Absence of the neutrophil serine protease cathepsin G decreases neutrophil granulocyte infiltration but does not change the severity of acute pancreatitis

Acute pancreatitis is characterized by an early intracellular protease activation and invasion of leukocytes into the pancreas. Cathepsins constitute a large group of lysosomal enzymes, that have been shown to modulate trypsinogen activation and neutrophil infiltration. Cathepsin G (CTSG) is a neutrophil serine protease of the chymotrypsin C family known to degrade extracellular matrix components and to have regulatory functions in inflammatory disorders. The aim of this study was to investigate the role of CTSG in pancreatitis. Isolated acinar cells were exposed to recombinant CTSG and supramaximal cholezystokinin stimulation. In CTSG−/− mice and corresponding controls acute experimental pancreatitis was induced by serial caerulein injections. Severity was assessed by histology, serum enzyme levels and zymogen activation. Neutrophil infiltration was quantified by chloro-acetate ersterase staining and myeloperoxidase measurement. CTSG was expessed in inflammatory cells but not in pancreatic acinar cells. CTSG had no effect on intra-acinar-cell trypsinogen activation. In CTSG−/− mice a transient decrease of neutrophil infiltration into the pancreas and lungs was found during acute pancreatitis while the disease severity remained largely unchanged. CTSG is involved in pancreatic neutrophil infiltration during pancreatitis, albeit to a lesser degree than the related neutrophil (PMN) elastase. Its absence therefore leaves pancreatitis severity essentially unaffected.

is further involved in detachment of adhering leukocytes from the endothelium and their transmigration into damaged tissue 11 . It also localizes to neutrophil extracellular traps, a mechnism recently found to be of relevance in pancreatitis 12 .
In view of the different interactions of CTSG under inflammatory conditions we investigated whether this serine protease also exerts effects on acute pancreatitis using a knockout mouse model for CTSG. Although CTSG reduces invasion of neutophil granulocytes this effect was not sufficient to decrease the severity of acute pancreatitis, giving it an inferior role to neutrophil granulocyte enzymes such as PNM elastase. These results suggest only a subordinate role of CTSG in progression of acute pancreatitis.

Results ctSG expression in leukocytes and pancreas.
Because of the known CTSG expression in leukocytes we used this cell population as a positive control. In acinar cells no CTSG expression was detected by reverse transcriptase PCR. GAPDH was used as a loading control to show equal cDNA input (Fig. 1a). Our observations were confirmed by immunohistochemistry of pancreatic tissues of wildtype mice, in which no CTSG was detected in pancreatic acinar cells (Fig. 1b). In contrast, CTSG expressing cells were found in spleen tissues especially in cells within the vessels of wildtype mice, while no expression was found in CTSC −/− spleens, as expected. Our findings indicate that CTSG expression is absent in excretory pancreatic cells.

Effects of CTSG on trypsinogen processing in-vitro and protease activation, and cellular necrosis in acinar cells.
To examine whether CTSG has a direct effect on the cleavage of trypsinogen in vitro we co-incubated trypsinogen with active CTSG in vitro and measured active trypsin activity by using the substrate R110-(CBZ-Ile-Pro-Arg) 2 . Even high CTSG concentrations did not lead to an increase in trypsin activity in contrast to enterokinase, a known trypsinogen activator, that induced a rapid increase in trypsin activity (Fig. 2a). In order to clarify whether CTSG has a direct effect on intracellular trypsin activation in vivo we isolated pancreatic acinar cells and co-incubated with recombinant CTSG and supramaximal doses of cholecystokinin. One fraction was treated with CTSG alone. In preliminary experiments we had shown that enzymatic CTSG activity is still present in the cultivation media of acinar cells (data not shown). Supramaximally CCK-stimulated acinar cells showed an increase Figure 1. Cathepsin G is expressed in leukocytes but not in pancreatic acinar cells. (a) Leucocytes extracted from spleen express CTSG as shown by RT-PCR. No expression was detected in acinar cells investigated in two different samples (#1 and #2). GAPDH served as a loading control. In the negative control only water was used. (b) Cells with positive labelling for CTSG were found inside blood vessels that contain leukocytes as shown on 400-fold magnification in the spleen tissues. Pancreatic acinar cells were deficient of CTSG. Specificity of CTSG was proven in CTSG −/− tissues that showed no labelling. Full length gels are presented in Supplementary Figs 2 and 3.
