- Email: [email protected]
Mechanisms of intracellular zymogen activation
BEST
Baillière’s Clinical Gastroenterology Vol. 13, No. 2, pp 227–240, 1999
B A I L L I È R E ’ S
2
PRACTICE & RESEARCH
Mechanisms of intracellular zymogen activation Fred S. Gorelick*
MD
Professor of Medicine and Cell Biology Department of Medicine and Cell Biology, VA Connecticut Healthcare System, Yale University School of Medicine, West Haven, Connecticut 06516, USA
Taiichi Otani
MD, PhD
Assistant Professor Department of Hepatobiliary and Pancreatic Surgery, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-0033, Japan
The pancreatic acinar cell is potentially the initial site of injury that begins the series of events leading to acute pancreatitis. Pathological intrapancreatic zymogen activation occurs in experimental pancreatitis in animals and in human pancreatitis. Intracellular activation has been clearly linked to aberrant zymogen processing in one form of hereditary pancreatitis; in this genetic disease a mutation in cationic trypsinogen may eliminate the degradation of any trypsin activated in the acinar cell. Recent studies have also provided the first direct evidence that trypsinogen activation takes place early in the course of caerulein-induced pancreatitis; parallel studies have used isolated pancreatic acini and conditions that simulate those that cause pancreatitis in vivo to demonstrate that zymogens can be pathologically activated in isolated cells. A unique acinar cell pathway regulates the intracellular proteinase processing of zymogens to their active forms. Stimulating the acinar cell with supramaximal concentrations of cholecystokinin (CCK) or carbamylcholine can activate this pathway. The activation depends on a low pH compartment within the acinar cell and activation of an intracellular serine protease. A marker of trypsinogen processing, the trypsinogen activation peptide (TAP), is generated in acinar cell compartments that do not overlap with secretory granules. This compartment overlaps with a marker of recycling endosomes and lysosomes. Thus, zymogen processing within the acinar cell proceeds in a distinct subcellular compartment and is dependent on a low pH environment and activation of serine proteases. Key words: PCA1; Trypsinogen; TAP; CCK; low pH; serine protease.
The generation of pancreatitis involves activating a cascade of pathological events. One of the critical initiating events appears to be the intracellular activation of pancreatic zymogens. The pancreas has developed a number of mechanisms that are designed to prevent zymogen activation within the acinar cell or quench the activity of small amounts of active enzymes. However, in pathological states these protective mechanisms may be overwhelmed and pancreatitis is initiated. Although investigators have long suggested that zymogen activation within the pancreas may initiate pancreatitis, only recently have studies demonstrated that such * Address correspondence to: Fred S. Gorelick MD, Research Building 27, VA Connecticut Healthcare, West Haven, CT 06516, USA. 1521–6918/99/020227 + 14 $12.00/00
© 1999, Baillière Tindall
228
F. S. Gorelick and T. Otani
activation takes place within the pancreas. These investigations have measured pancreatic enzyme activity and serum levels of markers for enzyme activation; most have used animal models of pancreatitis that generate oedematous pancreatitis. One of the most common uses high concentrations of cholecystokinin (CCK) or its analogue caerulein. In vivo, hyperstimulation (doses 10–100 times greater than those that elicit maximum secretion) by CCK causes suppressed secretion and pancreatitis in animal models (Gorelick et al, 1993). One of the early key features of hyperstimulation pancreatitis is decreased zymogen secretion into the pancreatic duct (Scheele et al, 1987); this is followed by pancreatic oedema, inflammation and necrosis. The caerulein-induced pancreatitis does not cause mortality, and there is usually complete recovery of function (Gorelick et al, 1993). Although this review focuses on zymogen activation within the acinar cell, other mechanisms undoubtedly contribute to pancreatic injury (Gorelick et al, 1993). Increased vascular permeability and endothelial injury are both features of acute pancreatitis and may contribute to pancreatic oedema, decreased pancreatic blood flow and ischaemic injury (Gress et al, 1990). Potentially damaging free radicals are detected within the pancreas soon after hyperstimulation (Niederan et al, 1992), and free radical scavengers have been reported to improve the course of hyperstimulation pancreatitis (Neuschwander-Tetri et al, 1992). Finally, cytokine release by both acinar cells and activated inflammatory cells, and the expression of vascular adhesion factors, promote further inflammation and injury (Gorelick et al, 1993; Norman, 1998). Advances in the study of acute pancreatitis have been aided by the development of new probes and techniques that detect zymogen activation using biochemical and immunocytochemical techniques. These approaches have been used to demonstrate that zymogen activation is an early feature of acute pancreatitis, that activation takes place within the pancreas, and, most recently, that zymogens can be activated within the acinar cell. EVIDENCE THAT ZYMOGENS ARE ACTIVATED WITHIN THE PANCREAS Many pancreatic digestive enzymes are synthesized and stored within the acinar cell as inactive zymogen precursors. Upon entering the small intestine, trypsinogen is converted to trypsin by the brush border hydrolase, enterokinase. Trypsin then activates the other pancreatic zymogens. Low levels of zymogen activation may take place within the acinar cell under physiological conditions, but protective mechanisms normally prevent cell damage. The protective mechanisms within the zymogen granules include the pancreatic trypsin inhibitor, pH conditions that are below the optimum for most enzymes, and proteases that can degrade activated enzymes. However, pathological activation of large amounts of trypsinogen and the resulting proteolytic cascade may proceed within the acinar cell and overwhelm protective mechanisms. The suggestion that zymogen activation plays a central role in the pathogenesis of pancreatitis is based upon the following observations: (a) trypsin and elastase activity increase in the pancreas early in the course of experimental pancreatitis (Bialek et al, 1991; Luthen et al, 1995), (b) the trypsinogen and carboxypeptidase A1 (PCA1) activation peptides, markers for zymogen activation are released into the serum early in the course of acute pancreatitis (Schmidt et al, 1992), (c) pre-treatment with gabexate mesilate, a serine protease inhibitor, reduces endoscopic retrograde
Mechanisms of intracellular zymogen activation
229
cholangiopancreatography (ERCP)-induced pancreatitis (Cavallini et al, 1996), (d) caerulein-hyperstimulation pancreatitis is ameliorated by serine protease inhibitors (Lasson and Ohlsson, 1984; Suzuki et al, 1992) and (e) a trypsinogen mutation that is likely to make trypsin resistant to degradation is associated with hereditary pancreatitis and (Whitcomb et al, 1996). These observations provide both indirect and direct evidence that zymogen activation is a critical feature of acute pancreatitis. Zymogen activation appears to occur very early in the course of acute pancreatitis. The trypsinogen and PCA1 activation peptides, markers for zymogen activation, are released into the serum within the first hours of acute pancreatitis (Gudgeon et al, 1990; Appelros et al, 1998). Similarly, pancreatic tryptic activity and levels of the trypsinogen activation peptide (TAP) increase soon (within 15 minutes) after the initiation of experimental hyperstimulation pancreatitis (Luthen et al, 1995; Mithofer et al, 1998). One study of experimental pancreatitis has observed that the generation of tryptic activity appears to be biphasic, with an early peak at about an hour followed by a second peak of activity several hours later (Luthen et al, 1995). This suggests that there may be more that one mechanism of zymogen activation. These studies demonstrate that zymogen activation takes place within the pancreas and at the onset of acute pancreatitis, but do not define the site of the activation. EVIDENCE THAT ZYMOGENS ARE ACTIVATED WITHIN THE ACINAR CELL Evidence that supports the hypothesis that zymogen activation takes place within the acinar cell comes from in vivo and in vitro experimental studies. In vivo hyperstimulation (a dose of secretagogue that is at least 10-fold greater than that required to generate a maximal secretory response) with CCK or its analogue, caerulein, results in oedematous pancreatitis. Following such hyperstimulation, there is an increase in tryptic activating in pancreatic fractions that are enriched in zymogen granules (Mithofer et al, 1998). However, the identity of the organelle associated with the tryptic activity has not been defined. Recent studies have used antibodies to TAP in immunofluorescence and immunoelectron microscopy studies to detect pancreatic sites involved in the processing of trypsinogen to trypsin (Otani et al, 1998). In untreated rat pancreas, there is no TAP immunoreactivity (Figure 1). This important control demonstrates that the activation peptide in trypsinogen is not accessible to the antibody. After exposure to a dose of caerulein that elicits maximal enzyme secretion (0.01 µg/kg/hour) little to no TAP immunoreactivity was appreciated. However, after hyperstimulation stimulation (5 µg/kg/hour) bright TAP immunoreactivity appeared in small vesicles that occupy the supranucler region of the acinar cell (Figure 1). Phase micrographs demonstrated that the immunoreactivity was at the base of the apical zymogen granule-rich region. Additional studies have shown that the number and size of TAP-positive structures increase up to 1 hour; after 120 minutes of stimulation, most TAP immunoreactivity is found in large vacuoles and throughout the cytoplasm of the acinar cell (Otani et al, 1998). To confirm that the appearance of TAP resulted from activating proteases, rats received the serine protease inhibitor FUT-175 prior to cerulein. This pre-treatment blocked the generation of TAP in the acinar cells (Figure 1). These findings demonstrate that TAP is generated within the acinar cell only under conditions that generate pancreatitis and that it requires protease activation. The morphology and distribution of the compartment suggests that it is not associated with zymogen granules.
