancreas of CP mice. Labeling of CD206 in pancreatic tissue of wild type and Stat6‐/‐ mice showed comparable numbers of alternatively activated CD206^+^ M2 like macrophages, whereas CCR2^+^ infiltrating cells as well as CD3^+^ T‐cells were significantly increased in Stat6‐/‐ mice (Figure 5A,B). A proteome analysis of CP tissue from wild type and Stat6‐/‐ mice provided further evidence that macrophage numbers were comparable in both mouse strains (Figure 5C). The expression of general macrophage marker proteins F4/80 and CSF1R was similar, whereas markers of alternatively activated macrophages showed an ambiguous pattern. While MRC1 and CD163 were slightly reduced in Stat6‐/‐ mice, others such as Chil3 (Ym1) were significantly elevated in CP tissue of Stat6‐/‐ mice. To further validate these unexpected findings, we performed quantitative RT‐PCR analysis and found, similar to the proteome data, the alternative activated macrophage related genes Arg1, Fizz, Ym1 and Il10 all upregulated in the absence of STAT6 (Figure 5D). Apparently, in Stat6‐/‐ mice an IL‐4/IL‐13‐independent pathway drives polarization into CD206^+^ macrophages, which have a slightly different marker setup compared to STAT6‐competent cells. Notably, CP pancreas tissue of Stat6‐/‐ animals had a more pro‐inflammatory immune cell composition that contained a significantly higher number of MMP9^+^ cells (Figure 5E). A quantitative RT‐PCR analysis of CP tissue confirmed this observation by the detection of a higher expression of Nos2, Il1b and Mmp9 in STAT6 deficient mice (Figure 5F). Surprisingly, we measured elevated levels of TGF‐β in the serum of both, Stat6‐/‐ and wild type animals (Figure 5G) and detected elevated mRNA levels of Tgfbi in the pancreas of Stat6‐/‐ mice. A higher level of Csf1 mRNA, which encodes macrophage colony stimulating factor 1 (M‐CSF) was also detected (Figure 5H). These observations indicate that in the development of CP M2 alternative macrophage polarization at least in part is independent of IL‐4/IL‐13 STAT6‐signaling. While in STAT6‐deficient mice the differentiated CD206^+^ M2‐like macrophages retained pro‐inflammatory properties corresponding to a mixed M1/M2 functional phenotype, the cells released TGF‐β as expected. This may explain why the fibrosis was only moderately reduced in the absence of STAT6.
CD206 M2‐like macrophages differentiate in the absence STAT6 but retain proinflammatory properties (A) Immunofluorescent labeling of CD206, CCR2 and CD3 in CP tissue. (B) Quantification of CD206+, CCR2+ and CD3+ cells in the pancreas of CP mice (n = 10). (C) Heat map illustrating different levels of inflammation related proteins between wild type and Stat6‐/‐ mice. The numbers in the boxes represents the change compared with untreated control mice, the amount of all red labeled proteins were significantly changed in Stat6‐/‐ compared to wild type mice after onset of CP (n = 5). (D) Quantitative gene expression analysis by RT‐qPCR of Arg1, Fizz, Ym1 and Il10 in CP tissue of wild type and Stat6‐/‐ mice (n = 5). (E) Immunofluorescent labeling of MMP9 (red) and α‐amylase (green) in CP tissue and quantification of MMP9+ cells in the pancreas of CP mice. (F) Quantitative RT‐PCR of Nos2, Il1b, Mmp9 and Il6 in CP tissue of wild type and Stat6‐/‐ mice (n = 10). (G) Measurement of TGF‐β in serum of untreated control mice and CP mice (n = 9). (H) Quantitative RT‐PCR of Tgfbi and Csf1 in CP tissue of wild type and Stat6‐/‐ mice (n = 5). Data in B and D‐H are shown as mean +/‐ SEM. Differences were tested for statistical significance by unpaired students t‐test or Mann–Whitney for non‐normally distributed samples for independent samples, significance levels of p < 0.05 are marked by an asterisk.
Pancreatic Fibroblasts Respond to IL‐4/IL‐13
2.6
As the detected differences in the extracellular matrix proteins of Stat6‐/‐ mice did not correlate with altered numbers of CD206^+^ alternatively activated macrophages in CP tissue, we next investigated a potential direct role of pancreatic fibroblasts (PAFs). Immunofluorescent co‐labeling with the fibroblast marker αSMA on wild type CP pancreatic tissue sections showed expression of the IL‐4 receptor on αSMA^+^ fibroblasts (Figure 6A). That PAFs express the IL‐4R, could also be confirmed in co‐labeling experiments on isolated fibroblasts from pancreatic tissue (Figure 6B). We then stimulated isolated pancreatic fibroblasts from wild type and Stat6‐/‐ mice with 10 ng/mL IL‐4/IL‐13(Figure 6C) and observed the induced expression of several extracellular matrix proteins. The STAT6 pathway was necessary for the expression of Col3a1, Col6a5 and Col4a2, but not of Col1a1, Acta2 and Vim (Figure 6D). IL‐4/IL13 stimulation caused a significantly elevated release of IL‐10 from wild type PAFs which was absent in PAFs from Stat6‐/‐, and a similar induction pattern could be observed for the growth factor G‐CSF. Surprisingly the basal level of IL‐6 release was higher in fibroblasts lacking STAT6, which indicates an elevated pro‐inflammatory response (Figure 6E). Furthermore, the IL‐4/IL‐13‐induced mRNA expression of Fn1, Socs1 and Csf1 was detected only in fibroblasts of wild type mice and only in wild type fibroblasts we observed an IL‐4/IL‐13 stimulated cell proliferation in an CCK8‐assay (Figure 6F). In addition to IL‐4/IL‐13, TGF‐β and IL‐6 have also been identified in chronically inflamed tissue. Co‐stimulation experiments demonstrate a partially antagonistic effect of IL4/IL13 on the TGF‐β‐mediated expression of Acta2 or Col1a1. In contrast, TGF‐β is able to suppress the IL‐4/IL‐13‐mediated expression of Col3a1, Col6a5, Socs1 and Fap (fibroblast activating protein). Conversely, we observed a modestly enhancing effect of co‐stimulation with IL‐6 on the expression of Col6a5 (Supplementary Figure S6). These experiments indicate that the differences in the extracellular matrix composition that we observed in Stat6‐/‐ CP can be explained by a direct stimulatory effect of IL‐4/IL‐13 on fibroblasts. This suggests the existence of a CD206^+^ macrophage‐independent mechanism of fibrogenesis. Furthermore, in human chronic pancreatic tissue, we labeled IL4Ra on the surface α‐SMA^+^ fibroblast, suggesting that IL‐4/IL‐13 can play a critical role in the fibrogenesis of CP patients (supplementary Figure S7A). The serum level of IL‐4 is not elevated in patients with pancreatitis, but a limited local trigger mechanism could be relevant (supplementary Figure S7B).
