Inflammatory mediators including chemokines play a critical role in acute pancreatitis. The precise nature of early inflammatory signals within the pancreas remains, however, unclear. We examined the ability of isolated pancreatic acini to synthesize CC chemokine monocyte chemotactic protein-1 (MCP-1) and CXC chemokine cytokine-induced neutrophil chemoattractant (CINC) and the response to the secretagogue cerulein at physiological and supraphysiological concentrations. Isolated rat pancreatic acini maintained in short-term (≤48 h) primary culture constitutively synthesized MCP-1 and CINC. Cerulein (10−7M; supramaximal dose) increased production of MCP-1 but not CINC. Cerulein-induced increase in MCP-1 synthesis was accompanied by increase in nuclear factor (NF)-κB activation shown by EMSA. Pretreatment with NF-κB inhibitors N-acetylcysteine (NAC) and N-tosylphenyalanine chloromethyl ketone (TPCK) blocked cerulein-induced NF-κB activation and abolished cerulein's effect on MCP-1 synthesis. Pretreatment with calcium antagonist BAPTA-AM also blocked cerulein's effect on MCP-1 synthesis. These results indicate that isolated acini synthesize MCP-1 and CINC and support the idea of acinar-derived chemokines as early mediators of inflammatory response in acute pancreatitis. Although cerulein hyperstimulation increased MCP-1 synthesis by a calcium-dependent mechanism involving NF-κB activation, CINC synthesis was not affected. This suggests that regulation of CC and CXC chemokines within acinar cells may be quite different.
- monocyte chemotactic protein-1
- cytokine-induced neutrophil chemoattractant
- nuclear factor-κB.
activation of digestive enzymes within pancreatic acinar cells is a critical initiating event in acute pancreatitis resulting in acinar cell damage and a localized inflammatory response (16, 28, 34). Secreted bioactive molecules from infiltrating leukocytes contribute to local damage and to the subsequent systemic inflammatory response, which may result in multiple organ dysfunction and death (5). The initial signals that recruit leukocytes into the pancreas are incompletely defined, although several inflammatory mediators have been implicated (5, 15).
The chemokines are a family of small (8–10 kDa) cytokines with activating and chemotactic effects on leukocyte subsets. CC chemokines, such as monocyte chemotactic protein-1 (MCP-1), principally affect monocytes, whereas CXC chemokines that possess the ELR motif at the amino terminal tend to act on neutrophils (1). Interleukin (IL)-8 is the best characterized of the ELR-positive CXC chemokines. Its plasma levels are elevated early in the course of acute pancreatitis and correlate with disease severity (3,17). Treatment with a monoclonal anti-human IL-8 antibody (WS-4) was shown recently to decrease acute lung injury in experimental pancreatitis in rabbits (26). The involvement of CC chemokines in acute pancreatitis is suggested by decreased pulmonary damage in CCR1 receptor knockout mice after induction of acute pancreatitis (12).
There are important differences in the profile of chemokines seen in different species. In the rat, for example, although there is a direct homolog of MCP-1, there is none for IL-8. Rather, cytokine-induced neutrophil chemoattractant (CINC), which is a homolog of the human ELR-positive CXC chemokine growth-related oncogene-α (GRO-α), is the best characterized of the rat CXC chemokines. We recently demonstrated (8) increased expression of MCP-1 and, to a lesser extent, of CINC on pancreatic acinar cells by immunohistochemistry soon after induction of acute pancreatitis by cerulein hyperstimulation or by infusion of sodium taurocholate into the biliopancreatic duct in the rat. mRNA for CINC and MCP 1 was detectable within pancreas as early as 1 h after induction of pancreatitis, suggesting that both are very early mediators of the inflammatory response in acute pancreatitis. In the present study we show constitutive expression of MCP-1 and CINC by isolated rat pancreatic acinar cells. In addition, we demonstrate increased production of MCP-1, but not CINC, by a nuclear factor (NF)-κB- and Ca2+-dependent mechanism after treatment with supramaximal doses of cerulein.