230
F. S. Gorelick and T. Otani
Figure 1. The tryspinogen activation peptide, TAP, is generated in vivo. Following 60 minutes in vivo stimulation with saline (A,B), cerulein (0.01 µg/kg/hour) (CD) or hyperstimulation (5 µg/kg/hour) (E, F, G, H), rat pancreas was fixed and processed for immunocytochemistry using affinity-purified antibody to TAP as described (Otani et al, 1998). Little TAP immunoreactivity was observed in unstimulated pancreas (A,B) or after the lower concentration of caerulein that generates a maximal secretory response (C, D). However, bright immunoreactivity appeared in the supranuclear region (arrowheads) after the high concentration of cerulein that induces pancreatitis (E, F). Immunoreactivity was not detected when hyperstimulated tissues were labelled in the absence of the TAP antibody (G, H). See Otani et al (1998) for additional information.
Mechanisms of intracellular zymogen activation
231
In vitro studies that use isolated pancreatic acini have directly demonstrated that hyperstimulation can cause zymogen activation in acinar cells. Exposure of isolated acini to CCK stimulation results in the conversion of PCA1 to CA1 in a concentrationdependent manner (Figure 2 and Leach et al, 1991). Under the conditions of hyperstimulation (0.1 µM), the active form is found within acinar cell and not the medium (Grady et al, in press). This indicates that zymogen processing is an intracellular event. Similar results are observed when the effects of CCK hyperstimulation on trypsin activity and the generation of TAP are used as markers of zymogen activation (Grady et al, in press). Activation of muscarinic receptors on the acinar cell also stimulates PCA1 processing (Saluja et al, 1997). In the rat, both zymogen processing and amylase secretion are regulated by a telenzepine-sensitive muscarinic receptor (Schmid et al, in press). These findings demonstrate that activation of more than one type of G-proteinlinked receptor can stimulate zymogen activation.
Figure 2. CCK-stimulated pro-carboxypeptidase A1 conversion to carboxypeptidase A1 is concentrationdependent. Increasing concentrations of CCK octapeptide were added to isolated rat acini for 30 minutes and the total amount of carboxypeptidase A1 (CA1) assayed and expressed as percentage of the unstimulated control. Adapted from Leach et al (1991). * P < 0.05 compared to unstimulated control.
As in in vivo studies, in vitro CCK hyperstimulation of isolated pancreatic acini has demonstrated that both trypsinogen and PCA1 conversion begins within 15 minutes of stimulation (Figure 3; Leach et al, 1991; Saluja et al, 1997). The amount of the active form tends to reach a plateau after 30 to 45 minutes and decrease after an hour. The fall in the CA1 levels at the later period may result from degradation of the enzyme (Gorelick et al, 1992). The levels of the active enzyme or active forms generated in these systems are usually only a fraction (<5%) of the total activity. The findings suggest that either a limited zymogen pool is susceptible to activation or that factors supporting zymogen activation may be lost during the process of in vitro activation. It also indicates that activation of only a small portion of the zymogen pool may be sufficient to initiate pancreatitis. A second line of evidence that zymogens can be activated within the acinar cell comes from studies that have measured the generation of TAP or trypsin in isolated pancreatic acini. As in studies of CA1 CCK hyperstimulation results in the rapid generation of TAP and tryptic activity in acini (Saluja et al, 1997). These biochemical and immunocytochemical studies in vivo and in vitro suggest that zymogen processing can take place within the acinar cell early in the course of acute pancreatitis, but they do not exclude processing in other sites such as the interstitial space.
232
F. S. Gorelick and T. Otani
Figure 3. CCK-stimulated PCA1 conversion to CA1 is time-dependent. A hyperstimulating concentration of CCK octapeptide (0.1 µM) was added to isolated rat acini and the total amount of carboxypeptidase A1 (CA1) assayed and expressed as a percentage of the unstimulated control. Adapted from Leach et al (1991). P < 0.05 compared to unstimulated control.
EFFECT OF ETHANOL ON ZYMOGEN PROCESSING Ethanol is one of the most frequent causes of acute and chronic pancreatitis. Ethanol does not itself cause pancreatitis in rats, but it has been reported to worsen caeruleinhyperstimulation pancreatitis, suggesting that the two agents may act synergistically (Quon et al, 1992). Administration of specific CCK antagonists have been reported to reduce pancreatic injury induced by diet (Saluja et al, 1989), bile salts (Bilechik et al, 1990), and ischaemia (Modlin et al, 1988). These studies suggest that CCK receptor stimulation may contribute to injury in many forms of pancreatitis. The ability of ethanol to sensitize the acinar cell to the effects of CCK has been examined in isolated pancreatic acini (Katz et al, 1996). Studies that examine the pathological effects of ethanol should employ levels that are pathologically relevant. Thus, the effects of levels of greater than 50 mM ethanol may not be pertinent to human disease. Although most in vivo studies of ethanol have not demonstrated a pathological effect on the pancreas, it is not clear that pathological levels were reached in these models. That only a small number of humans who abuse ethanol develop pancreatitis suggests that other genetic and environmental factors play a role in the development of alcoholic pancreatitis. Ethanol has been found to sensitize G-protein-linked receptors to agonists. Although a low concentration of ethanol (5 mM) has no effect on cAMP generation, it dramatically enhances beta-agonist-dependent stimulation of cAMP (Nagy et al, 1992). A similar pattern of sensitization has been observed with ethanol and cholecystokinin. While ethanol (25 mM) has no effect on PCA1 processing in isolated rat acini, it does sensitize the cells to CCK (Figure 4). This effect was dependent on the duration of ethanol treatment and its concentration. Thus, the addition of ethanol to physiological concentrations of CCK (0.1 nM) had effects on pro-carboxypeptidase processing similar to those of CCK hyperstimulation (0.1 µM). As shown in Figure 5, the addition of the CCK receptor inhibitor L364,718 blocked the enhanced conversion (Katz et al, 1996). Although ethanol sensitized the acinar cell to the effects of CCK, it had little effect on carbachol-stimulated PCA1 processing. Thus, ethanol appears to sensitize the acinar cell to selective agents. Because ethanol stimulates zymogen processing when CCK levels are within the physiological range, these observations may be relevant to human disease.
Mechanisms of intracellular zymogen activation
233
Figure 4. Ethanol sensitizes the acinar cell to the effects of CCK on PCA1 processing. (A) Effect of adding increasing concentrations of ethanol with a low dose of CCK (0.1 nM) on acini treated for 30 minutes (right) compared to a concentration of CCK (0.1 µM) (left) that generates maximal conversion (mean ±SEM). (B) The time course of CA1 generation by acini after adding ethanol (25 mM) 5, 10 and 30 minutes before terminating 30 minutes treatment with 0.1 nM CCK. For the 45 minute time point, ethanol was added 15 minutes prior to the addition of CCK. *P <0.001 versus CCK 0.1 nM. Adapted from Katz et al (1996). US = unstimulated.
THEORETICAL MECHANISMS OF ZYMOGEN ACTIVATION At least seven pathways have the potential to be involved in the intracellular conversion of pancreatic zymogens to active (mature) enzymes: (1) trypsinogen autoactivation to trypsin (Figarella et al, 1988), (2) lysosomal hydrolase cathepsin B cleavage of trypsinogen to trypsin (Greenbaum and Hirshkowitz, 1961; Figarella et al, 1988), (3) enhanced susceptibility of zymogens to proteolysis because of oxidation or decondensation (Pacifici et al, 1993), (4) leakage of zymogens and lysosomal enzymes into the cytoplasm and subsequent proteolytic activation (Wilson et al, 1992), (5) shunting of zymogens into membrane-bound compartments that contain active proteases (Steer, 1988; Gorelick and Matovcik, 1995), (6) uptake and processing of secreted zymogens by endocytic pathways (Lerch et al, 1995) and (7) diminished activity of the intracellular pancreatic trypsin inhibitor. The mechanisms that have received the most attention are trypsinogen auto-activation and cathepsin B activation of trypsinogen. Trypsinogen auto-activation requires an acidic pH and is enhanced in the presence of Ca2+ (Kassell and Kay, 1973; Figarella et al, 1988). The affinity of pancreatic trypsin inhibitor is greatest at a neutral pH and is reduced at an acidic pH. Thus, conditions favouring the intracellular formation of active trypsin require an acidic pH. The generation of low-pH compartments within the acinar cell during experimental pancreatitis may be important to trypsinogen auto-activation and the expression of tryptic activity (Niederau and Grendell, 1988). Cathepsin B activation of trypsinogen in vitro has been reported to proceed optimally at an acidic pH (Figarella et al, 1988) or at a neutral pH (Lerch et al, 1993a). Under basal conditions cathepsin B is found within zymogen granules (reviewed by Gorelick and Matovcik, 1995). During the early stages (within 1–2 hours) of induction of some forms of experimental pancreatitis, increased amounts of lysosomal
234
F. S. Gorelick and T. Otani
Figure 5. Ethanol stimulation of PCA1 processing requires CCK receptor occupancy. Acini were exposed to CCK in physiological (0.1 nM) or hyperstimulating (0.1 µM) concentrations for 30 minutes in the presence or absence of ethanol. Some acini were pre-treated for 10 minutes with the CCK receptor antagonist, L364,718. The generation of CA1 is expressed as per control. *P <0.001 versus paired value without L364,718. Adapted from Katz et al (1996). US = unstimulated.