Pancreatic fibroblasts respond to IL‐4/IL‐13. (A) Immunofluorescent co‐labeling of αSMA and the IL‐4R in CP tissue of wild type mice and (B) in isolated pancreatic fibroblasts. (C) Pancreatic fibroblasts were isolated from wild type and Stat6‐/‐ mice and stimulated with 10 ng/mL each of IL‐4 and IL‐13 for 24 h. (D) Quantitative RT‐PCR analysis using RNA prepared from pancreatic fibroblasts of wild type and Stat6‐/‐ after IL‐4/IL‐13 treatment (n = 4). (E) Cytokine secretion of fibroblast isolated from wild type and Stat6‐/‐ mice were measured in the supernatant after stimulation with 10 ng/mL each of IL‐4 and IL‐13 for 24 h (n = 3). (F) The CCK8 assay illustrates increased cell growth after IL‐4/IL‐13 treatment over 2 days in wild type fibroblasts but not in fibroblasts of Stat6‐/‐ mice (n = 3). Data in D‐F are shown as mean +/‐ SEM. Statistically significant differences were analyzed students t‐tests or by ANOVA followed by Holm‐Šídák test for multiple comparisons, significance levels of p < 0.05 are marked by an asterisk.
IL‐4/IL‐13 STAT6‐Signaling Promotes the Pancreatic Acinar Cell Regeneration
2.7
CP development is generally accompanied by a gradual loss of exocrine secretory tissue and its replacement by fibrotic tissue. In the following experiments we investigated the role of IL‐4/IL‐13 STAT6‐signaling on acinar cell regeneration. Proteome analysis of CP tissue revealed a stronger reduction of secretory enzymes chymotrypsin, pancreatic amylase, pancreatic lipase, elastase and trypsin in Stat6‐/‐ CP mice as compared with wild type animals (Figure 7A). Also, the amount of pancreatic triacylglycerol lipase and of chymotrypsin B were significantly reduced in Stat6‐/‐ CP mice. Interestingly, insulin‐1 and insulin‐2 levels were significantly higher in Stat6‐/‐ mice, pointing to a preserved function of the islands of Langerhans. Labeling of insulin and α‐amylase in CP tissue of wild type and Stat6‐/‐ mice supports the observation that the insulin producing pancreatic island cells remained intact while exocrine tissue was lost (Figure 7B). Weight curves show that Stat6‐/‐ mice lost significantly more body weight than wild type animals during CP development (Figure 7C). The fecal elastase activity, a marker of exocrine function, is significantly reduced compared to wild type mice, and chymotrypsin activity is also reduced but failed significance (Figure 7D). These results indicate a more pronounced loss of exocrine tissue and/or impaired acinar cell regeneration in the absence of STAT6. A dedifferentiation of acinar cells known as acinar to ductal metaplasia (ADMs) occurs in response to cellular damage and inflammation. Immunofluorescent labeling of α‐amylase and CK19 showed significantly increased numbers of CK19^+^ ductal‐like cells in CP tissue of Stat6‐/‐ mice (Figure 7E,F). To validate our observation from the immunofluorescent labeling at the transcriptional level we performed quantitative RT‐PCR from CP tissue (Figure 7G) and measured significantly higher transcript levels of ADM markers Ck19, Sox4, Sox9, Prom1 and Spp1 in CP tissue of Stat6‐/‐ mice, whereas Amyl mRNA levels were not different between the mouse strains. To evaluate acinar cell proliferation in the CP tissue we labeled pancreatic sections with α‐amylase and Ki67 and observed a significant decrease in the percentage of proliferating (Ki67^+^) acinar cells (α‐amylase^+^) in CP tissue of Stat6‐/‐ mice (Figure 7H,I). Thus, Stat6‐/‐ mice experience during CP development a reduction in both fibrosis and pancreatic acinar cell regeneration, which suggests that IL‐4/IL‐13 STAT6‐signaling has a general effect on tissue regeneration.
IL‐4/IL‐13 STAT6‐signaling promotes acinar cell regeneration during chronic pancreatitis. (A) Heat map illustrating different levels of secretory proteins of acinar cells as well as insulin1 and insulin2 between wild type and Stat6‐/‐ mice. The numbers in the boxes represents the changes compared to untreated wild type control mice, the amounts of all red labeled proteins were significantly changed in Stat6‐/‐ compared with wild type mice after onset of CP. (B) Labeling of insulin (white) and α‐amylase (red) in CP tissue of wild type and Stat6‐/‐ mice. (C) Differences in body weight development of mice during induction of CP (n = 10 each CP group, n = 5 each untreated control group). (D) Fecal elastase and chymotrypsin activity of CP mice (n = 8). (E) Labeling of α‐amylase (green) and CK19 (red) in pancreas of CP mice. (F) Quantification of CK19+ cells in pancreatic tissue of CP mice. (G) Quantitative RT‐PCR analysis of Ck19, Amyl, Sox4, Sox9, Prom1 and Spp1 in CP tissue of wild type and Stat6‐/‐ mice. (H) Labeling of Ki67 (white) and α‐amylase (red) in CP tissue of wild type and Stat6‐/‐ mice. (I) Quantification of Ki67+/ α‐amylase+ cells in pancreatic tissue of CP mice. Data in C, D, F, G and I are shown as mean +/‐ SEM. Differences were tested for statistical significance by unpaired students t‐test or Mann–Whitney for non‐normally distributed samples for independent samples, significance levels of p < 0.05 are marked by an asterisk.
Enhanced Pro‐Inflammatory Immune Cells Delay Acinar Cell Regeneration in Stat6‐/‐ Mice
2.8
In order to test whether the impairment of tissue regeneration in Stat6‐/‐ mice is a general phenomenon not restricted to CP, we induced acute pancreatitis and investigated pancreatic regeneration 72 h after the onset of the disease (Figure 8A). In H&E histology a de‐differentiation of exocrine pancreas tissue was more prominent in Stat6‐/‐ than in wild type mice (Figure 8B). These de‐differentiated parts of the pancreas were visible in azan blue staining and showed high levels of CK19, but only residual amounts of α‐amylase (Figure 8C). Quantification of CK19^+^/α‐amylase^+^ cells showed a significantly elevated number of de‐differentiated acinar cells in the Stat6‐/‐ mice 72 h after the disease onset (Figure 8D). ADM formation occurred in the later phase of acute pancreatitis and could be detected by CK19/α‐amylase labeling 72 h after onset of AP. At 8 and 24 h no clear ADM formation could be detected in the pancreas of wild type or Stat6‐/‐ mice (supplementary figure 8). By quantitative RT‐PCR analysis of pancreatic tissue we confirmed that Stat6‐/‐ mice exhibited a significantly higher expression of CK19, Sox4, Sox9, Prom1 and Spp1 (Figure 8E). The number of CD206^+^ macrophages as well as the number of infiltrating CCR2^+^ immune cells was not different between wild type and Stat6‐/‐ mice (Figure 8F,G). Healthy and de‐differentiated parts of the pancreas can be distinguished using Azan blue staining, and the immunofluorescence labeling of α‐amylase and CK19 verified this classification (Figure 8I). A known promotor of tissue remodeling is the matrix metalloproteinase MMP‐9, also known as type IV collagenase, which belongs to the family of zinc‐dependent metallo‐endopeptidases that degrade or cleave extracellular matrix (ECM). In pancreatic tissue pro‐MMP‐9 is produced especially by macrophages, neutrophils, fibroblast and epithelial cells. MMP9 mRNA and protein levels were increased in AP. Labeling of MMP9 showed that the enzyme was abundantly produced by infiltrating cells in the de‐differentiated areas of the pancreas but barely in healthy parts of the organ. Quantitative RT‐PCR analysis of Mmp9 demonstrated a significantly increased expression in pancreatic tissue of Stat6‐/‐ mice 72 h after the onset of pancreatitis (Figure 8J). MMP9^+^ cell numbers were significantly increased in Stat6‐/‐ mice compared to wild type animals (Figure 8K,L). NF‐κB p65 labeling revealed a prominent nuclear redistribution in infiltrating cells in the Stat6‐/‐ mice as an indicator of an ongoing pro‐inflammatory immune response (Figure 8M). Quantitative RT‐PCR analysis also a significantly increased mRNA expression level of pro‐inflammatory markers Il1b and Il6 (Figure 8N). Recent data provided evidence that IL‐6 mediated STAT3 activation results in the formation of ADM in pancreatic acinar cells [25]. Furthermore, an enhanced expression of STAT6 was detected in the de‐differentiated areas of pancreatic tissue (Figure 8O), suggesting a direct impact of pro‐inflammation on ADM formation and delayed regeneration. These results indicate that in the absence of IL4/IL13 STAT6‐signaling an enhanced pro‐inflammatory immune profile impairs the acinar cell regeneration.