MATERIALS AND METHODS
Preparation of pancreatic acini.
Pancreatic acini were obtained from rat pancreas by collagenase treatment as described previously (4, 11). Briefly, pancreata from Wistar rats (100–125 g) were removed under aseptic conditions, infused with collegenase buffer A (in mM: 140 NaCl, 4.7 KCl, 1.13 MgCl2, 1 CaCl2, 10 glucose, and 10 HEPES, pH 7.2) containing 200 IU/ml collagenase and 0.5 mg/ml soybean trypsin inhibitor, and incubated in a shaking water bath for 10 min at 37°C. The digested tissue was passed through 50 mg/ml BSA and washed twice with collegenase buffer A before culture.
Short-term primary culture of pancreatic acini.
Pancreatic acini were suspended in Weymouth's MB 752/1 medium containing 0.1% BSA, 0.5 mg/ml soybean trypsin inhibitor, 25 ng/ml epidermal growth factor, and antibiotics (9). Equal volumes (∼4 × 105 cells/ml) were then aliquoted into 24-well plates and incubated at 37°C in a humidified atmosphere of 10% CO2-90% air. Acini were cultured for periods of ≤48 h. Cell viability was assessed by trypan blue exclusion.
Experiments were performed to examine the effects of culture time and cerulein (Research Plus), tumor necrosis factor (TNF)-α (Sigma), and thapsigargin (Sigma) treatment on MCP-1 and CINC synthesis and on amylase secretion. Further experiments were undertaken to investigate effects of the NF-κB antagonists N-acetylcysteine (NAC; Sigma), an antioxidant, and N-tosylphenyalanine chloromethyl ketone (TPCK; Sigma), an inhibitor of serine proteases, on amylase secretion, cerulein- and TNF-α-induced MCP-1 synthesis, and NF-κB activation. Experiments were also undertaken to investigate effects of the intracellular Ca2+ chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA)-AM (Sigma) on cerulein- and TNF-α-induced MCP-1 synthesis.
Transmission electron microscopy.
Small pieces of tissue (∼1 mm) were fixed in 3% glutaraldehyde in 0.1 M cacodylate buffer pH 7.4 for 1 h (minimum) at room temperature. The specimens were washed in 0.1 M cacodylate buffer (pH 7.4) for three washes of 5 min each. Specimens were then fixed in 1% osmium tetroxide in 0.1 M cacodylate buffer (pH 7.4) for 1 h at room temperature and washed in 0.1 M cacodylate buffer (pH 7.4) for three washes of 5 min each. The samples were then dehydrated and embedded in epon-araldite resin. The blocks were sectioned on a Reichert Ultracut E ultramicrotome set to give 120-nm-thick sections. The sections were picked up onto 200-mesh hexagonal thin bar copper grids and stained in 2% uranyl acetate for 20 min followed by Reynolds lead citrate for 5 min. The sections were examined in a Philips CM10 transmission electron microscope operated at 80 kV. Negatives were recorded on Kodak 4489 film.
Amylase activity was measured with a kinetic spectrophotometric assay. Media were incubated with the substrate 4,6-ethylidene-(G7)-p-nitrophenyl(G1)-1-d-maltoheptoside (Sigma) for 2 min at 37°C, and absorbance at 405 nm was measured every minute for a subsequent 2 min (4, 27).
MCP-1 and CINC assays.
Conditioned media were assayed for MCP-1 and CINC with a sandwich ELISA according to the manufacturer's instructions. Matched antibody pairs against rat MCP-1 and recombinant rat MCP-1 were obtained from Pharmingen. Matched antibody pairs against rat CINC and recombinant CINC were obtained from R&D Systems. Briefly, anti-CINC or anti-MCP-1 primary antibody was aliquoted onto ELISA plates and incubated at 4°C overnight. Samples and standards were incubated for 2 h, the plates were washed, and a biotinylated anti-CINC or anti-MCP-1 antibody was added for 1 h. Plates were washed again, and streptavidin bound to horseradish peroxidase was added for 30 min. After a further wash, tetramethylbenzidine was added for color development and the reaction was terminated with 0.18 M H2SO4. Absorbance was measured at 450 nm.