hydrolases are found in zymogen-containing compartments (Saluja et al, 1987; Lerch et al, 1993b). STUDIES RELATING TO THE MECHANISM AND SITE OF PROCESSING Processing takes place in a secretory compartment As with CCK or carbachol, bombesin stimulates secretion from the acinar cell. However, in contrast to the biphasic secretory response to CCK or carbachol, acini stimulated by bombesin exhibit a monophasic response (Figure 6A). These two patterns of secretory responsiveness correlate with the effects of these two agents on the apical actin cytoskeleton. An intact actin cytoskeleton is required for secretion from the pancreatic acinar cell (Muallem et al, 1995). High concentrations of CCK result in disruption of the apical actin cytoskeleton of the acinar cell, but high concentrations of bombesin do not disrupt the actin cytoskeleton (O’Konski and Pandol, 1993). Exposure of acini to each of these agents results in enhanced zymogen conversion (Figure 6B; Grady et al, 1998). However, in contrast to CCK or carbachol, the active enzyme forms generated by bombesin treatment are released into the medium (Figure 7). Even after bombesin stimulation, approximately 20% of CA1 is associated with the acinar pellet (not shown). Because bombesin-treated cells demonstrated intact morphology by transmission electron microscopy, excluded trypan blue, and did not release LDH, it is unlikely that leakage of active enzyme from cells accounts for these findings (Grady et al, 1998). There are two important implications of the bombesin studies. First, they suggest that proteolytic conversion of PCA1 to mature CA1 takes place within the secretory
Mechanisms of intracellular zymogen activation
235
Figure 6. Effects of CCK (䊐) and bombesin (䉬) on amylase secretion and PCA1 processing. Isolated pancreatic acini were stimulated with increasing concentrations of CCK or bombesin for 30 minutes to measure secretion and for 45 minutes to measure conversion. (A) Amylase secretion into the medium. (B) Generation of CA1 expressed as a percentage of the control. Adapted from Grady et al (1998). Acini
Medium PCA1 CA1
A
B
C
D
Figure 7. CA1 generated by CCK hyperstimulation is not secreted. Isolated pancreatic acini were hyperstimulated using cholecystokinin (CCK) (0.1 µM) (A, C), or treated with saline (B) or bombesin (0.1 µM) (D) for 30 minutes. The medium and acini were separated, and proteins were subjected to SDS–PAGE and processed to immunoblot analysis. The upper band is pro-carboxypeptidase A1 (PCA1) and the lower band is mature carboxypeptidase A1 (CA1). Both CCK (A) and bombesin (not shown) stimulate the generation of CA1. In the case of CCK, CA1 is not released into the medium (C), but after bombesin (D), CA1 appears in the medium. Adapted from Grady et al (1998).
pathway. Second, they suggest that the signals for zymogen conversion (elicited with both CCK and bombesin) may be different from those that lead to an inhibition of enzyme secretion (observed only with CCK). These observations may have physiological relevance and correlate with the effects of high concentrations of CCK or bombesin on the pancreas in vivo. While CCK hyperstimulation causes pancreatitis, bombesin does not. A conclusion that follows from these observations is that both zymogen conversion and the retention of enzymes within the acinar cell are required to initiate pancreatitis. Dependence of processing on protease activation Proteases differ in their mechanism of activation. This fact can be used to help define the mechanism of zymogen activation in the pancreatic acinar cell. Trypsin belongs to the serine protease family and cathepsin B is a thiol protease. Protease inhibitors can selectively block the activity of members of each family. The addition of soybean
236
F. S. Gorelick and T. Otani
trypsin inhibitor, a high-molecular-weight serine protease inhibitor that would not be expected to enter the acinar cell, has no effect on PCA1 processing (Leach et al, 1991). This finding suggests that zymogen processing does not take place in the incubation medium. The low-molecular-weight thiol protease inhibitor E-64 enters cells and is a potent inhibitor of cathepsin B. Although addition of this inhibitor to acinar cells blocked virtually all cathepsin B activity, it did not affect the processing of PCA1 associated with CCK hyperstimulation (Figure 8). In contrast, benzamidine, a permeant serine protease inhibitor, reduced the processing of PCA1 to basal levels. These findings suggest that activation of a serine protease within the acinar cell is required to stimulate PCA1 processing. Similar to its affect on PCA1 caerulein hyperstimulation of isolated pancreatic acini results in increased trypsinogen processing assayed using tryptic activity and the generation of TAP (Klonowski-Stump et al, 1998). The serine protease inhibitor FUT-175 blocked the activation but not the thiol protease inhibitors NCO-700 or E-64. Notably, these inhibitors blocked >95% of cathepsin B activity. Another study has shown that a high concentration of a more permeable form of E-64, E-64d, does block the generation of tryptic activity in cerulein-stimulated acini (Saluja et al, 1997). Finally, preliminary studies have suggested that inhibition of thiol proteases may result in increased levels of CA1 in hyperstimulated acini (Gorelick et al, 1992). The last observation suggests that thiol proteases may degrade activated enzymes. As discussed elsewhere (Gorelick and Matovcik, 1995), additional studies are required to resolve the differences between the results of these investigations.
Figure 8. The processing of PCA1 is dependent on activation of serine proteases. Acini were pre-treated for 30 minutes with either the serine protease inhibitor benzamidine (10 mM) or the thiol protease inhibitor, E-64 (0.1 mM) prior to hyperstimulation by CCK (0.1 µM). *P <0.05 versus CCK 0.1 µM. Adapted from Leach et al (1991).
Dependence on a low-pH compartment To examine the role of intracellular pH on zymogen processing, the pH of acidic intracellular compartments has been increased using the weak base chloroquine and the proton inophore monesin (Leach et al, 1991). Using concentrations of these agents that did not affect the secretory response to cholecystokinin, their effects on CCKinduced PCA1 processing was examined. As shown in Figure 9, both agents dramatically reduced the processing of this zymogen. The effects of these agents on the
Mechanisms of intracellular zymogen activation
237
Figure 9. The processing of PCA1 is dependent on a low pH compartment. Acini were pre-treated for 30 minutes with either monensin (10 µM) or chloroquine (40 µM) to increase intracellular pH and then hyperstimulated by CCK. *P < 0.05 versus CCK 0.1 µM. Adapted from Leach et al (1991).
generation of tryptic activity or TAP have not been reported. These findings suggest that a low pH compartment play a role in zymogen processing in the pancreatic acinar cell. The pH of compartments in the secretory pathway has been estimated using a weak base and an immunoelectron microscopy technique (Orci et al, 1987). In that study the Golgi complex was mildly acidic (pH 6.8) and condensing vacuoles were more acidic (pH 6.5). Zymogen granules were less acidic (pH ~ 6.9). Although the effect of hyperstimulation on pH in these compartments has not been examined, it is possible that a decrease in the pH initiates processing. Alternatively, the movement of nascent proteins through these compartments may be delayed during hyperstimulation so that they are exposed to a lower pH for a longer period. In either event, a low pH appears to play a central role in the processing of PCA1 to its active form. The importance of a low pH environment in the processing of other zymogens needs to be examined. Trypsinogen processing takes place in a distinct cellular compartment To examine the compartments associated with zymogen processing, double labelling experiments were performed using antibodies to TAP and antibodies to GRAMP-92 (granule membrane protein-92) an integral membrane protein and marker of lysosomes and recycling endosomes and in vivo caerulein hyperstimulation (Laurie et al, 1992; Otani et al, in press). High-magnification immunofluorescence studies demonstrated that TAP immunoreactivity was localized to structures just below the zymogen granule-enriched region (Figure 10). Localization of TAP immunoreactivity using immunogold techniques and electron microscopy has confirmed that TAP is generated in irregular vesicular structures and not in zymogen granules (Otani et al, 1998). Double labelling studies demonstrated that most TAP immunoreactivity is confined to GRAMP-92-positive structures. These findings suggest that zymogen processing takes place in a compartment that has overlap with endosomal or lysosomal markers. It is likely that this is also the compartment responsible for the processing of PCA1 to CA1. Together with data demonstrating that bombesin stimulates the processing of PCA1 in a secretory compartment (see Grady et al, in press and Figure 7), the findings suggest that processing might proceed in a secretory compartment that has overlap with
238
F. S. Gorelick and T. Otani
Figure 10. TAP is generated in a distinct subcellular compartment. After 45 minutes of in vivo hyperstimulation with caerulein, the pancreas was fixed and processed for phase microscopy (A), TAP (B) and GRAMP-92 (C) immunoreactivity. TAP appeared in a rounded structure just below the zymogen granule compartment (arrowheads). TAP immunoreactivity was contained in a GRAMP-92-positive structure.
endosomes or lysosomes. Recent studies have described a regulated secretion from lysosomes in several types of cell, including cultured fibroblasts and epithelial cells (Rodriguez et al, 1997). Whether zymogen processing in the acinar cell takes place in a similar compartment remains unclear. SUMMARY Zymogen processing can take place within the pancreatic acinar cell. This processing can be stimulated by supraphysiological concentrations of naturally occurring ligands. In vivo and in vitro studies demonstrate that processing is enhanced soon (15 minutes) after hyperstimulation with CCK or caerulein. Ethanol sensitizes the acinar cell to the effects of CCK; physiological concentrations of CCK cause enhanced zymogen processing in the presence of physiologically attainable ethanol levels. Zymogen processing is dependent on activation of serine proteases and a low pH compartment. However, the role of thiol proteases, including cathepsin B, remains unclear. Zymogen processing takes place within the acinar cell and not in zymogen granules. The processing may take place in a distinct secretory compartment. Finally, zymogen activation alone may not be sufficient to cause acinar cell injury and pancreatitis. Disruption of the secretory mechanisms and the apical actin cytoskeleton may cause the active enzymes to be retained in the acinar cell and be required for the initiation of pancreatitis. REFERENCES Appelros S, Thim L & Borgstorm A (1998) Activation peptide of carboxypeptidase B in serum and urine in acute pancreatitis. Gut 42: 97–102. Bialek R, Willemer S, Arnold R & Adler G (1991) Evidence of intracellular activation of serine proteases in acute cerulein-induced pancreatitis in rats. Scandinavian Journal of Gastroenterology 26: 190–196.