In Stat6‐/‐ mice enhanced pro‐inflammatory immune cells delay the acinar cell regeneration. (A) AP was induced by 8 hourly injections of caerulein in wild type and Stat6‐/‐ mice and all animals were sacrificed 72 h after the first injection. (B) H&E histology of the pancreas. (C) Azan blue staining of the pancreas and immunofluorescent co‐labeling of α‐amylase (green) and CK19 (red) illustrates acinar to ductal metaplasia (ADM) in the pancreas. (D) Quantification of CK19+ cells in the pancreas (n = 7). (E) Quantitative RT‐PCR of Ck19, Sox4, Sox9, Prom1 and Spp1 in pancreatic tissue of wild type and Stat6‐/‐ mice. (F) Immunofluorescent labeling of CD206 and CCR2 in pancreatic tissue sections. (G) Quantification of Immunofluorescent labelling of CD206+ macrophages and CCR2+ infiltrating leukocytes. (I) Azan blue staining together with α‐amylase/CK19 labeling and MMP9 labeling shows a redistribution of MMP9+ cells in the pancreas of wild type mice 72 h after onset of pancreatitis in consecutive sections (the white‐circled annotations mark the de‐differentiated areas in the pancreas). (J) Quantitative RT‐PCR of Mmp9 in pancreatic tissue of wild type and Stat6‐/‐ mice. (L) Quantification of MMP9+ cells in the pancreas, subdivided according to healthy and ADM rich areas (marked by annotation in panel I). (M) In acute pancreatitis labeling of NFκBp65 in pancreatic tissue shows a nuclear redistribution in Stat6‐/‐ mice. (N) Quantitative RT‐PCR of Il1b and Il6 in pancreatic tissue of wild type and Stat6‐/‐ mice. (O) Labeling of STAT3 in wild type pancreatic tissue 72 h after onset of pancreatitis. Data in D, E, G, J, L and N are shown as mean +/‐ SEM. Differences were tested for statistical significance by unpaired students t‐test or Mann–Whitney for non‐normally distributed samples for independent samples, significance levels of p < 0.05 are marked by an asterisk.
Discussion
3
The induction of an acute episode of pancreatitis triggers a prominent local pro‐inflammatory immune response in which pro‐inflammatory mediators like TNF‐α [21], IL‐1β [26] or ROS [27] cause acinar cell damage. Pancreatic macrophages are fundamentally involved in the amplification of the pro‐inflammatory response [8,12]. In the disease progression a switch from pro‐inflammatory macrophages (M1) toward alternative activated macrophages (M2) occurs, which could serve as a promising therapeutic strategy to prevent hyperinflammation and to mitigate the disease severity [12,28,29]. Cytokines IL‐4 and IL‐13 are crucial mediators for the induction of the alternatively activated macrophage polarization [30]. The IL‐4/IL‐13 pathway depends on the presence of the transcription factor STAT6, which becomes phosphorylated during IL‐4 receptor activation [18], leading to dimerization and translocation into the nucleus. There STAT6 regulates the transcription of various genes involved in the type 2 immune response and attenuates the pro‐inflammatory gene expression [31]. In AP this pathway gets activated early, and mice that lack STAT6 expression develop significantly more pancreatic damage, as indicated by higher numbers of activated CD68^+^ pancreatic macrophages. Early after disease onset (8 h) these mice have elevated serum levels of pro‐inflammatory cytokines IL‐6 and IL‐1β, which are released from macrophages. We know, that in the initial immune response these macrophages have a decisive role, as they translate the local inflammation into a systemic immune reaction [9,13,28]. In our study we demonstrate that STAT6 has an important role in the suppression of the pancreatitis‐associated pro‐inflammation, which is compromised in Stat6‐/‐ mice. In line with our results, the group of Czimmerer et al. could show that IL‐4‐mediated STAT6 activation limits the response of macrophages and prevents inflammasome activation and the release of mature IL‐1β [32]. In an episode of AP the inflammasome activation contributes directly to local and systemic complications [9,26]. Pancreatic macrophages also release pro‐inflammatory mediators like TNF‐α, which can directly act on acinar cells and mediate cell necrosis/necroptosis [21,22,33,34]. The suppression of this TNF‐α release by IL‐4/IL‐13 STAT6 activation [35] represents a critical immunosuppressive role of the IL‐4/IL‐13 signaling, in the early phase of AP similar to the anti‐inflammatory activity described for IL‐10 during AP [36,37]. In absence of the immunosuppressive action of IL‐4/IL‐13 we observed an elevated number of MMP9^+^ cells in Stat6‐/‐ mice. MMP9 is expressed by pro‐inflammatory macrophages (M1) [38] and neutrophils [39], thus, elevated MMP9 levels mark a prolonged pancreatic inflammatory state in the absence of STAT6. Apparently STAT6‐signaling suppresses the pro‐inflammatory immune response during the disease manifestation of AP and ameliorates the disease severity. In a recent patient study an elevated serum Th2 cytokine profile, including IL‐4 and IL‐13, was associated with mild AP, whereas a Th1 cytokine profile was associated with a severe course of the disease [40].
The activation of STAT6 depends mainly on the cytokines IL‐4 and IL‐13, which are the signature cytokines of the type 2 immune response. They are released by Th2 cells or ILC2s, but also basophils, eosinophils and mast cells are a source of these cytokines [41]. In a previous study our group has shown, that a systemic activation of GATA3^+^ Th2 cells occurs early during AP [11], and in parallel with the pro‐inflammatory response that originates in the inflamed pancreatic tissue [9]. Apparently, the immune response to AP involves a balanced counter regulation of pro‐ and anti‐inflammatory mechanisms.