of intracellular trypsin activity peaking at 20 minutes while the addition of CTSG, even at a high concentration (10 mM), did not affect trypsinogen activation. Intracellular cathepsin B (CTSB), a known trypsinogen activator, also rapidly increased in its activity within the first 20 minutes after CCK-stimulation and steadily decreased thereafter. It did not show any change of activity in the presence of CTSG (Fig. 2b). Moreover, the extent of cellular necrosis, measured by propidium iodide (PI) exclusion, was not altered by addition of CTSG (Fig. 2c). These results indicate that early protease activation in acinar cells and cell injury are independent of CTSG.
Decreased neutrophil invasion in ctSG −/− mice during acute pancreatitis. Since CTSG is expressed in neutrophil granulocytes and is responsible for activation of several other neutrophil proteases 13 we investigated whether depletion of CTSG would change the infiltration of neutrophils in the pancreas. Activities of myeloperoxidase (MPO), a neutrophil marker, increased during acute pancreatitis in pancreas homogenates showing the highest activities at 24 hours. In CTSG −/− mice this increase was delayed leading to a significantly lower MPO activity at 8 hours, which was no longer evident at 24 h (Fig. 3a). In the lungs MPO activities were highest at 8 hours and strongly decreased at 24 hours. In CTSG −/− mice MPO activity was significantly lower at  www.nature.com/scientificreports www.nature.com/scientificreports/ Severity of acute pancreatitis in ctSG −/− mice. To investigate whether depletion of CTSG alters the disease severity in acute pancreatitis mice we induced acute experimental pancreatitis using the caerulein-model. Both 129S2/SvPasCrl wild type and CTSG −/− mice received serial intraperitoneal injections of the CCK-analogue caerulein. Serum lipase and amylase rapidly increased at 8 h, but did not differ significantly between wild type and CTSG knock out animals (Fig. 4a). At 8 hours the digestive proteases trypsin and elastase were increased in pancreatic tissue but not significantly different in CTSG deficient mice (Fig. 4b). Local pancreatic injury, determined by histological damage, consisting of vacuolization, necrosis, and inflammatory cell infiltration, increased during acute pancreatitis but was independent of the presence of CTSG, as no significant changes were observed between www.nature.com/scientificreports www.nature.com/scientificreports/ the groups (Fig. 4c). It is well known that neutrophil granulocytes increase pancreatic damage by release of reactive oxygen species (ROS) 14 . Neutrophil cytosolic factor 1 (NCF-1) and NCF-2 are part of the NADPH oxidase pathway and are responsible for the production of superoxide anions. A defect of NADPH oxidase mediated ROS production in infiltrating neutrophils could not be observed in our model. Wildtype mice as well as CTSG deficient animals show a clear positive signal for NCF-1 (p47-phox) or NCF-2 (p67-phox) in infiltrating neutrophils (Fig. 4d). These results indicate that the course and severity of pancreatitis do not depend on CTSG expression in mice, despite the transient decrease of neutrophil invasion into the pancreas seen at 8 hours (Fig. 3). Moreover, serum activities of cytokines were not significantly altered at 8 hours in CTSG knockout mice. Apparently systemic cytokine levels depend on more sources than only neutrophil granulocytes ( Supplementary Fig. 1).