Mechanisms of intracellular zymogen activation
239
Bilchik AJ, Zucker KA, Adrian TE & Modlin IM (1990) Amelioration of cholinergic-induced pancreatitis with a selective cholecystokinin receptor antagonist. Archives of Surgery 125: 1546–1549. Cavallini G, Tittobello A, Frulloni E et al (1996) Gabexate for the prevention of pancreatic damage related to endoscopic retrograde cholaniopancreatography. New England Journal of Medicine 335: 919–923. Figarella C, Miszczuk-Jamska B & Barret AJ (1988) Possible lysosomal activation of pancreatic zymogens: activation of both human trypsinogens by cathepsin B and spontaneous acid activation of human trypsinogen. Biological Chemistry Hopp-Seyler 369: 293–298. Gorelick F & Matovcik L (1995) Lysosomal enzymes and pancreatitis. Gastroenterology 109: 620–625. Gorelick FS, Adler G & Kern HF (1993) Cerulein-induced pancreatitis. In Go VLW et al (eds) The Pancreas: Biology, Pathobiology, and Disease, pp 501–526. New York: Raven Press. Gorelick FS, Modlin IM, Leach SD et al (1992) Intracellular proteolysis of pancreatic zymogens. Yale Journal of Biology and Medicine 65: 407–420. *Grady T, Mah’moud M, Otani T et al (1998) Zymogen proteolysis within the pancreatic acinar cell is associated with cellular injury. American Journal of Physiology 275: G1010–G1017. Greenbaum LA & Hirshkowitz A (1961) Endogenous cathepsin activaties trypsinogen in extracts of dog pancreas. Proceedings of the Society for Experimental Biology and Medicine 107: 74–76. Gress TM, Arnold R & Adler G (1990) Structural alterations of pancreatic microvasculature in ceruleininduced pancreatitis in the rat. Research in Experimental Medicine 190: 401–412. Gudgeon AM, Heath DI, Hurley P et al (1990) Trypsinogen activation peptides assay in the early prediction of severity of acute pancreatitis. Lancet 335: 4–8. Kassell B & Kay J (1973) Zymogens of proteolytic enzymes. Science 180: 1022–1027. *Katz M, Carangelo R, Miller LJ & Gorelick F (1996) Effect of ethanol on cholecystokinin-stimulated zymogen conversion in pancreatic acinar cells. American Journal of Physiology 270: G171–G175. Klonowski-Stump H, Lüthen R, Han B et al (1998) Inhibition of cathepsin B does not affect the intracellular activation of trypsinogen by cerulein hyperstimulation in isolated rat pancreatic acinar cells. Pancreas 16: 96–101. Lasson A & Ohlsson K (1984) Protease inhibitors in acute pancreatitis: correlation between biochemical changes and clinical course. Scandinavian Journal of Gastroenterology 19: 779–786. Laurie SM, Mixon MB, Brand SH & Castle JD (1992) A secretion granule membrane protein (GRAMP 92) is found in non-granule membranes including those of the endocytic pathway. European Journal of Cell Biology 58: 12–27. Leach SD, Modlin IM, Scheele GA & Gorelick FS (1991) Intracellular activation of digestive zymogens in rat pancreatic acini. Stimulation by high doses of cholecystokinin. Journal of Clinical Investigation 87: 362–366. Lerch MM, Saluja AK, Dawra R et al (1993a) The effect of chloroquine administration on two experimental models of acute pancreatitis. Gastroenterology 104: 1768–1779. Lerch MM, Saluja AK, Runzi M et al (1993b) Pancreatic duct obstruction triggers acute necrotizing pancreatitis in the opossum [see comments]. Gastroenterology 104: 853–861. Lerch MM, Saluja AK, Runzi M et al (1995) Luminal endocytosis and intracellular targeting by acinar cells during early biliary pancreatitis in the opossum. Journal of Clinical Investigation 95: 2222–2238. Luthen R, Niederau C & Grendell JH (1995) Intrapancreatic zymogen activation and levels of ATP and glutathione during caerulein pancreatitis in rats. American Journal of Physiology 268: G592–G604. *Mithofer K, Fernandex-del Castillo C, Rattner D & Warshaw, AL (1998) Subcellular kinetic of early typsinogen activation in acute rodent pancreatitis. American Journal of Physiology 274: G71–G79. Modlin I, Bilchik A, Zucker K et al (1988) Cholecystokinin augmentation of ‘surgical’ pancreatitis. Archives of Surgery 124: 574–578. Muallem S, Kwiatkowska K, Xu X & Yin HL (1995) Actin filament disassembly is a sufficient final trigger for exocytosis in nonexcitable cells. Journal of Cell Biology 128: 589–598. Nagy LE & De Silva SE (1992) Ethanol increases receptor-dependent cyclic AMP production in cultured hepatocytes by decreasing G(i)-mediated inhibition. Biochemical Journal 286: 681–686. Neuschwander-Tetri BA, Ferrell LD, Sukhabote RJ & Grendell JH (1992) Glutathione monoethyl ester ameliorates caerulein-induced pancreatitis in the mouse. Journal of Clinical Investigation 89: 109–116. Niederau C & Grendell JH (1988) Intracellular vacuoles in experimental acute pancreatitis in rats and mice are an acidified compartment. Journal of Clinical Investigation 81: 229–236. Niederau C, Niederau M, Borchard F et al (1992) Effects of antioxidants and free radical scavengers in three different models of acute pancreatitis. Pancreas 7: 486–496. *Norman J (1998) The role of cytokines in the pathogenesis of acute pancreatitis. American Journal of Surgery 175: 76–83. O’Konski MS & Pandol SJ (1993) Cholecystokinin JMV-180 and caerulein effects on the pancreatic acinar cell cytoskeleton. Pancreas 8: 638–646. Orci L, Ravazzola M & Anderson R (1987) The condensing vacuole of exocrine cells is more acidic than the mature secretory vesicle. Nature 326: 77–79.
240
F. S. Gorelick and T. Otani
*Otani TS, Chepilko M, Grendell JH & Gorelick FS (in press) Co-distribution of trypsinogen activation peptide and the granule membrane protein, GRAMP-92, in rat cerulein-induced pancreatitis. American Journal of Physiology 275: G999–G1009. Pacifici RE, Kono Y & Davies KJA (1993) Hydrophobicity as the signal for selective degradation of hydroxyl radical-modified hemoglobin by the multicatalytic proteinase complex, proteosome. Journal of Biological Chemistry 268: 15 405–15 411. Quon MG, Kugelmas M, Wisner JR et al (1992) Chronic alcohol consumption intensifies caerulein-induced acute pancreatitis in the rat. International Journal of Pancreatology 12: 31–39. Rodriguez A, Webster P, Ortego J & Andrews NW (1997) Lysosomes behave as Ca2+-regulated exocytic vesicles in fibroblasts and epithelial cells. Journal of Cell Biology 137: 93–104. Saluja A, Sadamitsu H, Saluja M et al (1987) Subcellular redistribution of lysosomal enzymes during caeruleininduced pancreatitis. American Journal of Physiology 253: G508–G516. Saluja AK, Saluja M, Printz H et al (1989) Experimental pancreatitis is mediated by low-affinity cholecystokinin receptors that inhibit digestive enzyme secretion. Proceedings of the National Academy of Sciences of the USA 86: 8968–8971. Saluja AK, Donovan EA, Yamanaka K et al (1997) Cerulein-induced in vivo activation of trypsinogen in rat pancreatic acini is mediated by cathepsin B. Gastroenterology 113: 304–311. Scheele G, Adler G & Kern H (1987) Exocytosis occurs at the lateral plasma membrane of the pancreatic acinar cell during supramaximal secretagogue stimulation. Gastroenterology 92: 345–353. Schmid S, Modlin IM, Stoch A et al (in press) Telenzepine-sensitive muscarinic receptors on the rat pancreatic acinar cell. American Journal of Physiology Schmidt J, Fernandez-del Castillo C, Rattner DW et al (1992) Trypsinogen-activation peptides in experimental rat pancreatitis: prognostic implications and histopathologic correlates. Gastroenterology 103: 1009–1016. Steer M (1988) Pathogenesis of acute pancreatitis. Annual Review of Medicine 39: 95–105. Suzuki M, Isaji S, Stanten R et al (1992) Effect of protease inhibitor FUT-175 on acute hemorrhagic pancreatitis in mice. International Journal of Pancreatology 11: 59–65. Whitcomb DC, Gorry MC, Preston RA et al (1996) Hereditary pancreatitis is caused by a mutation on the cationic trypsinogen gene. Nature Genetics 14: 141–145. Wilson JS, Apte MV, Thomas MC et al (1992) Effects of ethanol, acetaldehyde and cholesteryl esters on pancreatic lysosomes. Gut 33: 1099–1104.