In most cases CP develops from recurrent episodes of AP which result in a gradual fibro‐inflammatory replacement of exocrine tissue [3]. Also, in CP Th2 cells migrate into the pancreas and regulate the local macrophage polarization. The release of IL‐4/IL‐13, drives the polarization of alternatively activated macrophages (M2) [15], which is associated with pancreatic fibrogenesis and organ regeneration [15,42]. Pancreatic remodeling and fibrogenesis are regulated by the type 2 immune response [14], where macrophages in particular play a decisive role [12,15]. Alternatively activated macrophages (M2) release TGF‐β which drives the fibrogenesis by the activation of pancreatic fibroblasts [15,43]. M2‐macrophages are characterized by an anti‐inflammatory profile and the expression of the mannose receptor MRC‐1 (CD206) [15]. In our study we show that the genetic deletion of STAT6 completely prevented M2 differentiation in response to IL‐4/IL‐13 stimulation in vitro in isolated BMDMs. In vivo the situation is more complex. Obviously, M2 polarization is not complete in the absence of STAT6. While the expression of TGF‐β or YM1 is induced, a suppression of pro‐inflammatory genes does not occur, which results in a mixed functional phenotype of M2‐like macrophages in pancreatitis. The disbalance between pro‐ and anti‐inflammation causes more tissue destruction and impairs organ regeneration. Normally, IL‐4/IL‐13 signaling is known to induce polarization of M2a macrophages, a subtype of the M2‐macrophage continuum [44,45], which is associated with wound healing and tissue repair. Surprisingly, in the absence of STAT6 the number of CD206^+^ macrophages was not decreased in CP tissue, and the expression of hallmark M2 markers such as Fizz1, Ym1 or Arg1 was even elevated. This result indicates a replacement of the IL‐4/IL‐13 induced M2a subtype by other more pro‐inflammatory M2‐macrophage phenotypes. In addition to the elevated M2 markers, a significantly elevated TGF‐β expression in the pancreas of Stat6‐/‐, may explain the observed fibrogenesis in the absence of IL‐4/IL‐13 induced M2 macrophages. A recently published study reported a similar phenomenon: mice with a double knockout for the cytokines IL‐4 and IL‐13 showed increased fibrosis in the model of alcohol and smoking‐induced CP [46]. In summary, these data show that the induction of fibrosis during CP involves a complex network of different signals. Blocking individual signals therefore does not always have a beneficial effect. The fact that CP is a relapsing‐remitting disease which mostly involves acute phases are also part of the clinical picture makes treatment even more complicated.
Our proteome analysis provided a detailed insight into the composition of extracellular matrix proteins during CP. IL‐4/IL‐13 STAT6‐signaling increased the synthesis of the matrix proteins COL3A1, COL6A5 or COL4A2, while others such as COL1A1, COL14A1, COL12A1 or COL18A1 were not affected. We were able to demonstrate that the transcriptional control of Col3a1, Col6a5 and Col4a2 depends on IL‐4/IL‐13 STAT6‐signaling in pancreatic fibroblasts. It is known that IL‐4/IL‐13 induces collagen synthesis also in lung fibroblasts [47] as well as in hepatic stellate cells [48]. Our results suggest that Th2 cells and ILC2s have a direct impact on fibrogenesis via the IL4/IL13 STAT6‐signaling in pancreatic fibroblasts, independent of macrophages. Comparable to the polarization of alternatively activated macrophages by multiple signaling pathways, also the activation of PAFs can be mediated by various signals like TGF‐β, pro‐inflammatory cytokines and, as demonstrated, by IL‐4/IL‐13. Our proteome data suggest that the extracellular matrix composition, specifically the presence of distinct collagen subtypes, may allow predictions about the activation mechanism. PAFs seem to represent a continuum of different subtypes, as already demonstrated in pancreatic cancer and cancer associated fibroblasts (CAFs) [49,50].
In CP, functional exocrine tissue is successively replaced by fibrotic tissue [3,28]. Beside Th2 cells and ILC2s [14] also activated pancreatic stellate cells release IL‐4 after activation by TGF‐β [15]. IL‐4/IL‐13 cytokines induce the polarization of alternatively activated macrophages which then activate fibroblasts [15]. The blockage of the IL‐4 signaling pathway by an IL‐4/IL‐13‐blocking peptide [15], by antibodies against IL‐4 [14], or by the genetic deletion of the IL‐4Rα [15] has been shown to reduce pancreatic fibrosis in a mouse model of CP. These studies indicate that the IL‐4 signaling pathway could represent a therapeutic target to limit pancreatic fibrosis and preserve organ‐function, but the question, if the prevention of fibrosis can also enhance acinar cell regeneration is currently unclear. Among various clinically approved drugs that inhibit IL‐4 signaling, Dupilumab, a monoclonal antibody which blocks the α‐chain of the IL‐4 receptor, is used for the treatment of atopic dermatitis [51], asthma [52] and COPD [53], but not for patients with CP. Likewise, the IL‐4/STAT6 axis in the type 2 immune response might represent a promising target for an anti‐fibrotic organ persevering treatment in CP. In general, CP progress is marked by recurrent acute episodes of AP. A therapeutic use of antibodies, such as Dupilumab, to block IL‐4R could impede the immunosuppressive effect of IL‐4 during the acute phase, thereby increasing the risk of severe AP development. Dupilumab has a half‐life of 14 days or longer in patients [54], which poses a significant risk, as the blockage of the immunosuppressive effect cannot be rapidly terminated in case of an acute pancreatitis episode. Small molecule inhibitors with a shorter half‐life could represent a therapeutic alternative e for CP patients. However the polarization of alternatively activated macrophages involves beside IL‐4/IL‐13, various other stimuli like IL‐10 [44] or TGF‐β [55]. As a result, alternatively activated macrophages embody a continuum of cells which share specific cell surface markers like CD206 but differ regarding their response to activation signaling. By proteome profiling, we demonstrated that Stat6‐/‐ mice showed elevated MAFB activation, which is a known driver of M2 polarization. The deletion of STAT6 completely blocks IL‐4 signaling but did not prevent CD206^+^ M2 polarization and TGF‐β release. Consequently, we observed a reduction, but not a complete prevention of pancreatic tissue fibrosis in Stat6‐/‐ CP.