Discussion
Lysosomal enzymes are important regulators in the early phase of acute pancreatitis. They can intracellularly activate as well as degrade the zymogen trypsinogen which is physiologically activated by the brush border enzyme enterokinase in the duodenum 15 . Once proteolysed to its active form trypsin causes an activation of multiple other digestive proteases which leads to cellular injury, ultimately resulting in pancreatitis 16 . During the initial events digestive enzymes and lysosomal hydrolases co-localize in the same subcellular compartment of acinar cells with subsequent release of active trypsin to the cytosol 4 . In addition to acini, trypsinogen activation also occurs in macrophages that harbor CTSB and ingest acinar cell components including trypsinogen once they have infiltrated the pancreas 17 .
Subsequently, inflammatory cells are recruited to the pancreas that further aggravate the severity of pancreatitis. The two main cell populations entering the pancreas are neutrophil granulocytes and macrophages 14,17 . This infiltration already takes place at early time points, even before pancreatic necroses have evolved 18 . Tumor necrosis factor α (TNF-α) is one of the key mediators that cause trypsin activation and pancreatic injury. TNF-α is released by immune cells, and its inhibition 3 or deletion of the corresponding receptor 19 abrogates disease severity. The TNF-α mediated trypsin activation is dependent on CTSB 3 . Furthermore, nicotinamide adenine dinucleotide phosphate (NADPH) oxidase regulates trypsin activation by generating reactive oxygen species making trypsinogen more susceptible to hydrolysis by serine proteases 14 .
Recently CTSC was identified as a lysosomal hydrolase, whose deficiency ameliorated the severity of pancreatitis 8 . CTSC processes the neutrophil serine proteases elastase (NE), proteinase 3 (PR3) and cathepsin G (CTSG), which are stored in primary neutrophil granules and participate in the progression of many inflammatory disorders via a variety of mechanisms 20 : One of their primary functions is to degrade engulfed microorganisms but they are also released directly at sites of inflammation and play an important role in cell signalling, degradation of extracellular matrix and processing of cytokines 21,22 .
The role of neutrophil elastase (NE), syn. polymorphonuclear (PMN) granulocyte elastase, in acute pancreatitis has been investigated in earlier studies that showed its ability to cleave the cell contact molecule E-cadherin and to facilitate the dissociation of cell contacts, which than permits an increased transmigration of leukocytes into the pancreas. Conversely, elastase inhibition significantly reduced the translocation of leukocytes and decreased the disease severity 23 .
The impact of other serine proteases in acute pancreatitis is incompletely understood. CTSG is abundantly expressed in neutrophils which makes this enzyme an interesting candidate. Our study showed, that its absence had no effect on the severity of pancreatitis despite a reduction of neutrophil invasion into the pancreas. The latter was transient. Moreover, protease activation in acinar cells, a hallmark of early pancreatitis, was independent of CTSG.
One explanation for the reduced neutrophil invasion into the pancreas of CTSG −/− mice might be the capacity of CTSG to cleave vascular endothelial (VE)-cadherin, a cell contact protein expressed on the surface of the endothelial cells in blood vessels 24 . Absence of CTSG will maintain the barrier of the endothelial cells and thus impair the invasion of neutrophil granulocytes. In contrast, the adherens junction protein E-cadherin between acinar cells is not substantially cleaved by CTSG as we could show previously and dissociation of acinar cells was independent of presence of CTSG 8 . These findings may explain why the severity of acute pancreatitis is not ameliorated in CTSG −/− mice compared to the wildtype.
Another possible explanation would be, that the moderately reduced number of neutrophils is not sufficient to improve the course of acute pancreatitis. Furthermore, this reduction occurred only transiently at 8 hours, and was no longer evident at 24 hours. The inflammatory infiltrate also consists of invading macrophages, which ingest digestive zymogens from injured acinar cells and activate zymogens in a CTSB dependent manner 17 . As long as macrophages are entering the pancreas, perpetuation of acute pancreatitis occurs due to persistent trypsinogen activation. Despite a reduced secretion of pro-inflammatory cytokines in isolated CTSG −/− neutrophils in vitro systemic cytokine levels in serum were not significantly different between wildtype and CTSG knockout mice. Other inflammatory cells appear to contribute to a greater extend to cytokine secretion which could make the reduced cytokine release from CTSG −/− neutrophils clinically irrelevant. In addition to mast cells and macrophages, also lymphocytes, including B-and T-cells and NK-cells are known to have a role in systemic inflammation [25][26][27] .