Baillière’s Clinical Gastroenterology Vol. 13, No. 2, pp 227–240, 1999
B A I L L I È R E ’ S
2
PRACTICE & RESEARCH
Mechanisms of intracellular zymogen activation Fred S. Gorelick*
MD
Professor of Medicine and Cell Biology Department of Medicine and Cell Biology, VA Connecticut Healthcare System, Yale University School of Medicine, West Haven, Connecticut 06516, USA
Taiichi Otani
MD, PhD
Assistant Professor Department of Hepatobiliary and Pancreatic Surgery, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-0033, Japan
The pancreatic acinar cell is potentially the initial site of injury that begins the series of events leading to acute pancreatitis. Pathological intrapancreatic zymogen activation occurs in experimental pancreatitis in animals and in human pancreatitis. Intracellular activation has been clearly linked to aberrant zymogen processing in one form of hereditary pancreatitis; in this genetic disease a mutation in cationic trypsinogen may eliminate the degradation of any trypsin activated in the acinar cell. Recent studies have also provided the first direct evidence that trypsinogen activation takes place early in the course of caerulein-induced pancreatitis; parallel studies have used isolated pancreatic acini and conditions that simulate those that cause pancreatitis in vivo to demonstrate that zymogens can be pathologically activated in isolated cells. A unique acinar cell pathway regulates the intracellular proteinase processing of zymogens to their active forms. Stimulating the acinar cell with supramaximal concentrations of cholecystokinin (CCK) or carbamylcholine can activate this pathway. The activation depends on a low pH compartment within the acinar cell and activation of an intracellular serine protease. A marker of trypsinogen processing, the trypsinogen activation peptide (TAP), is generated in acinar cell compartments that do not overlap with secretory granules. This compartment overlaps with a marker of recycling endosomes and lysosomes. Thus, zymogen processing within the acinar cell proceeds in a distinct subcellular compartment and is dependent on a low pH environment and activation of serine proteases. Key words: PCA1; Trypsinogen; TAP; CCK; low pH; serine protease.
The generation of pancreatitis involves activating a cascade of pathological events. One of the critical initiating events appears to be the intracellular activation of pancreatic zymogens. The pancreas has developed a number of mechanisms that are designed to prevent zymogen activation within the acinar cell or quench the activity of small amounts of active enzymes. However, in pathological states these protective mechanisms may be overwhelmed and pancreatitis is initiated. Although investigators have long suggested that zymogen activation within the pancreas may initiate pancreatitis, only recently have studies demonstrated that such * Address correspondence to: Fred S. Gorelick MD, Research Building 27, VA Connecticut Healthcare, West Haven, CT 06516, USA. 1521–6918/99/020227 + 14 $12.00/00
© 1999, Baillière Tindall
228
F. S. Gorelick and T. Otani
activation takes place within the pancreas. These investigations have measured pancreatic enzyme activity and serum levels of markers for enzyme activation; most have used animal models of pancreatitis that generate oedematous pancreatitis. One of the most common uses high concentrations of cholecystokinin (CCK) or its analogue caerulein. In vivo, hyperstimulation (doses 10–100 times greater than those that elicit maximum secretion) by CCK causes suppressed secretion and pancreatitis in animal models (Gorelick et al, 1993). One of the early key features of hyperstimulation pancreatitis is decreased zymogen secretion into the pancreatic duct (Scheele et al, 1987); this is followed by pancreatic oedema, inflammation and necrosis. The caerulein-induced pancreatitis does not cause mortality, and there is usually complete recovery of function (Gorelick et al, 1993). Although this review focuses on zymogen activation within the acinar cell, other mechanisms undoubtedly contribute to pancreatic injury (Gorelick et al, 1993). Increased vascular permeability and endothelial injury are both features of acute pancreatitis and may contribute to pancreatic oedema, decreased pancreatic blood flow and ischaemic injury (Gress et al, 1990). Potentially damaging free radicals are detected within the pancreas soon after hyperstimulation (Niederan et al, 1992), and free radical scavengers have been reported to improve the course of hyperstimulation pancreatitis (Neuschwander-Tetri et al, 1992). Finally, cytokine release by both acinar cells and activated inflammatory cells, and the expression of vascular adhesion factors, promote further inflammation and injury (Gorelick et al, 1993; Norman, 1998). Advances in the study of acute pancreatitis have been aided by the development of new probes and techniques that detect zymogen activation using biochemical and immunocytochemical techniques. These approaches have been used to demonstrate that zymogen activation is an early feature of acute pancreatitis, that activation takes place within the pancreas, and, most recently, that zymogens can be activated within the acinar cell. EVIDENCE THAT ZYMOGENS ARE ACTIVATED WITHIN THE PANCREAS Many pancreatic digestive enzymes are synthesized and stored within the acinar cell as inactive zymogen precursors. Upon entering the small intestine, trypsinogen is converted to trypsin by the brush border hydrolase, enterokinase. Trypsin then activates the other pancreatic zymogens. Low levels of zymogen activation may take place within the acinar cell under physiological conditions, but protective mechanisms normally prevent cell damage. The protective mechanisms within the zymogen granules include the pancreatic trypsin inhibitor, pH conditions that are below the optimum for most enzymes, and proteases that can degrade activated enzymes. However, pathological activation of large amounts of trypsinogen and the resulting proteolytic cascade may proceed within the acinar cell and overwhelm protective mechanisms. The suggestion that zymogen activation plays a central role in the pathogenesis of pancreatitis is based upon the following observations: (a) trypsin and elastase activity increase in the pancreas early in the course of experimental pancreatitis (Bialek et al, 1991; Luthen et al, 1995), (b) the trypsinogen and carboxypeptidase A1 (PCA1) activation peptides, markers for zymogen activation are released into the serum early in the course of acute pancreatitis (Schmidt et al, 1992), (c) pre-treatment with gabexate mesilate, a serine protease inhibitor, reduces endoscopic retrograde
Mechanisms of intracellular zymogen activation
229
cholangiopancreatography (ERCP)-induced pancreatitis (Cavallini et al, 1996), (d) caerulein-hyperstimulation pancreatitis is ameliorated by serine protease inhibitors (Lasson and Ohlsson, 1984; Suzuki et al, 1992) and (e) a trypsinogen mutation that is likely to make trypsin resistant to degradation is associated with hereditary pancreatitis and (Whitcomb et al, 1996). These observations provide both indirect and direct evidence that zymogen activation is a critical feature of acute pancreatitis. Zymogen activation appears to occur very early in the course of acute pancreatitis. The trypsinogen and PCA1 activation peptides, markers for zymogen activation, are released into the serum within the first hours of acute pancreatitis (Gudgeon et al, 1990; Appelros et al, 1998). Similarly, pancreatic tryptic activity and levels of the trypsinogen activation peptide (TAP) increase soon (within 15 minutes) after the initiation of experimental hyperstimulation pancreatitis (Luthen et al, 1995; Mithofer et al, 1998). One study of experimental pancreatitis has observed that the generation of tryptic activity appears to be biphasic, with an early peak at about an hour followed by a second peak of activity several hours later (Luthen et al, 1995). This suggests that there may be more that one mechanism of zymogen activation. These studies demonstrate that zymogen activation takes place within the pancreas and at the onset of acute pancreatitis, but do not define the site of the activation. EVIDENCE THAT ZYMOGENS ARE ACTIVATED WITHIN THE ACINAR CELL Evidence that supports the hypothesis that zymogen activation takes place within the acinar cell comes from in vivo and in vitro experimental studies. In vivo hyperstimulation (a dose of secretagogue that is at least 10-fold greater than that required to generate a maximal secretory response) with CCK or its analogue, caerulein, results in oedematous pancreatitis. Following such hyperstimulation, there is an increase in tryptic activating in pancreatic fractions that are enriched in zymogen granules (Mithofer et al, 1998). However, the identity of the organelle associated with the tryptic activity has not been defined. Recent studies have used antibodies to TAP in immunofluorescence and immunoelectron microscopy studies to detect pancreatic sites involved in the processing of trypsinogen to trypsin (Otani et al, 1998). In untreated rat pancreas, there is no TAP immunoreactivity (Figure 1). This important control demonstrates that the activation peptide in trypsinogen is not accessible to the antibody. After exposure to a dose of caerulein that elicits maximal enzyme secretion (0.01 µg/kg/hour) little to no TAP immunoreactivity was appreciated. However, after hyperstimulation stimulation (5 µg/kg/hour) bright TAP immunoreactivity appeared in small vesicles that occupy the supranucler region of the acinar cell (Figure 1). Phase micrographs demonstrated that the immunoreactivity was at the base of the apical zymogen granule-rich region. Additional studies have shown that the number and size of TAP-positive structures increase up to 1 hour; after 120 minutes of stimulation, most TAP immunoreactivity is found in large vacuoles and throughout the cytoplasm of the acinar cell (Otani et al, 1998). To confirm that the appearance of TAP resulted from activating proteases, rats received the serine protease inhibitor FUT-175 prior to cerulein. This pre-treatment blocked the generation of TAP in the acinar cells (Figure 1). These findings demonstrate that TAP is generated within the acinar cell only under conditions that generate pancreatitis and that it requires protease activation. The morphology and distribution of the compartment suggests that it is not associated with zymogen granules.