Our study demonstrated that IL‐4/IL‐13 signaling has a protective role in AP by suppressing the pro‐inflammation reaction. Necrotic cell death of acinar cells triggers an inflammatory reaction that induces the clearance of necrotic debris by macrophages [28,56]. It has previously been shown that the stimulation of macrophages with IL4 increased their capacity to phagocytose necrotic debris [57] or induced them to proliferate [23]. We did not observe significantly increased phagocytosis or proliferation in our study, only a tendency. Our AP experiments showed ongoing acinar cell necrosis in STAT6‐deficient mice and a prolonged pro‐inflammatory response, which delayed or disrupted cell proliferation and organ regeneration [58]. It also caused acinar cell de‐differentiation and ADM formation via secretion of pro‐inflammatory cytokines like IL‐6 [25] or TNF‐α [59], which were released during AP [21,60] and as we could show were suppressed by IL‐4/IL‐13/STAT6 signaling. Prolonged pancreatic inflammation associated with diminished acinar cell regeneration, and enhanced ADM development resulted in an accelerated loss of exocrine function. The group of J Cosín‐Roger et al. was able to show that IL‐4‐induced M2a macrophages are essential for wound healing and the induction of cell proliferation by the activation of WNT‐signaling in a mouse model of IBD [61]. Both processes, fibrosis and regeneration are part of the wound healing process that is regulated by the type 2 immune response. Wound healing typically occurs in three phases: 1) inflammation and necrotic clearance, 2) regeneration and new tissue formation, and 3) remodeling [62]. The promotion of organ regeneration while reducing fibrosis would be of therapeutic interest. However, the fact that both processes are regulated by the same type 2 signaling pathway poses a therapeutic challenge. Taken together, our therapeutic interference with IL‐4/IL‐13 STAT6‐signaling restricted the alternative macrophage polarization during CP but had only a moderate impact on fibrosis. A timely or more efficient suppression of the inflammatory reaction may be needed to trigger pancreas regeneration.
Our study has certain limitations. The utilization of cell‐specific STAT6‐deficient animals was not feasible within this project. A cell‐specific knockout could be able to demonstrate the precise influence of the IL‐4/IL‐13/STAT6 signaling pathway in macrophages and fibroblasts, but our data suggest that STAT6 directly controls the production of specific collagens, such as Col6a5, in pancreatic fibroblasts. Additionally, IL‐4/IL‐13 stimulation of macrophages suppresses inflammation and induces the differentiation of alternatively activated macrophages. However, we also observed the presence of CD206^+^ macrophages in the absence of STAT6. This suggests that diverse signaling pathways can induce an M2 phenotype. To investigate this in detail, single‐cell analysis of macrophage‐specific Stat6‐/‐ animals would provide insight into incomplete polarization of M2 macrophages.
In summary, we could demonstrate that STAT6‐signaling has a protective function in AP where it facilitates the suppression of inflammation and the promotion of acinar cell regeneration. During CP, STAT6‐signaling affects macrophages and also PAFs. Surprisingly in the absence of STAT6, the differentiation of CD206^+^ macrophages was not completely suppressed but resulted in a mixed M2‐like phenotype, which indicates a contribution of multiple pathways to M2 macrophage polarization. Furthermore, we discovered that IL‐4/IL‐13 secretion presumably from Th2 cells and ICL2s stimulates the secretion of certain collagens via a direct activation of STAT6 signaling in fibroblasts. This indicates the presence of a macrophage‐independent fibrogenic activation mechanism of PAFs. Our results reveal a complex cellular regulation network in tissue fibrosis and organ regeneration. A detailed analysis of the involved mechanisms will be crucial for the development of an anti‐fibrotic treatment in pancreatitis patients that does not interfere with organ regeneration.
Material and Methods
4
Ethics Declaration
4.1
Acute pancreatitis and control blood donor serum samples were collected at the university medicine Greifswald after approval by the Ethical committee of the university medicine Greifswald (III UV 91/03). All patients gave written and informed consent. Human chronic pancreatitis tissue samples were collected in the context of the ChroPac trial (ISRCTN38973832) [63].
All experimental procedures regarding mouse experiments were approved by the local animal care committee (Landesamt für Landwirtschaft, Lebensmittelsicherheit und Fischerei Mecklenburg‐Vorpommern, protocol number: 7221.3‐1‐030/22). All animal experiments were performed in accordance with the ARRIVE guidelines.
Mice Model
4.2
Mice were housed at the central animal facility at University Medicine Greifswald (UMG) in accordance with German Animal Protection Law. Mice were kept at a maximum of 5 mice per cage at pathogen‐free conditions (SPF) in a temperature‐ and humidity‐controlled environment with free access to food and water ad libitum. Stat6‐/‐ mice (B6.129S2(C)‐Stat6^1Gru^/J, strain code # 005977) were purchased from The Jackson Laboratory and used for breeding. C57Bl/6‐J mice were obtained from Janvier Labs. AP was induced by hourly repetitive caerulein i.p. injection (50 µg/kg/bodyweight) (4030451, Bachem) over 8 h and CP over 6 h, three times per week over a time period of 4 weeks in 8‐ to 16‐week‐old Stat6‐/‐ and wild type mice, as previously described [11,14]. At the indicated times, mice were anaesthetized using a combination of ketamine and xylazine, administered intraperitoneally. Following confirmation of deep anesthesia retro‐orbital blood sample was taken and mice were euthanized by cervical dislocation. Pancreas and lung were removed immediately for further analysis.
Reagents and Antibodies
4.3
Anti‐Mrc1/CD206 (dilution 1:200, OASA05048, aviva‐sysbio), CD68 (dilution 1:200, ABIN181836, antibody‐online), CCR2 (dilution 1:200, ab273050, abcam), anti‐αSMA (dilution 1:200, M0851, Dako), anti‐α‐amylase (dilution 1:200, sc‐46657, Santa Cruz), anti‐collagen 1 (dilution 1:200, ab34710, abcam), anti‐Cytokeratin 19 (CK19) (dilution 1:200, ab195872, abcam), anti‐Insulin (dilution 1:200, #4590, Cell Signaling) anti‐Ki‐67 (dilution 1:200, IHC‐00375, Bethyl), anti‐IL‐4R (dilution 1:200, PA5‐103142, Invitrogen), anti‐MMP9 (dilution 1:200, ab228402), anti‐FGF Receptor (dilution 1:200, 9740S, cell signaling), anti‐STAT6 (dilution 1:1000, ab32520, abcam), anti‐pSTAT6 (dilution 1:1000, 700247, Invitrogen), anti‐NFkBp65 (dilution 1:100, #8242, Cell Signaling), anti‐CD3 (dilution 1:100, Cat 100202 Biolegend), anti‐MPO (dilution 1:200, AF3667, R&D systems), anti‐STAT3 (dilution 1:100, D3Z2G, cell signaling), anti‐TGF‐β (dilution 1:1000, ab215715, abcam), anti‐GAPDH (dilution 1:1000, H86504M, Meridian OH, USA), anti‐mouse‐Cy5 (dilution 1:200, 115‐175‐146, Jackson ImmunoResearch), anti‐mouse‐FITC (dilution 1:200, 115‐095‐146, Jackson ImmunoResearch), anti‐rabbit‐AF488 (dilution 1:200, A21206, Invitrogen), anti‐rabbit‐Cy3 (dilution 1:200, 111‐165‐144, Jackson ImmunoResearch), anti‐rabbit‐Cy5 (dilution 1:200, 111‐175‐144, Jackson ImmunoResearch) anti‐rat‐Cy3 (dilution 1:200, 112‐165‐062, Jackson ImmunoResearch), ECL anti‐mouse‐HRP (NA931V, Sigma‐Aldrich), ECL anti‐rabbit‐HRP (NA934VS, Sigma‐Aldrich). Recombinant Mouse IL‐4 Protein (#404‐ML, R&D Systems), Recombinant Mouse IL‐13 Protein (#413‐ML, R&D Systems), Lipopolysaccharide Escherichia coli O26:B6 (L8274, Sigma Aldrich), Recombinant Mouse IL‐6 Protein (406‐ML, R&D systems), Recombinant Mouse TGF‐β Protein (7666‐MB, R&D systems), Recombinant mouse IL‐10 (575804, BioLegend).