Mice deficient in both neutrophil elastase and cathepsin G have been investigated in a previous study, that showed a normal recruitment of neutrophils to the site of inflammation upon induction of sterile peritonitis. In addition, normal TNF-α levels were produced in these mice 28 . These findings are in line with our observations as acute pancreatits is a primarily sterile infection. However, susceptibility to fungal infections was more severe in double NE and CTSG knockout mice 28 . CTSG −/− mice were also investigated after infection with Mycobacterium bovis bacillus Calmette-Guérin (BCG) and showed an impaired pathogen elimination 29 .
Rats that underwent ischemia and reperfusion (I/R)-induced pancreatitis and were treated with Lex032, a broad serine protease inhibitor, preserved capillary perfusion and showed reduced serum IL-6 and amylase concentrations. www.nature.com/scientificreports www.nature.com/scientificreports/ These results indicate that serine protease inhibition in general has an anti-inflammatory potential 11 but probably CTSG inhibtion alone is not sufficient to alter the severity of the disease. However, comparability of these results is challenging, as the authors used a different experimental model for acute pancreatitis than in our study.
In conclusion, our results indicate that the absence of cathepsin G leads to a transient reduction in neutrophil granulocyte invasion but leaves pancreatitis severity largely unaffected. This effect, however, is less prominent than for the related protease PNM elastase 23 , or the calcium binding protein S110A9 18 , both of which facilitates leucocyte infiltration via cleavage of cell-cell contacts between pancreatic acinar cells, which CTSG does not. The former therefore represent more promising treatment targets than CTSG.

Materials and Methods
Reagents, substrates and antibodies. Collagenase from Clostridium histolyticum (EC.3.4.24.3) was purchased from SERVA (lot no. 14007, Heidelberg, Germany) and used for acinar cell isolation. Cathepsin G enzyme (human neutrophil) was obtained from Merck (Darmstadt, Germany). Myeloperoxidase (MPO) enzyme from human polymorphonuclear leukocytes, enterokinase from porcine intestine, caerulein, and cholecystokinin (CCK) were obtained from Sigma (Taufkirchen, Germany). Ketamine and xylazine were from Selectavet (Weyarn-Holzolling, Germany). Amylase and lipase were quantified using a kit from Roche-Hitachi (Grenzach-Wyhlen, Germany). Anti-cathepsin G (ab64891) was purchased from abcam (Cambridge, UK). Anti-p47-phox (sc-17844) and anti-p67-phox (sc-374510) were purchased from Santa Cruz Biotechnology (Dallas, TX). Protease activity was measured by adding the following substrates: for trypsin R110-(CBZ-Ile-Pro-Arg) 2  Immunohistochemical staining of p47 and p67. 2 µm thick slides of paraffin embedded tissue were used for immunohistochemical staining. Antigen retrieval was performed by cooking the slides in DAKO Antigen retrieval buffer (DAKO, Santa Clara, CA) for 30 min. Endogeneous peroxidase activity was blocked by 3% H 2 O 2 . Slides were blocked with PBS containing 5% bovine serum albumine (BSA) and 0.3% Triton X-100 for 1 h. Antibodies were diluted in PBS containing 1% BSA and 0.3% Triton X-100 and added on the slides and incubated over night at 4 °C. Anti-mouse secondary antibody was used from DAKO. DAB staining kit was used from Vector Laboratories (Burlingame, CA). After DAB reaction the slides were counterstained with Hematoxylin and mounted with VectaMount mounting media from Vector Laboratories.