230
F. S. Gorelick and T. Otani
Figure 1. The tryspinogen activation peptide, TAP, is generated in vivo. Following 60 minutes in vivo stimulation with saline (A,B), cerulein (0.01 µg/kg/hour) (CD) or hyperstimulation (5 µg/kg/hour) (E, F, G, H), rat pancreas was fixed and processed for immunocytochemistry using affinity-purified antibody to TAP as described (Otani et al, 1998). Little TAP immunoreactivity was observed in unstimulated pancreas (A,B) or after the lower concentration of caerulein that generates a maximal secretory response (C, D). However, bright immunoreactivity appeared in the supranuclear region (arrowheads) after the high concentration of cerulein that induces pancreatitis (E, F). Immunoreactivity was not detected when hyperstimulated tissues were labelled in the absence of the TAP antibody (G, H). See Otani et al (1998) for additional information.
Mechanisms of intracellular zymogen activation
231
In vitro studies that use isolated pancreatic acini have directly demonstrated that hyperstimulation can cause zymogen activation in acinar cells. Exposure of isolated acini to CCK stimulation results in the conversion of PCA1 to CA1 in a concentrationdependent manner (Figure 2 and Leach et al, 1991). Under the conditions of hyperstimulation (0.1 µM), the active form is found within acinar cell and not the medium (Grady et al, in press). This indicates that zymogen processing is an intracellular event. Similar results are observed when the effects of CCK hyperstimulation on trypsin activity and the generation of TAP are used as markers of zymogen activation (Grady et al, in press). Activation of muscarinic receptors on the acinar cell also stimulates PCA1 processing (Saluja et al, 1997). In the rat, both zymogen processing and amylase secretion are regulated by a telenzepine-sensitive muscarinic receptor (Schmid et al, in press). These findings demonstrate that activation of more than one type of G-proteinlinked receptor can stimulate zymogen activation.
Figure 2. CCK-stimulated pro-carboxypeptidase A1 conversion to carboxypeptidase A1 is concentrationdependent. Increasing concentrations of CCK octapeptide were added to isolated rat acini for 30 minutes and the total amount of carboxypeptidase A1 (CA1) assayed and expressed as percentage of the unstimulated control. Adapted from Leach et al (1991). * P < 0.05 compared to unstimulated control.
As in in vivo studies, in vitro CCK hyperstimulation of isolated pancreatic acini has demonstrated that both trypsinogen and PCA1 conversion begins within 15 minutes of stimulation (Figure 3; Leach et al, 1991; Saluja et al, 1997). The amount of the active form tends to reach a plateau after 30 to 45 minutes and decrease after an hour. The fall in the CA1 levels at the later period may result from degradation of the enzyme (Gorelick et al, 1992). The levels of the active enzyme or active forms generated in these systems are usually only a fraction (<5%) of the total activity. The findings suggest that either a limited zymogen pool is susceptible to activation or that factors supporting zymogen activation may be lost during the process of in vitro activation. It also indicates that activation of only a small portion of the zymogen pool may be sufficient to initiate pancreatitis. A second line of evidence that zymogens can be activated within the acinar cell comes from studies that have measured the generation of TAP or trypsin in isolated pancreatic acini. As in studies of CA1 CCK hyperstimulation results in the rapid generation of TAP and tryptic activity in acini (Saluja et al, 1997). These biochemical and immunocytochemical studies in vivo and in vitro suggest that zymogen processing can take place within the acinar cell early in the course of acute pancreatitis, but they do not exclude processing in other sites such as the interstitial space.
232
F. S. Gorelick and T. Otani
Figure 3. CCK-stimulated PCA1 conversion to CA1 is time-dependent. A hyperstimulating concentration of CCK octapeptide (0.1 µM) was added to isolated rat acini and the total amount of carboxypeptidase A1 (CA1) assayed and expressed as a percentage of the unstimulated control. Adapted from Leach et al (1991). P < 0.05 compared to unstimulated control.
EFFECT OF ETHANOL ON ZYMOGEN PROCESSING Ethanol is one of the most frequent causes of acute and chronic pancreatitis. Ethanol does not itself cause pancreatitis in rats, but it has been reported to worsen caeruleinhyperstimulation pancreatitis, suggesting that the two agents may act synergistically (Quon et al, 1992). Administration of specific CCK antagonists have been reported to reduce pancreatic injury induced by diet (Saluja et al, 1989), bile salts (Bilechik et al, 1990), and ischaemia (Modlin et al, 1988). These studies suggest that CCK receptor stimulation may contribute to injury in many forms of pancreatitis. The ability of ethanol to sensitize the acinar cell to the effects of CCK has been examined in isolated pancreatic acini (Katz et al, 1996). Studies that examine the pathological effects of ethanol should employ levels that are pathologically relevant. Thus, the effects of levels of greater than 50 mM ethanol may not be pertinent to human disease. Although most in vivo studies of ethanol have not demonstrated a pathological effect on the pancreas, it is not clear that pathological levels were reached in these models. That only a small number of humans who abuse ethanol develop pancreatitis suggests that other genetic and environmental factors play a role in the development of alcoholic pancreatitis. Ethanol has been found to sensitize G-protein-linked receptors to agonists. Although a low concentration of ethanol (5 mM) has no effect on cAMP generation, it dramatically enhances beta-agonist-dependent stimulation of cAMP (Nagy et al, 1992). A similar pattern of sensitization has been observed with ethanol and cholecystokinin. While ethanol (25 mM) has no effect on PCA1 processing in isolated rat acini, it does sensitize the cells to CCK (Figure 4). This effect was dependent on the duration of ethanol treatment and its concentration. Thus, the addition of ethanol to physiological concentrations of CCK (0.1 nM) had effects on pro-carboxypeptidase processing similar to those of CCK hyperstimulation (0.1 µM). As shown in Figure 5, the addition of the CCK receptor inhibitor L364,718 blocked the enhanced conversion (Katz et al, 1996). Although ethanol sensitized the acinar cell to the effects of CCK, it had little effect on carbachol-stimulated PCA1 processing. Thus, ethanol appears to sensitize the acinar cell to selective agents. Because ethanol stimulates zymogen processing when CCK levels are within the physiological range, these observations may be relevant to human disease.
Mechanisms of intracellular zymogen activation
233
Figure 4. Ethanol sensitizes the acinar cell to the effects of CCK on PCA1 processing. (A) Effect of adding increasing concentrations of ethanol with a low dose of CCK (0.1 nM) on acini treated for 30 minutes (right) compared to a concentration of CCK (0.1 µM) (left) that generates maximal conversion (mean ±SEM). (B) The time course of CA1 generation by acini after adding ethanol (25 mM) 5, 10 and 30 minutes before terminating 30 minutes treatment with 0.1 nM CCK. For the 45 minute time point, ethanol was added 15 minutes prior to the addition of CCK. *P <0.001 versus CCK 0.1 nM. Adapted from Katz et al (1996). US = unstimulated.
THEORETICAL MECHANISMS OF ZYMOGEN ACTIVATION At least seven pathways have the potential to be involved in the intracellular conversion of pancreatic zymogens to active (mature) enzymes: (1) trypsinogen autoactivation to trypsin (Figarella et al, 1988), (2) lysosomal hydrolase cathepsin B cleavage of trypsinogen to trypsin (Greenbaum and Hirshkowitz, 1961; Figarella et al, 1988), (3) enhanced susceptibility of zymogens to proteolysis because of oxidation or decondensation (Pacifici et al, 1993), (4) leakage of zymogens and lysosomal enzymes into the cytoplasm and subsequent proteolytic activation (Wilson et al, 1992), (5) shunting of zymogens into membrane-bound compartments that contain active proteases (Steer, 1988; Gorelick and Matovcik, 1995), (6) uptake and processing of secreted zymogens by endocytic pathways (Lerch et al, 1995) and (7) diminished activity of the intracellular pancreatic trypsin inhibitor. The mechanisms that have received the most attention are trypsinogen auto-activation and cathepsin B activation of trypsinogen. Trypsinogen auto-activation requires an acidic pH and is enhanced in the presence of Ca2+ (Kassell and Kay, 1973; Figarella et al, 1988). The affinity of pancreatic trypsin inhibitor is greatest at a neutral pH and is reduced at an acidic pH. Thus, conditions favouring the intracellular formation of active trypsin require an acidic pH. The generation of low-pH compartments within the acinar cell during experimental pancreatitis may be important to trypsinogen auto-activation and the expression of tryptic activity (Niederau and Grendell, 1988). Cathepsin B activation of trypsinogen in vitro has been reported to proceed optimally at an acidic pH (Figarella et al, 1988) or at a neutral pH (Lerch et al, 1993a). Under basal conditions cathepsin B is found within zymogen granules (reviewed by Gorelick and Matovcik, 1995). During the early stages (within 1–2 hours) of induction of some forms of experimental pancreatitis, increased amounts of lysosomal
234
F. S. Gorelick and T. Otani
Figure 5. Ethanol stimulation of PCA1 processing requires CCK receptor occupancy. Acini were exposed to CCK in physiological (0.1 nM) or hyperstimulating (0.1 µM) concentrations for 30 minutes in the presence or absence of ethanol. Some acini were pre-treated for 10 minutes with the CCK receptor antagonist, L364,718. The generation of CA1 is expressed as per control. *P <0.001 versus paired value without L364,718. Adapted from Katz et al (1996). US = unstimulated.