Histology, Immunohistochemistry, and Immunofluorescence
4.4
Sequential immunostaining was performed on 2‐µm‐thick sections of FFPE tissue or fresh frozen OCT‐embedded tissue. Deparaffinization, rehydration, and antigen retrieval steps were performed for FFPE samples and cryosections were fixed with acetone. Samples were blocked and permeabilized in 0.02% Tween20 and 10% FCS in 1X PBS (blocking buffer) for 1 h. Samples were incubated with primary antibodies (1:200) diluted in blocking buffer at 4°C overnight, followed by incubation with appropriate secondary antibodies (1:200) for 1 h at room temperature. Slides were finally counterstained with DAPI to visualize nuclei.
FFPE sections of pancreas and lung were also stained with hematoxylin and eosin (12156, Morphisto) and Azan staining (12079, Morphisto) according to standard protocol. A blinded quantitative scoring system was used for pancreatitis scoring encompassing necrosis, edema and infiltration (0 = absent, 1 = minor, 2 = present, 3 = abundant).
Staining was visualized using fluorescence microscope Olympus IX81 and slides were scanned on a Pannoramic MIDI II (Sysmex) platform. The QuantCentre software from Sysmex (Kobe, Japan) was used to quantify the indicated markers in the pancreatic tissue.
Quantitative Analysis of Serum Cytokines and Chemokines
4.5
Multiplex analysis was used to determine the concentrations of chemokines (CCL5, CXCL5, CXCL11, CCL7, CCL2, LEGEDplex Custom Mouse Panel 745) and cytokines (IL‐6, TGF‐ß, IL‐1β, LEGEDplex Custom Mouse Panel 744) in the indicated serum samples. The cytokine levels of IL‐4 in human serum samples were measured by LEGENDplex Human CD8/NK Panel (741187, BioLegend). Measurement of amylase and lipase activity in the serum was performed using appropriate colorimetric kits from Roche (AMYL2 Ref: 03183742122; LIPC Ref: 03029590322) according to the manufacturer's instructions.
Myeloperoxidase Measurement
4.6
The measurement of myeloperoxidase activity in homogenized lung samples was conducted in a 50 mM potassium phosphate buffer solution (pH 6) containing 0.53 mM O‐dianisidine and 0.15 mM H_2_O_2_ at 30°C for 10 min in a Spectrophotometer (Spectramax, Molecular Devices, San Jose, CA USA). The enzyme activity was measured against a standard (Cat# 475911, Calbiochem) and adjusted to the protein content of samples [21].
Isolation of Primary Macrophages
4.7
To generate mouse bone marrow‐derived macrophages, bone marrow cells were isolated from C57BL/6‐J and Stat6 ‐/‐ mice and were cultured in RPMI medium supplemented with 10% FBS, 1% P/S and 20 ng/mL murine M‐CSF (576406, BioLegend) for 7–10 days to stimulate macrophage differentiation.
Isolation of Pancreatic Fibroblasts
4.8
Primary fibroblasts were isolated from the pancreas of C57BL/6‐J and Stat6‐/‐ mice. The pancreas was transferred to the buffer and enzyme mix of the Multi Tissue Dissociation Kit 3 (130‐110‐204, Miltenyi Biotec). Following incubation at 37°C for 10 min under continuous shaking (80 rpm), the tissue was dissociated in DMEM medium supplemented with 10% FBS, 1% P/S within the gentleMACS Dissociator (130‐093‐235, Miltenyi Biotec) at 287 runs for 52 s. Acinar cells were removed by filtration through a 40 µm cell strainer, followed by additional centrifugation (2300 rpm, 2:30 min). Pancreatic fibroblasts were cultured in DMEM medium supplemented with 10% FBS, 1% P/S and 10 ng/mL FGF‐2 (130‐105‐786, Miltenyi Biotec) for 7–10 days.
Isolation of Pancreatic Acini
4.9
Pancreatic acinar cells were isolated from murine pancreas of wild type mice under sterile conditions by collagenase digestion (Collagenase of Clostridium histolyticum (EC.3.4.24.3) from Serva, Heidelberg, Germany) [6]. Cells were maintained and stimulated for 30 min with 0.001 mM Cholecystokinin (CCK), CCK‐Octapeptid (J66669.EXD Thermo‐Fisher Scientific, Bremen, Germany) in Dulbecco's modified Eagle's medium containing 10 mM HEPES and 2% BSA and PenStrep, and subsequently co‐incubated with BMDMs.
Western Blot Analysis
4.10
Protein was extracted from cultured macrophages using a lysis buffer containing 50 mM NaF, 10 mM Na_4_P_2_O_7_, 1 mM PMSF and 10 µg/mL aprotinin. The cell lysates were mixed with 2X Laemmli buffer containing 10% 2‐mercaptoethanol and were denatured at 95°C for 10 min. SDS‐PAGE was performed using Tris/glycine gels with a 12.5% acrylamide separation gel and 5% acrylamide stacking gel at a current of 0.01 A in 1X Tris/glycine/SDS running buffer. A PageRuler pre‐stained protein ladder (#26616, Thermo Fisher) was used as a molecular weight marker. Proteins were then transferred to a nitrocellulose membrane via semi‐dry transfer (150 mA, 30 min) using transfer buffer (50 mM Tris, 40 mM glycine, 0.38% SDS and 20% methanol). Membranes were blocked at room temperature for 1 h with in NET (150 mM NaCl, 5 mM EDTA, 50 mM Tris‐HCl, and 0.05% Triton X‐100) containing 2% gelatine. Proteins of interest were detected using antibodies and enhanced chemiluminescence (Thermo Fisher Scientific) with the Fusion‐FX system (Vilber, Collégien, France).
Proliferation Assay
4.11
The Cell Counting Kit 8 (CCK 8, GK10001, GlpBio) was used to quantify the cell proliferation of pancreatic fibroblasts over a 2‐day period. Accordingly, 1.000 cells were seeded in 100 µL of culture medium in a 96‐well plate under appropriate stimulation conditions. Following the addition of 10 µL of the CCK8 solution, the fibroblasts were incubated for 2 h at 37°C. The absorption was subsequently measured at a wavelength of 450 nm.
Phagocytosis Measurement
4.12
The measurement of phagocytosis of BMDM was done using the Phagocytosis Assay Kit (IgG PE) from Cayman Chemicals (cat. 600540, Ann Arbor, MI, USA), with the use of a fluorometer (FLUOstar Optima from BLabtech, Ortenburg, Germany).