Nuclear protein extraction.
Nuclear proteins were extracted as previously described (23). Briefly, pancreatic acini were washed once with PBS and the pellet was resuspended in buffer A (in mM: 50 NaCl, 10 HEPES, pH 8.0, 1 EDTA, 0.5 spermidine, and 0.15 spermine with 0.5 M sucrose and 0.2% Triton X-100). After a 10-min incubation at 4°C, nuclei were pelleted by centrifugation (3,500 g for 3 min at 4°C), washed once with buffer B (in mM: 50 NaCl, 10 HEPES, pH 8.0, 1 EDTA, 0.5 spermidine, and 0.15 spermine with 25% glycerol) and incubated for 30 min in buffer C (in mM: 350 NaCl, 10 HEPES, pH 8.0, 1 EDTA, 0.5 spermidine, and 0.15 spermine with 25% glycerol). After a 1-min centrifugation (3,500 g, 4°C) the supernatant was aliquoted, snap-frozen in liquid nitrogen, and stored at −80°C.
Electrophoretic mobility shift assay.
Electrophoretic mobility shift assays (EMSAs) were carried out essentially as described previously (23). Reaction mixtures (in a final volume of 10 μl) containing 4% Ficoll, 20 mM HEPES, pH 7.5, 35 mM NaCl, 60 mM KCl, 0.01% NP-40, 2 mM dithiothreitol, 0.1 mg/ml BSA, 1 μg Poly(dI-dC), and the indicated amount of nuclear protein extract were incubated for 10 min at room temperature. For supershift assays, antibodies against p50, p65, and c-rel (Santa Cruz Biotechnology), when used, were added at this time and the incubation period was extended to 20 min.32P-labeled oligonucleotide (0.02 pmol) was added, and the reaction continued for a further 20 min. In competition experiments, unlabeled oligonucleotide was added before addition of32P-labeled oligonucleotide. The reaction mixtures were loaded onto 5% polyacrylamide gels containing 0.5× TBE (in mM: 89 Tris, 89 boric acid, and 0.2 EDTA). Electrophoresis was performed at 200 V in the same buffer (0.5× TBE), and the gels were then transferred to Whatman 3MM filter paper, dried, and autoradiographed. The following oligonucleotides were used: 5′-AGTTGAGGGGACTTTCCCAGGC-3′ (containing the NF-κB consensus binding site) or 5′-AGTTGAGGCCACTAACCCAGGC-3′ (containing a mutated NF-κB site).
Results are presented as means ± SE of a typical experiment with six replicates for each condition. Each experiment was repeated at least three times. In Figs. 1-4 and 6-10, error bars denote SE and the absence of such bars indicates that the SE is too small to illustrate. Initial statistical comparisons were made with a Kruskal-Wallis test and subsequent paired comparisons with a Mann-Whitney U-test; the level of significance was set at 5%.
Viability and biological activity of rat pancreatic acinar cells in primary culture.
The morphology of rat pancreatic acini in primary culture, as evidenced by the ultramicroscopic examination of epon-embedded sections, was unchanged up to 48 h, the latest time point studied (Fig.1 A). At 48 h >95% of cells were viable, as determined by trypan blue exclusion. Moreover, amylase levels as a function of DNA remained constant for up to 48 h (Fig. 1 B). Amylase secretion in response to cerulein treatment demonstrated a biphasic dose-response curve: there was an increase in amylase secretion with increasing dose of cerulein up to the physiological dose of 10−10 M, followed by a reduction in secretion with supramaximal stimulation with up to 10−7M cerulein. The biphasic dose-response curve was observed in acini even after 24 h in culture (Fig. 2).