hydrolases are found in zymogen-containing compartments (Saluja et al, 1987; Lerch et al, 1993b). STUDIES RELATING TO THE MECHANISM AND SITE OF PROCESSING Processing takes place in a secretory compartment As with CCK or carbachol, bombesin stimulates secretion from the acinar cell. However, in contrast to the biphasic secretory response to CCK or carbachol, acini stimulated by bombesin exhibit a monophasic response (Figure 6A). These two patterns of secretory responsiveness correlate with the effects of these two agents on the apical actin cytoskeleton. An intact actin cytoskeleton is required for secretion from the pancreatic acinar cell (Muallem et al, 1995). High concentrations of CCK result in disruption of the apical actin cytoskeleton of the acinar cell, but high concentrations of bombesin do not disrupt the actin cytoskeleton (O’Konski and Pandol, 1993). Exposure of acini to each of these agents results in enhanced zymogen conversion (Figure 6B; Grady et al, 1998). However, in contrast to CCK or carbachol, the active enzyme forms generated by bombesin treatment are released into the medium (Figure 7). Even after bombesin stimulation, approximately 20% of CA1 is associated with the acinar pellet (not shown). Because bombesin-treated cells demonstrated intact morphology by transmission electron microscopy, excluded trypan blue, and did not release LDH, it is unlikely that leakage of active enzyme from cells accounts for these findings (Grady et al, 1998). There are two important implications of the bombesin studies. First, they suggest that proteolytic conversion of PCA1 to mature CA1 takes place within the secretory
Mechanisms of intracellular zymogen activation
235
Figure 6. Effects of CCK (䊐) and bombesin (䉬) on amylase secretion and PCA1 processing. Isolated pancreatic acini were stimulated with increasing concentrations of CCK or bombesin for 30 minutes to measure secretion and for 45 minutes to measure conversion. (A) Amylase secretion into the medium. (B) Generation of CA1 expressed as a percentage of the control. Adapted from Grady et al (1998). Acini
Medium PCA1 CA1
A
B
C
D
Figure 7. CA1 generated by CCK hyperstimulation is not secreted. Isolated pancreatic acini were hyperstimulated using cholecystokinin (CCK) (0.1 µM) (A, C), or treated with saline (B) or bombesin (0.1 µM) (D) for 30 minutes. The medium and acini were separated, and proteins were subjected to SDS–PAGE and processed to immunoblot analysis. The upper band is pro-carboxypeptidase A1 (PCA1) and the lower band is mature carboxypeptidase A1 (CA1). Both CCK (A) and bombesin (not shown) stimulate the generation of CA1. In the case of CCK, CA1 is not released into the medium (C), but after bombesin (D), CA1 appears in the medium. Adapted from Grady et al (1998).
pathway. Second, they suggest that the signals for zymogen conversion (elicited with both CCK and bombesin) may be different from those that lead to an inhibition of enzyme secretion (observed only with CCK). These observations may have physiological relevance and correlate with the effects of high concentrations of CCK or bombesin on the pancreas in vivo. While CCK hyperstimulation causes pancreatitis, bombesin does not. A conclusion that follows from these observations is that both zymogen conversion and the retention of enzymes within the acinar cell are required to initiate pancreatitis. Dependence of processing on protease activation Proteases differ in their mechanism of activation. This fact can be used to help define the mechanism of zymogen activation in the pancreatic acinar cell. Trypsin belongs to the serine protease family and cathepsin B is a thiol protease. Protease inhibitors can selectively block the activity of members of each family. The addition of soybean
236
F. S. Gorelick and T. Otani
trypsin inhibitor, a high-molecular-weight serine protease inhibitor that would not be expected to enter the acinar cell, has no effect on PCA1 processing (Leach et al, 1991). This finding suggests that zymogen processing does not take place in the incubation medium. The low-molecular-weight thiol protease inhibitor E-64 enters cells and is a potent inhibitor of cathepsin B. Although addition of this inhibitor to acinar cells blocked virtually all cathepsin B activity, it did not affect the processing of PCA1 associated with CCK hyperstimulation (Figure 8). In contrast, benzamidine, a permeant serine protease inhibitor, reduced the processing of PCA1 to basal levels. These findings suggest that activation of a serine protease within the acinar cell is required to stimulate PCA1 processing. Similar to its affect on PCA1 caerulein hyperstimulation of isolated pancreatic acini results in increased trypsinogen processing assayed using tryptic activity and the generation of TAP (Klonowski-Stump et al, 1998). The serine protease inhibitor FUT-175 blocked the activation but not the thiol protease inhibitors NCO-700 or E-64. Notably, these inhibitors blocked >95% of cathepsin B activity. Another study has shown that a high concentration of a more permeable form of E-64, E-64d, does block the generation of tryptic activity in cerulein-stimulated acini (Saluja et al, 1997). Finally, preliminary studies have suggested that inhibition of thiol proteases may result in increased levels of CA1 in hyperstimulated acini (Gorelick et al, 1992). The last observation suggests that thiol proteases may degrade activated enzymes. As discussed elsewhere (Gorelick and Matovcik, 1995), additional studies are required to resolve the differences between the results of these investigations.
Figure 8. The processing of PCA1 is dependent on activation of serine proteases. Acini were pre-treated for 30 minutes with either the serine protease inhibitor benzamidine (10 mM) or the thiol protease inhibitor, E-64 (0.1 mM) prior to hyperstimulation by CCK (0.1 µM). *P <0.05 versus CCK 0.1 µM. Adapted from Leach et al (1991).
Dependence on a low-pH compartment To examine the role of intracellular pH on zymogen processing, the pH of acidic intracellular compartments has been increased using the weak base chloroquine and the proton inophore monesin (Leach et al, 1991). Using concentrations of these agents that did not affect the secretory response to cholecystokinin, their effects on CCKinduced PCA1 processing was examined. As shown in Figure 9, both agents dramatically reduced the processing of this zymogen. The effects of these agents on the
Mechanisms of intracellular zymogen activation
237
Figure 9. The processing of PCA1 is dependent on a low pH compartment. Acini were pre-treated for 30 minutes with either monensin (10 µM) or chloroquine (40 µM) to increase intracellular pH and then hyperstimulated by CCK. *P < 0.05 versus CCK 0.1 µM. Adapted from Leach et al (1991).
generation of tryptic activity or TAP have not been reported. These findings suggest that a low pH compartment play a role in zymogen processing in the pancreatic acinar cell. The pH of compartments in the secretory pathway has been estimated using a weak base and an immunoelectron microscopy technique (Orci et al, 1987). In that study the Golgi complex was mildly acidic (pH 6.8) and condensing vacuoles were more acidic (pH 6.5). Zymogen granules were less acidic (pH ~ 6.9). Although the effect of hyperstimulation on pH in these compartments has not been examined, it is possible that a decrease in the pH initiates processing. Alternatively, the movement of nascent proteins through these compartments may be delayed during hyperstimulation so that they are exposed to a lower pH for a longer period. In either event, a low pH appears to play a central role in the processing of PCA1 to its active form. The importance of a low pH environment in the processing of other zymogens needs to be examined. Trypsinogen processing takes place in a distinct cellular compartment To examine the compartments associated with zymogen processing, double labelling experiments were performed using antibodies to TAP and antibodies to GRAMP-92 (granule membrane protein-92) an integral membrane protein and marker of lysosomes and recycling endosomes and in vivo caerulein hyperstimulation (Laurie et al, 1992; Otani et al, in press). High-magnification immunofluorescence studies demonstrated that TAP immunoreactivity was localized to structures just below the zymogen granule-enriched region (Figure 10). Localization of TAP immunoreactivity using immunogold techniques and electron microscopy has confirmed that TAP is generated in irregular vesicular structures and not in zymogen granules (Otani et al, 1998). Double labelling studies demonstrated that most TAP immunoreactivity is confined to GRAMP-92-positive structures. These findings suggest that zymogen processing takes place in a compartment that has overlap with endosomal or lysosomal markers. It is likely that this is also the compartment responsible for the processing of PCA1 to CA1. Together with data demonstrating that bombesin stimulates the processing of PCA1 in a secretory compartment (see Grady et al, in press and Figure 7), the findings suggest that processing might proceed in a secretory compartment that has overlap with
238
F. S. Gorelick and T. Otani
Figure 10. TAP is generated in a distinct subcellular compartment. After 45 minutes of in vivo hyperstimulation with caerulein, the pancreas was fixed and processed for phase microscopy (A), TAP (B) and GRAMP-92 (C) immunoreactivity. TAP appeared in a rounded structure just below the zymogen granule compartment (arrowheads). TAP immunoreactivity was contained in a GRAMP-92-positive structure.
endosomes or lysosomes. Recent studies have described a regulated secretion from lysosomes in several types of cell, including cultured fibroblasts and epithelial cells (Rodriguez et al, 1997). Whether zymogen processing in the acinar cell takes place in a similar compartment remains unclear. SUMMARY Zymogen processing can take place within the pancreatic acinar cell. This processing can be stimulated by supraphysiological concentrations of naturally occurring ligands. In vivo and in vitro studies demonstrate that processing is enhanced soon (15 minutes) after hyperstimulation with CCK or caerulein. Ethanol sensitizes the acinar cell to the effects of CCK; physiological concentrations of CCK cause enhanced zymogen processing in the presence of physiologically attainable ethanol levels. Zymogen processing is dependent on activation of serine proteases and a low pH compartment. However, the role of thiol proteases, including cathepsin B, remains unclear. Zymogen processing takes place within the acinar cell and not in zymogen granules. The processing may take place in a distinct secretory compartment. Finally, zymogen activation alone may not be sufficient to cause acinar cell injury and pancreatitis. Disruption of the secretory mechanisms and the apical actin cytoskeleton may cause the active enzymes to be retained in the acinar cell and be required for the initiation of pancreatitis. REFERENCES Appelros S, Thim L & Borgstorm A (1998) Activation peptide of carboxypeptidase B in serum and urine in acute pancreatitis. Gut 42: 97–102. Bialek R, Willemer S, Arnold R & Adler G (1991) Evidence of intracellular activation of serine proteases in acute cerulein-induced pancreatitis in rats. Scandinavian Journal of Gastroenterology 26: 190–196.