Fecal Elastase and Chymotrypsin Measurement
4.13
Fecal elastase and chymotrypsin activities were measured in stool samples from CP animals. Fecal samples were weighed and resuspend in 500 mmol/L NaCl, 100 mmol/L CaCl_2_ containing 0.1% Triton X‐100. The fecal samples were resuspended and sonicated two times. Elastase activity was determined by fluorometric enzyme kinetic over 1 h at 37°C by the usage of 0.12 mM elastase substrate (CBZ‐Ala‐Ala‐Ala‐Ala)2‐R110 (R6506, ThermoFisher Scientific). Chymotrypsin activity was measured as enzyme kinetic over 1 h at 37°C by the usage of 0.12 mM substrate Suc‐Ala‐Ala‐Pro‐Phe‐AMC (CAS Number: 71973‐79‐0, Echelon Biosciences). Kinetics were measured in 100 mmol/l Tris buffer containing 5 mmol/L CaCl_2_ at pH 8.0.
RNA Isolation, cDNA Preparation and Quantitative Real‐Time PCR
4.14
Total RNA was extracted from macrophages, fibroblasts as well as pancreatic tissue using Trizol reagent (15596026, life technologies) according to the manufacturer's instructions. The RNA samples with an A_260_/A_280_ ratio between 1.5 and 2.0 were used for the following cDNA synthesis reaction: 2 µg RNA; 5 µM OligodT primers; 75 ng random primers; 0.5 µM dNTP Mix; 1× First Strand Buffer (18080044, Invitrogen); 10 µM DTT; 40 Units RNasin Ribonuclease Inhibitor (N251B, Promega) and 200 Units M‐MLV RT (28025013, Invitrogen). The obtained cDNA was used as a template for quantitative PCR (qPCR) with SYBR Green PCR Master Mix (4309155, applied biosystems).
Relative expression levels were calculated using the 2^−∆∆Ct^‐method, normalized against Rn5s and the relative expression in control mice. The following oligonucelotides used in this study were purchased from Invitrogen: Rn5s forward 5′‐GCCCGATCTCGTCTGATCTC‐3′ reverse 5′‐GCCTACAGCACCCGGTATTC‐3′, Acta2 forward 5′‐GCCAGTCGCTGTCAGGAACCC‐3′ reverse 5′‐CCAGCGAAGCCGGCCTTACA‐3′, Amy2a4 forward 5′‐CAAAATGGTTCTCCCAAGGA‐3′ reverse 5′‐ACATCTTCTCGCCATTCCAC‐3′, Arg1 forward 5′‐ GTGGCTTTAACCTTGGCTTG‐3′ reverse 5′‐CTGTCTGCTTTGCTGTGATG‐3′, Ck19 forward 5′‐ ACCCTCCCGAGATTACAACC‐3′ reverse 5′‐CAAGGCGTGTTCTGTCTCAA‐3′, Csf1 forward 5′‐GCCTCCTGTTCTACAAGTGGAAG‐3′ reverse 5′‐ACTGGCAGTTCCACCTGTCTGT‐3′, Col1a1 forward 5′‐CAGACTGGCAACCTCAAGAA‐3′ reverse 5′‐CAAGGGTGCTGTAGGTGAAG‐3′, Col3a1 forward 5′‐ ATGGCTCACCAGGACAAAG‐3′ reverse 5′‐CACCAGGACTGCCGTTATT‐3′, Col4a2 forward 5′‐ CACGCTTCAGCACCATGCC‐3′ reverse 5′‐ACTTGTCGTTGCGGCTGGC‐3′, Col5a2 forward 5′‐ GCGACAGGAGATAAAGGTCCC‐3′ reverse 5′‐TTCTCCAACAGCACCATCCCG‐3′, Col6a5 forward 5′‐ GCGCCAACCAGTCTGAATTC‐3′ reverse 5′‐TCCTCATCTGCTCAATGGCG‐3′, Fap forward 5′‐ CAAAGGCTGGGGCTAAGAATCC‐3′ reverse 5′‐GCCACTGCAAGCATACTCGTT‐3′, Fgf2 forward 5′‐TGCGCATCCATCCCGACG‐3′ reverse 5′‐ACACTTAGAAGCCAGCAGCCGTCC‐3′, Fizz forward 5′‐ GCTGATGGTCCCAGTGAATA‐3′ reverse 5′‐CGTTACAGTGGAGGGATAGTTAG‐3′, Fn1 forward 5′‐ GCCTGAGGTGGACCCCGCTA‐3′ reverse 5′‐GGGCCCAAGTGACCCGCATC‐3′, Il1b forward 5′‐GAGGACATGAGCACCTTCTTT‐3′ reverse 5′‐GCCTGTAGTGCAGTTGTCTAA‐3′, Il4 forward 5′‐AGATCATCGGCATTTTGAACG‐3′ reverse 5′‐TTTGGCACATCCATCTCCG‐3′, Il6 forward 5′‐ CCAGAGTCCTTCAGAGAGATACA‐3′ reverse 5′‐CCTTCTGTGACTCCAGCTTATC‐3′, Il10 forward 5′‐TTGAATTCCCTGGGTGAGAAG‐3′ reverse 5′‐TCCACTGCCTTGCTCTTATTT‐3′, Mmp9 forward 5′‐ CTGGAACTCACACGACATCTT‐3′ reverse 5′‐TCCACCTTGTTCACCTCATTT‐3′, Nos2 forward 5′‐ AGAGTGAAAAGTCCAGCCG‐3′ reverse 5′‐ACAACTCGCTCCAAGATTCC‐3′, Prom1 forward 5′‐ GCCCAAGCTGGAAGAATATG‐3′ reverse 5′‐GCATGCTTGTCATAGCCAAA ‐3′, Socs1 forward 5′‐ CTCCGTGACTACCTGAGTTCCT‐3′ reverse 5′‐ATCTCACCCTCCACAACCACT‐3′, Sox4 forward 5′‐ GTGAGCGAGATGATCTCGGG‐3′ reverse 5′‐CAGGTTGGAGATGCTGGACTC‐3′, Sox9 forward 5′‐ CACACGTCAAGCGACCCATGAA‐3′ reverse 5′‐TCTTCTCGCTCTCGTTCAGCAG‐3′, Spp1 forward 5′‐CAAGCAATTCCAATGAAAGCC‐3′ reverse 5′‐ATCCGAGTCCACAGAATCC‐3′, Tgfbi forward 5′‐TCTTCAAACAGGCGTCAGCG‐3′ reverse 5′‐AAACTGAGAGAAACTGGCG‐3′, Vim forward 5′‐AACACCCGCACCAACGAGAAG‐3′ reverse 5′‐TCCTCTCTCTGGAGCATCTCCTC‐3′, Ym1 forward 5′‐TCCAGAAGCAATCCTGAAGAC‐3′ reverse 5′‐ GTCCTTAGCCCAACTGGTATAG‐3′.