Reverse transcriptase pcR. Purification of RNA was done using TRIzol reagent (Thermo Fisher Scientific, Waltham, MA) from isolated acinar cells and leukocytes from spleen tissue. Single strand cDNA synthesis was generated from 500 ng RNA as starting material using the Reverse Transcriptase System from Promega (Mannheim, Germany). For cathepsin G (NM_007800) the following gene specific primers were used: sense primer: 5′-GAG TCC AGA AGG GCT GAG TG-3′ and antisense primer: 5′-CCT TTC TCG CAT TTG GAT GT-3′ amplifying a 136 bp fragment 30 . Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a reference for equal cDNA input. The PCR mixture contained 2x FastStart Master Mix (Roche, Basel, Switzerland), primers (200 nM each), cDNA (approximately 100 ng) in a final volume of 50 μl. Amplification was carried out in an Eppendorf cycler for 35 cycles, annealing temperatures were 62 °C for CTSG primers and 60 °C for GAPDH primers. Negative controls without DNA were set up for each PCR reaction. PCR products were visualized under ultraviolet light on a 2.0% agarose gel containing ethidium bromide.
induction of acute pancreatitis in mice. All animal experiments were performed according to state institutional guidelines and animal facility protocols after prior approval by the institutional animal care committee. Wildtype 129S2/SvPasCrl wildtype mice were obtained from Charles River Laboratories (Sulzfeld, Germany). CTSG −/− mice [129-CtsgG tm1.1Roes/(H) ] were purchased from the European Mouse Mutant Archive (EMMA). Acute pancreatitis was induced by hourly intraperitoneal injections of caerulein (50 µg/kg/body weight) for up to 8 hours 31 . All mice were starved over night with access to water ad libitum. Mice were sacrified at 0 h, 8 h, and 24 h after the first caerulein injection.
Tissue was harvested immediately after sacrificing the animals. Pancreas and lung samples were frozen in liquid nitrogen and stored at −80 °C for later analysis. For histology, tissue was fixed in 4.5% formalin for paraffin embedding. Serum samples were and stored at −20 °C. isolation of Acinar cells. Acinar cells were prepared by collagenase digestion as previously described 32  www.nature.com/scientificreports www.nature.com/scientificreports/ activity was monitored for up to 60 min after CCK stimulation and aliquots were taken every 20 minutes. Untreated cells served as controls. In vivo measurement of protease activation in living acinar cells was performed in cell medium (pH 7.4) containing 24.5 mM HEPES, 96 mM NaCl, 11.5 mM glucose, 6 mM KCl, 1 mM MgCl 2 6H 2 O, 0.5 mM CaCl 2 2H 2 O, 2.5 mM NaH 2 PO 4 H 2 O, 5 mM sodium fumarate, 5 mM sodium glutamate, 5 mM sodium pyruvate, and 1% BSA and DMEM in a flurometer (FLUOStar Omega, BMG Labtech GmbH, Ortenberg, Germany) as described previously 8 . For trypsin activity the substrate R110-(CBZ-IPR) 2 (10 µM), and for CTSB activity the substrate AMC-Arg 2 (20 µM) were used. Necrosis was determined by measurement of propidium iodide incorporation in acinar cells. All measurements were done in triplicates.
Histology and immunohistochemistry. Paraffin sections from the pancreas and spleen were used for haematoxylin and eosin (H&E) staining. For assessment of damage a modified score, adapted from Niederau et al., was used 33 . The scoring was based on the extent of necrosis, vacuolization of acinar cells and invasion of inflammatory cells into the pancreas. A minimum of 10 visual fields from each animal was investigated. For immunohistochemistry antigen retrieval was performed with Target Retrieval Solution (Dako). CTSG was stained using an anti-cathepsin G antibody (ab64891), diluted 1:50 in 20% fetal calf serum (FCS) from PAN Biotech (Aidenbach, Germany) and incubated over night at + 4 °C followed by staining with a peroxidase-HRP labeled antibody (1:1000 dilution). CTSG stained cells were visualized using the DAB (3,3′-diaminobenzidine) Substrate Kit (Vector Laboratories, Burlingame, CA, USA) according to the manufacturer's instructions.