Production of MCP-1 and CINC by rat pancreatic acini as a function of time.
In the absence of any stimulation, the levels of both MCP-1 and CINC in conditioned media increased with time up to 48 h, the latest time point analyzed (Fig. 3).
Production of MCP-1 and CINC by rat pancreatic acini in response to cerulein treatment.
To investigate the effect of cerulein treatment on chemokine production, pancreatic acini were incubated with cerulein for 30 min (37°C). Conditioned media were then analyzed by ELISA. Treatment of pancreatic acini with physiological doses of cerulein (10−10 M) had little effect on MCP-1 synthesis, but higher doses (10−7 M) of cerulein increased MCP-1 synthesis (Fig.4 A). After 24 h in culture, treatment of acini with 10−7 M cerulein still increased synthesis of MCP-1 (Fig. 4 A). Synthesis of CINC, however, was not affected after treatment with cerulein in both fresh acini and acini after 24 h in culture (Fig. 4 B).
NF-κB activation by supramaximal stimulation with cerulein.
Detectable levels of NF-κB DNA binding were observed in nuclear extracts from nontreated acini (Fig.5 A). An increase in NF-κB DNA binding was, however, observed in nuclear extracts from acini treated with a supramaximal dose of cerulein (Fig. 5 A). The addition to EMSA reactions of an excess of unlabeled competitor oligonucleotide containing the NF-κB DNA binding sequence resulted in a decrease in the binding of complexes arbitrarily designateda, b, c, and d (Fig.5 B). These complexes were not affected by the addition of unlabeled competitor oligonucleotide containing a mutated NF-κB DNA binding sequence (Fig. 5 B). Supershift assays with antibodies against p65, p50, and c-rel revealed thatcomplexes b and c contained p65 whereascomplexes a, b, and d contained p50 (Fig. 5 C). Addition of anti-c-Rel antibody to reaction mixes had no effect on the pattern of binding (Fig. 5 C).Complex b may represent the classic NF-κB dimer p65-p50 and complex d the p50-p50 homodimer. Cerulein treatment resulted in increases in complexes b and c but not in complex d.
Effect of NF-κB inhibitors on MCP-1 and CINC production.
Experiments were then performed to examine the effect of two NF-κB inhibitors, NAC and TPCK, on synthesis of MCP-1 and CINC by pancreatic acini. Pretreatment of pancreatic acini with either antagonist blocked both NF-κB activation and the upregulation of MCP-1 synthesis that occurred in response to cerulein hyperstimulation in freshly isolated acini or in acini after 24 h in culture (Fig.6, A and C). Neither inhibitor had any effect on basal production of MCP-1 from nonstimulated acini. In parallel experiments, both NAC and TPCK inhibited the cerulein-induced increase in NF-κB complexes b and c (Fig. 6 B). Neither inhibitor had any effect on amylase secretion in response to different doses of cerulein (data not shown).
Effect of TNF-α on MCP-1 and CINC production.
To investigate the effect of TNF-α treatment on chemokine production, pancreatic acini were incubated with TNF-α for 30 min (37°C). Conditioned media were then analyzed by ELISA. Treatment of pancreatic acini with TNF-α increased MCP-1 synthesis (Fig.7). Synthesis of CINC, however, was not affected after treatment with TNF-α (Fig. 7). Pretreatment of acini with the NF-κB inhibitors NAC and TPCK blocked the upregulation of MCP-1 synthesis in response to TNF-α (Fig.8).
Effect of thapsigargin and BAPTA-AM on MCP-1 and CINC synthesis.
To investigate the effect of the endoplasmic reticulum Ca2+-ATPase inhibitor thapsigargin, which increases intracellular Ca2+ levels, on chemokine production, pancreatic acini were incubated with thapsigargin for 30 min (37°C). Conditioned media were then analyzed by ELISA. Treatment of pancreatic acini with thapsigargin increased MCP-1 synthesis in fresh acini or after 24 h in culture. CINC production was not affected at either time point (Fig. 9).