Mechanisms of intracellular zymogen activation
239
Bilchik AJ, Zucker KA, Adrian TE & Modlin IM (1990) Amelioration of cholinergic-induced pancreatitis with a selective cholecystokinin receptor antagonist. Archives of Surgery 125: 1546–1549. Cavallini G, Tittobello A, Frulloni E et al (1996) Gabexate for the prevention of pancreatic damage related to endoscopic retrograde cholaniopancreatography. New England Journal of Medicine 335: 919–923. Figarella C, Miszczuk-Jamska B & Barret AJ (1988) Possible lysosomal activation of pancreatic zymogens: activation of both human trypsinogens by cathepsin B and spontaneous acid activation of human trypsinogen. Biological Chemistry Hopp-Seyler 369: 293–298. Gorelick F & Matovcik L (1995) Lysosomal enzymes and pancreatitis. Gastroenterology 109: 620–625. Gorelick FS, Adler G & Kern HF (1993) Cerulein-induced pancreatitis. In Go VLW et al (eds) The Pancreas: Biology, Pathobiology, and Disease, pp 501–526. New York: Raven Press. Gorelick FS, Modlin IM, Leach SD et al (1992) Intracellular proteolysis of pancreatic zymogens. Yale Journal of Biology and Medicine 65: 407–420. *Grady T, Mah’moud M, Otani T et al (1998) Zymogen proteolysis within the pancreatic acinar cell is associated with cellular injury. American Journal of Physiology 275: G1010–G1017. Greenbaum LA & Hirshkowitz A (1961) Endogenous cathepsin activaties trypsinogen in extracts of dog pancreas. Proceedings of the Society for Experimental Biology and Medicine 107: 74–76. Gress TM, Arnold R & Adler G (1990) Structural alterations of pancreatic microvasculature in ceruleininduced pancreatitis in the rat. Research in Experimental Medicine 190: 401–412. Gudgeon AM, Heath DI, Hurley P et al (1990) Trypsinogen activation peptides assay in the early prediction of severity of acute pancreatitis. Lancet 335: 4–8. Kassell B & Kay J (1973) Zymogens of proteolytic enzymes. Science 180: 1022–1027. *Katz M, Carangelo R, Miller LJ & Gorelick F (1996) Effect of ethanol on cholecystokinin-stimulated zymogen conversion in pancreatic acinar cells. American Journal of Physiology 270: G171–G175. Klonowski-Stump H, Lüthen R, Han B et al (1998) Inhibition of cathepsin B does not affect the intracellular activation of trypsinogen by cerulein hyperstimulation in isolated rat pancreatic acinar cells. Pancreas 16: 96–101. Lasson A & Ohlsson K (1984) Protease inhibitors in acute pancreatitis: correlation between biochemical changes and clinical course. Scandinavian Journal of Gastroenterology 19: 779–786. Laurie SM, Mixon MB, Brand SH & Castle JD (1992) A secretion granule membrane protein (GRAMP 92) is found in non-granule membranes including those of the endocytic pathway. European Journal of Cell Biology 58: 12–27. Leach SD, Modlin IM, Scheele GA & Gorelick FS (1991) Intracellular activation of digestive zymogens in rat pancreatic acini. Stimulation by high doses of cholecystokinin. Journal of Clinical Investigation 87: 362–366. Lerch MM, Saluja AK, Dawra R et al (1993a) The effect of chloroquine administration on two experimental models of acute pancreatitis. Gastroenterology 104: 1768–1779. Lerch MM, Saluja AK, Runzi M et al (1993b) Pancreatic duct obstruction triggers acute necrotizing pancreatitis in the opossum [see comments]. Gastroenterology 104: 853–861. Lerch MM, Saluja AK, Runzi M et al (1995) Luminal endocytosis and intracellular targeting by acinar cells during early biliary pancreatitis in the opossum. Journal of Clinical Investigation 95: 2222–2238. Luthen R, Niederau C & Grendell JH (1995) Intrapancreatic zymogen activation and levels of ATP and glutathione during caerulein pancreatitis in rats. American Journal of Physiology 268: G592–G604. *Mithofer K, Fernandex-del Castillo C, Rattner D & Warshaw, AL (1998) Subcellular kinetic of early typsinogen activation in acute rodent pancreatitis. American Journal of Physiology 274: G71–G79. Modlin I, Bilchik A, Zucker K et al (1988) Cholecystokinin augmentation of ‘surgical’ pancreatitis. Archives of Surgery 124: 574–578. Muallem S, Kwiatkowska K, Xu X & Yin HL (1995) Actin filament disassembly is a sufficient final trigger for exocytosis in nonexcitable cells. Journal of Cell Biology 128: 589–598. Nagy LE & De Silva SE (1992) Ethanol increases receptor-dependent cyclic AMP production in cultured hepatocytes by decreasing G(i)-mediated inhibition. Biochemical Journal 286: 681–686. Neuschwander-Tetri BA, Ferrell LD, Sukhabote RJ & Grendell JH (1992) Glutathione monoethyl ester ameliorates caerulein-induced pancreatitis in the mouse. Journal of Clinical Investigation 89: 109–116. Niederau C & Grendell JH (1988) Intracellular vacuoles in experimental acute pancreatitis in rats and mice are an acidified compartment. Journal of Clinical Investigation 81: 229–236. Niederau C, Niederau M, Borchard F et al (1992) Effects of antioxidants and free radical scavengers in three different models of acute pancreatitis. Pancreas 7: 486–496. *Norman J (1998) The role of cytokines in the pathogenesis of acute pancreatitis. American Journal of Surgery 175: 76–83. O’Konski MS & Pandol SJ (1993) Cholecystokinin JMV-180 and caerulein effects on the pancreatic acinar cell cytoskeleton. Pancreas 8: 638–646. Orci L, Ravazzola M & Anderson R (1987) The condensing vacuole of exocrine cells is more acidic than the mature secretory vesicle. Nature 326: 77–79.
240
F. S. Gorelick and T. Otani
*Otani TS, Chepilko M, Grendell JH & Gorelick FS (in press) Co-distribution of trypsinogen activation peptide and the granule membrane protein, GRAMP-92, in rat cerulein-induced pancreatitis. American Journal of Physiology 275: G999–G1009. Pacifici RE, Kono Y & Davies KJA (1993) Hydrophobicity as the signal for selective degradation of hydroxyl radical-modified hemoglobin by the multicatalytic proteinase complex, proteosome. Journal of Biological Chemistry 268: 15 405–15 411. Quon MG, Kugelmas M, Wisner JR et al (1992) Chronic alcohol consumption intensifies caerulein-induced acute pancreatitis in the rat. International Journal of Pancreatology 12: 31–39. Rodriguez A, Webster P, Ortego J & Andrews NW (1997) Lysosomes behave as Ca2+-regulated exocytic vesicles in fibroblasts and epithelial cells. Journal of Cell Biology 137: 93–104. Saluja A, Sadamitsu H, Saluja M et al (1987) Subcellular redistribution of lysosomal enzymes during caeruleininduced pancreatitis. American Journal of Physiology 253: G508–G516. Saluja AK, Saluja M, Printz H et al (1989) Experimental pancreatitis is mediated by low-affinity cholecystokinin receptors that inhibit digestive enzyme secretion. Proceedings of the National Academy of Sciences of the USA 86: 8968–8971. Saluja AK, Donovan EA, Yamanaka K et al (1997) Cerulein-induced in vivo activation of trypsinogen in rat pancreatic acini is mediated by cathepsin B. Gastroenterology 113: 304–311. Scheele G, Adler G & Kern H (1987) Exocytosis occurs at the lateral plasma membrane of the pancreatic acinar cell during supramaximal secretagogue stimulation. Gastroenterology 92: 345–353. Schmid S, Modlin IM, Stoch A et al (in press) Telenzepine-sensitive muscarinic receptors on the rat pancreatic acinar cell. American Journal of Physiology Schmidt J, Fernandez-del Castillo C, Rattner DW et al (1992) Trypsinogen-activation peptides in experimental rat pancreatitis: prognostic implications and histopathologic correlates. Gastroenterology 103: 1009–1016. Steer M (1988) Pathogenesis of acute pancreatitis. Annual Review of Medicine 39: 95–105. Suzuki M, Isaji S, Stanten R et al (1992) Effect of protease inhibitor FUT-175 on acute hemorrhagic pancreatitis in mice. International Journal of Pancreatology 11: 59–65. Whitcomb DC, Gorry MC, Preston RA et al (1996) Hereditary pancreatitis is caused by a mutation on the cationic trypsinogen gene. Nature Genetics 14: 141–145. Wilson JS, Apte MV, Thomas MC et al (1992) Effects of ethanol, acetaldehyde and cholesteryl esters on pancreatic lysosomes. Gut 33: 1099–1104.