isolation of neutrophils. Leukocytes were isolated from murine spleens of CTSG −/− and WT mice as described previously 8 . Briefly, after passing the cell suspension over a 70 μm sterile nylon filter (BD Falcon, Amsterdam, The Netherlands) to remove cell aggregates and debris erythrocytes were lysed with 155 mM NH 4 Cl, 10 mM KHCO 3 , and 0.1 mM EDTA 34 . Neutrophils were isolated from total leukocytes using MACS ® cell separation kit according to the manufacturer's instructions (Miltenyi Biotech, Bergisch-Gladbach, Germany) by using Ly6g as a specific antibody.
Biochemical assays. Serum lipase and amylase activities were measured by photometric assays (Roche Hitachi, Grenzach-Whylen, Germany) as kinetics over 30 min with an absorbance at 570 nm at 37 °C. Purified enzymes (purchased from Sigma) served as controls. Trypsin and elastase activity were determined from pancreatic tissue homogenates and were measured at 37 °C in a buffer containing 100 mM Tris, 5 mM CaCl 2 pH 8.0, by adding 10 µM of R110-Ile-Pro-Arg for Trypsin or (CBZ-Ala-Ala-Ala-Ala) 2 -R110 for elastase as enzyme kinetic over 1 h at 37 °C at an excitation wavelength of 485 nm and an emission wavelength of 520 nm.
For in vitro proteolytic cleavage of trypsinogen either enterokinase (40 mU/ml) or cathepsin G enzyme (50 mU/ml) were added to 0.05 pg/ml bovine trypsinogen and incubated for 1 hour. Active trypsin was measured as enzyme kinetic by adding 10 µM of R110-Ile-Pro-Arg in 100 mM Tris, 5 mM CaCl 2 pH 8.0 buffer.
Myeloperoxidase (MPO) activity measurement was performed as previously described 3 . Briefly pancreatic tissue was homogenized on ice in 20 mM potassium phosphate buffer (pH 7.4) and centrifuged at 20,000 x g at 4 °C for 10 min. The pellet was resuspended in 50 mM potassium phosphate buffer (pH 6.0) containing 0.5% cetyltrimethylammonium bromide. The suspension was frozen in liquid nitrogen and thawed in cycles, sonicated twice and centrifuged at 20,000 x g at 4 °C for 10 min. MPO activity was measured in 50 mM potassium phosphate buffer (pH 6.0) containing 0.53 mM O-dianisidine and 0.15 mM H 2 O 2 . The initial increase in absorbance was measured at room temperature with a Spectramax Spectrophotometer (Molecular Devices, Sunnyvale, CA). The results were expressed in units of MPO activity on the basis of 1 unit being able to oxidize 1 μmol H 2 O 2 per minute per milligram pancreatic protein 32 . Cytokines TNF-α, IL-6, IL-12, IFN γ, and MCP-1 were measured from serum samples by FACS analysis with the CBA mouse-inflammation-kit according to the manufacturer's instructions (Becton Dickinson, Heidelberg, Germany). Statistical analysis. All data are expressed as means ± SEM from at least four animals per experiment in each group. Statistical analyses were performed by SigmaPlot (Systat Software Inc, Erkrath, Germany) and SigmaStat (Systat Software Inc, Erkrath, Germany) using the unpaired 2-tailored student t-test or analysis of variances (ANOVA) for samples without normality. Differences were considered significant for p < 0.05. ethical approval and informed consent. All experimental protocols were approved by the institutional animal care committee (Landesamt für Landwirtschaft, Lebensmittelsicherheit und Fischerei Mecklenburg-Vorpommern).

Data availability
The datasets generated during and analyzed during the current study are available from the corresponding author on request.