Experiments were then performed to examine the effect of the Ca2+ chelator BAPTA-AM on MCP-1 synthesis by pancreatic acini. Pretreatment of pancreatic acini with BAPTA-AM blocked the upregulation of MCP-1 synthesis that occurred in response to cerulein hyperstimulation (Fig.10 A) or TNF-α treatment (Fig. 10 B).
Administration of supramaximal doses of the CCK analog cerulein to experimental animals leads to acute pancreatitis, which early on is associated with activation of trypsinogen, vacuole formation, and disrupted release of zymogen granules (16, 29, 33). Abnormalities of calcium signaling are thought to play a critical role in these events (33). Hyperstimulation of isolated pancreatic acini with CCK and its analog cerulein causes morphological changes that are similar to those seen in vivo and provides an ideal model with which to dissect the intra-acinar events in vitro. Whereas a physiological concentration of CCK does not cause intracellular trypsin activation, exposure to supramaximal CCK does have such an effect. This activation depends on a rise in intracellular Ca2+concentration and can be blocked by the intracellular Ca2+chelator BAPTA-AM (28, 33).
Most cells have the ability to express and synthesize chemokines, and typically this occurs rapidly after cellular injury (1). The profile of chemokines expressed at a site of injury is one of the major factors that determine the nature of the subsequent leukocyte infiltrate (1, 2). Inflammatory mediators involved in acute pancreatitis, particularly those involved in the early stages of an attack, may be derived directly from pancreatic acinar cells. For example, pancreatic acinar cells have been shown to produce and release TNF-α and via specific receptors these cells respond (by apoptosis) to TNF-α (18).
Recent work in our laboratory (8) demonstrated expression of MCP-1 on pancreatic acini in cerulein-induced pancreatitis in the rat. Using the same model, Grady et al. (15) showed an increase in pancreatic MCP-1 mRNA (by Northern blot analysis) on induction of acute pancreatitis, although the cellular origin of MCP-1 was not established. Two groups have reported production of MCP-1 by isolated pancreatic acini (7, 34). In the study of Blinman et al. (7), MCP-1 mRNA levels were upregulated in individual isolated acinar cells. Inhibitors of p38 mitogen-activated protein kinase (MAPK) and NF-κB reduced levels of MCP-1 mRNA. In contrast to isolated acini, no or very low MCP-1 expression was detected in normal rat pancreas, suggesting thereby that activation of p38 MAPK, activation of NF-κB, and synthesis of cytokines occur as a result of removal of the pancreas from the animal and isolation of acini (7). Our results confirm production of MCP-1, a CC chemokine, by isolated acini in short-term primary culture, and although physiological doses of cerulein did not affect its synthesis, treatment with supramaximal doses of cerulein increased MCP-1 synthesis.
The transcription factor NF-κB is involved in the regulation of several cytokines and chemokines that together have proinflammatory effects (13, 30). In nonstimulated cells NF-κB/Rel proteins are sequestered in the cytoplasm in ternary complexes of NF-κB homo- or heterodimers bound to specific inhibitor proteins (IκBs). After stimulation, IκB is degraded by proteasomes and active NF-κB/Rel is translocated to the nucleus (13,30). Cerulein hyperstimulation has been reported to cause NF-κB activation in isolated acini and in pancreas in vivo (19,20, 32).
We observed that the increase in MCP-1 production by isolated pancreatic acini correlated with an increase in nuclear NF-κB DNA binding activity. Interestingly, not all of the observed NF-κB-containing complexes were affected by cerulein hyperstimulation. The quickly migrating complex d, which contained p50 (Fig. 5 C) and likely represents p50-p50 homodimer, was not altered by cerulein hyperstimulation; p50 homodimers have been associated with transcription repression (10,14). In contrast, the potently transactivating heterodimer p65-p50 (complex b; Fig. 5) exhibited increased DNA binding activity. These data are consistent with the observed increase in MCP-1 production.