Microarray‐based Transcriptome Analysis of BMDM From WT and STAT6‐/‐ mice After IL4/IL13 Stimulation
4.15
Three replicates of bone marrow derived macrophages from wild type mice and STAT6‐/‐ mice were either stimulated with IL4/IL13 [10 ng/mL] or left unstimulated (control condition) for 24 h, respectively.
All incubated cells were lysed in 1 mL TRIzol and snap frozen before RNA extraction. Total RNA was isolated following the manufacturer's instructions for total RNA isolation from TRIzol (Thermo Fisher Scientific, Waltham, MA, USA). The aqueous RNA containing solution was further processed according to the protocol RNA Clean‐up and Concentration from Phenol/Guanidine‐based RNA isolation (RNA_Clean‐Up and Concentration Kit, NorgenBiotek, Thorold, Canada). After purification and quality assessment of the total RNA preparations (average RNA‐integrity number, RIN = 10, concentration 263 ± 50 ng/µL), transcriptional profiling of stimulated and unstimulated macrophages derived from wild type and STAT6‐/‐ mice were performed using the WT Plus Kit and Clariom S mouse arrays (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions.
For probe set extraction and normalization, expression raw data were transferred to the Transcriptome Analysis Console (TAC, 4.0.2.15, Thermo Fisher Scientific, Waltham, Massachusetts, USA). The generated gene expression data sets have been submitted to Gene Expression Omnibus (GEO). Statistical analysis of the case control study based on three biological replicates for each group was performed using analysis of variance (ANOVA) with e‐Bayesian correction. Statistically significant differential gene expression in terms of transcript levels was defined as a fold change of > |1.5| between the compared conditions and an FDR < 0.05. Microarray data have been deposited in the National Center for Biotechnology Information GEO database and are accessible through the following GEO accession number: GSE297550.
Proteome Analysis of Pancreatic Tissue
4.16
The murine pancreas tissue samples were processed in accordance with the protocols that had been published previously [64,65]. In brief, the tissue samples were subjected to a process of homogenization in liquid nitrogen, utilizing a bead mill homogenizer equipped with an 8 mm tungsten carbide ball. The resulting homogenates were solubilized in 40 mM HEPES buffer containing 2% SDS, followed by a nuclease treatment to eliminate nucleic acids. The total protein concentration was determined by using the BCA method. Following reduction and alkylation, a quantity of 4 µg of each sample was digested using a Trypsin/Lys‐C Mix at an enzyme‐to‐protein ratio of 1:25 using the SP3 protocol.
As previously outlined by Reder et al. [65]., LC‐ESI MS/MS measurements were conducted in accordance with the established protocol. In summary, the peptide mixtures were separated via nano‐LC (UltiMateTM 3000) and analyzed on an Exploris 480 mass spectrometer (Thermo Electron, Bremen, Germany) in DIA mode (see also supplemental Tables S5 and S6).
The data analysis was conducted using Spectronaut (Biognosys AG, Zurich, Switzerland) version 19, employing the reviewed murine database (version 9/2024, 17212 entries) provided by UniProt [66]. Subsequent analysis was conducted utilizing R version 4.4.2 (R Core Team, 2025) with the SpectroPipeR package version 0.40 [67]. Proteome data have been deposited in the MassIVE database (https://massive.ucsd.edu) and are available as dataset MSV000098430. (Reviewer account: “MSV000098430_reviewer”; password: “pancreatitis”). Statistical significance of differently protein abundance was tested by ROPECA [68].
Data Analysis and Statistics
4.17
Data analysis, statistics, graphing and figure design were performed using GraphPad Prism (version 5.0.0 or later), R and PowerPoint (Office 2019 or later). Data are presented as means ± SEM. The Shapiro–Wilk normality test was performed to assess the normality of the data. Statistical differences between groups were analyzed using a Student's t‐test or one‐way ANOVA for normally distributed samples and Mann–Whitney or Kruskal–Wallis tests for non‐normally distributed samples. Two‐tail p values of <0.05 were considered statistically significant.
Author Contributions
Concept of the study M.S., F.U.W., and J.G. Data acquisition and interpretation: H.E., A.S., J.P., S.B., R.L., S.A., L.S., G.H., E.H., U.V., B.B., F.U.W., M.S., and J.G. Writing committee M.S., F.U.W., and J.G. Correction of manuscript and approval of final version: all.
Funding
This work was supported by Deutsche Forschungsgemeinschaft (DFG SE 2702/2‐3, GL 1096/1‐1, GRK 2719 A1, C1).
Conflicts of Interest
The authors declare no conflict of interest.
Supporting information
Supporting File: advs73523‐sup‐0001‐SuppMat.docx.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1S. M. van Dijk , N. D. L. Hallensleben , H. C. van Santvoort , et al., Dutch Pancreatitis Study Group , “Acute pancreatitis: Recent Advances Through Randomised Trials,” Gut 66 (2017): 2024.28838972 10.1136/gutjnl-2016-313595 · doi ↗ · pubmed ↗
- 2P. K. Garg and V. P. Singh , “Organ Failure Due to Systemic Injury in Acute Pancreatitis,” Gastroenterology 156 (2019): 2008.30768987 10.1053/j.gastro.2018.12.041PMC 6486861 · doi ↗ · pubmed ↗
- 3G. Beyer , A. Habtezion , J. Werner , M. M. Lerch , and J. Mayerle , “Chronic Pancreatitis,” Lancet 396 (2020): 499–512.32798493 10.1016/S 0140-6736(20)31318-0 · doi ↗ · pubmed ↗
- 4A. B. Lowenfels , P. Maisonneuve , G. Cavallini , et al., “Prognosis of Chronic Pancreatitis: An International Multicenter Study,” American Journal of Gastroenterology 89 (1994): 1467.8079921 · pubmed ↗
- 5D. Yadav and A. B. Lowenfels , “The Epidemiology of Pancreatitis and Pancreatic Cancer,” Gastroenterology 144 (2013): 1252–1261.23622135 10.1053/j.gastro.2013.01.068PMC 3662544 · doi ↗ · pubmed ↗
- 6J. M. Modenbach , C. Möller , S. Asgarbeik , et al., “Biochemical Analyses of Cystatin‐C Dimers and Cathepsin‐B Reveals a Trypsin‐Driven Feedback Mechanism in Acute Pancreatitis,” Nature Communications 16 (2025): 1702.10.1038/s 41467-025-56875-x PMC 1183308139962054 · doi ↗ · pubmed ↗
- 7I. Gukovsky , A. S. Gukovskaya , T. A. Blinman , V. Zaninovic , and S. J. Pandol , “Early NF‐κB Activation is Associated With Hormone‐Induced Pancreatitis,” American Journal of Physiology 275 (1998): G 1402.9843778 10.1152/ajpgi.1998.275.6.G 1402 · doi ↗ · pubmed ↗
- 8M. Sendler , F.‐U. Weiss , J. Golchert , et al., “Cathepsin B‐Mediated Activation of Trypsinogen in Endocytosing Macrophages Increases Severity of Pancreatitis in Mice,” Gastroenterology 154 (2018): 704–718.29079517 10.1053/j.gastro.2017.10.018PMC 6663074 · doi ↗ · pubmed ↗