We further investigated the role of NF-κB in cerulein-induced MCP-1 production by using two inhibitors of NF-κB activation, the antioxidant NAC (24) and the proteasome inhibitor TPCK (22). Reactive oxygen intermediates are produced and released concurrently with NF-κB activation during the inflammatory response (13), and antioxidants such as NAC are known to be inhibitors of NF-κB activation (24). The importance of reactive oxygen species in the early pathogenesis of acute pancreatitis is well established (30). Proteasome inhibition prevents activation of NF-κB by blocking the degradation of the inhibitor protein IκB. Consequently, NF-κB is retained in the cytoplasm in an inactive form. Pretreatment with either of the NF-κB antagonists, NAC or TPCK, prevented both cerulein-induced activation of NF-κB and cerulein-mediated production of MCP-1.
Treatment of pancreatic acini with thapsigargin, which increases cytosolic Ca2+ levels, leads to increased MCP-1 production. Pretreatment of isolated acini with the intracellular Ca2+chelator BAPTA-AM prevented cerulein-mediated increase in MCP-1 synthesis, suggesting that the increase in cytosolic Ca2+seen after cerulein hyperstimulation is a prerequisite for increased MCP-1 production.
TNF-α has been shown to induce MCP-1 production in many cell types. It is also an important inflammatory mediator in acute pancreatitis. Treatment of isolated acini with TNF-α increased production of MCP-1. This was NF-κB and Ca2+ dependent because pretreatment of isolated acini with BAPTA-AM and NF-κB inhibitors inhibited TNF-α-mediated production of MCP-1. There are thus similarities between the regulation of chemokine production in response to cerulein and TNF-α. Furthermore, TNF-α had no effect on CINC production by isolated acini, reinforcing the concept of a differential regulation of these two chemokines in pancreatic acinar cells.
Supraphysiological concentrations of CCK have recently been shown to induce expression of mob-1, a CXC ELR-negative chemokine, in rat pancreatic acini through the activation of the transcription factor NF-κB. CCK induced NF-κB nuclear translocation, and DNA binding was also blocked by BAPTA-AM (21). ELR-negative CXC chemokines have little effect on neutrophils, and their profile of activity resembles in many ways that of CC chemokines such as MCP-1. For example, ELR-negative CXC chemokines, like CC chemokines, recruit and activate lymphocytes and monocytes (25).
Production of chemokines, including CINC, by nonstimulated isolated acini may in part be explained by the stress of isolation and culture (7). Increased CINC mRNA levels, however, were demonstrated in pancreas after induction of acute pancreatitis by cerulein hyperstimulation in vivo in rats (8, 19). We showed previously (6) that plasma levels of CINC are elevated in acute pancreatitis and that a blocking anti-CINC antibody reduces lung injury. Furthermore, we have shown that circulating levels of GRO-α, the human homolog of CINC, are elevated in severe pancreatitis (31). Together these findings suggest that CINC is an important mediator in the inflammatory response in acute pancreatitis. In our experiments, however, neither cerulein hyperstimulation nor TNF-α treatment stimulated CINC production in vitro from pancreatic acini. These results suggest that CINC production in pancreatic acinar cells is regulated by a mechanism independent of NF-κB or Ca2+.
This study demonstrates that MCP-1 and CINC are produced by isolated pancreatic acini and supports the idea that they may be involved in the early inflammatory response in acute pancreatitis. Furthermore, the differential regulation of CINC compared with MCP-1 suggests that there are fundamental differences in the regulation of chemokine synthesis within the acinar cell.
This work was supported by Wellcome Trust Project Grant 003393.
Address for reprint requests and other correspondence: M. Bhatia, Dept. of Pharmacology, National Univ. of Singapore, Faculty of Medicine, 10 Kent Ridge Crescent, Singapore 119260.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society