Pathologically relevant trypsinogen activation in pancreatitis.

In a recent intriguing report, Chen and colleagues (2022) investigated the effect of pancreas-specific deletion of cathepsin B ( Ctsb Δpan mice) and cathepsin L ( Ctsl Δpan mice) on cerulein-induced trypsin activity and pancreatitis severity [ 25 ]. As expected, intrapancreatic trypsin activity was reduced in Ctsb Δpan mice and increased in Ctsl Δpan mice when compared to C57BL/6N controls. Furthermore, the double-knockout Ctsb Δpan , Ctsl Δpan mice exhibited slightly elevated trypsin activity relative to wild-type controls. Interestingly, however, when pancreatitis severity was evaluated by measurement of serum amylase activity, pancreas histology, pancreatic cytokine mRNA expression and assessment of autophagy, the authors found that pancreas-specific CTSB and CTSL deficiencies did not alter disease severity. Acinar cell apoptosis was also unaffected, consistent with the findings of Halangk et al. (2000) on global Ctsb-KO mice. In the authors' interpretation, these results challenged the trypsin-central dogma by suggesting that progression and severity of pancreatitis was independent of trypsinogen activation. Concurrently with this study, we published our findings on novel trypsinogen mutant mouse strains specifically designed to dissect the role of cathepsin B-mediated trypsinogen activation and autoactivation, in the initiation and severity of pancreatitis [ 26 - 29 ]. Since both cathepsin B and trypsin share the same activation site within the trypsinogen activation peptide ( Figure 1B ), our aim was to develop mouse models that allow for experimental separation and selective study of the two activation processes. To this end, we introduced mutations into the native mouse cationic trypsinogen (isoform T7) to inhibit autoactivation while either preserving ( T7D23A,K24G mice) or enhancing ( T7D22A,K24G mice) sensitivity to cathepsin B-mediated activation ( Table 1 ). In these constructs, the K24G mutation changes the trypsin-sensitive Lys24 site to Gly, which is resistant to trypsin but still cleaved by cathepsin B. Replacement of the adjacent Asp23 and Asp22 amino-acid residues with Ala modifies the rate by which cathepsin B can activate the mutant trypsinogens [ 26 ]. After a single cerulein injection, we observed significantly increased trypsin activity in the T7D23A,K24G, T7D22A,K24G and C57BL/6N mice compared to the saline-treated controls. Trypsin activation was markedly higher in the cerulein-treated T7D22A,K24G mice compared to T7D23A,K24G and C57BL/6N mice, which showed similar trypsin levels. In contrast, trypsin activity remained at basal levels in the pancreas of cerulein-injected Ctsb-KO mice, in which the Ctsb gene was deleted globally. Remarkably, the considerable differences in intrapancreatic trypsin activity did not translate to changes in disease severity when acute pancreatitis was induced with repeated cerulein injections. We found that pancreas weight, amylase activity and histomorphological alterations were similar across all strains [ 27 ]. While these results reinforced the findings of Chen et al. (2022), our interpretation differs significantly. We concluded that in cerulein-induced experimental pancreatitis of C57BL/6N mice, intrapancreatic trypsin activity is entirely dependent on cathepsin B, and trypsinogen autoactivation plays no role in this process. More importantly, our results demonstrated that cathepsin B-mediated trypsin activity is not required for pancreatitis onset, nor does it influence disease severity. Thus, this early rise in intrapancreatic trypsin activity appears to represent a consequence or marker of disease development rather than a primary driver. The data presented by Chen et al. (2022) and our own observations provide context for earlier studies highlighting the discrepancy between intrapancreatic trypsin activity and disease pathology. Previously, Dawra and colleagues (2011) generated a murine model lacking the main mouse cationic (T7) trypsinogen isoform. These mice exhibited diminished intrapancreatic trypsin activity and partially ameliorated acinar cell necrosis when challenged with cerulein, both in vivo and in vitro [ 30 ]. However, despite the absence of T7 trypsinogen expression, T7 deficient mice did not show significant changes in pancreas and lung MPO content, indicating unchanged neutrophil infiltration. The authors concluded that while trypsin activation was necessary for early pancreatic damage, the progression of local and systemic injury did not depend on intrapancreatic trypsinogen activation. We believe that the lack of a strong protective phenotype in this trypsinogen-deficient model was due to the fact that cerulein-induced trypsinogen activation was mediated through cathepsin B. Based on our observations discussed above, intracellular trypsin activity elicited by cathepsin B has no significant effect on disease severity. Furthermore, studies on mice deficient in the trypsin-degrading enzyme cathepsin L ( Ctsl-KO ) showed markedly increased intrapancreatic trypsin activity following cerulein administration; while the severity of pancreatitis was unexpectedly mitigated, as judged by the decrease in serum amylase, lipase, and cytokines levels [ 12 ]. In addition, cerulein-treated Ctsl-KO mice exhibited less necrosis but developed a much greater extent of apoptosis compared to the wild-type controls. Similar findings were published on a mouse model lacking the cation-independent mannose 6- phosphate/insulin-like growth factor II receptor (CI-MPR) [ 31 ]. CI-MPR was shown to mediate trafficking of cathepsin B to lysosomes, thus CI-MPR deletion led to the redistribution of cathepsin B to the zymogen granule. Cerulein injections resulted in approximately 40% increase in intrapancreatic trypsin activity in CI-MPR-deficient mice compared to wild type, however this increase yet again failed to affect pancreatitis severity. The authors of this study suggested that co-localization of cathepsin B and trypsinogen was not sufficient to induce pancreatitis or contribute to pancreatic injury. Importantly, immunofluorescence experiments revealed that CI-MPR deficient mice injected with saline also exhibited large clusters of cytoplasmic vesicles containing both cathepsin B and trypsinogen, indicating that co-localization of these two enzymes can occur even without an exogenous stimulus. Moreover, the missorting of lysosomal enzymes to cytosolic vacuoles did not induce spontaneous intracellular activation of trypsinogen. This observation aligns with accounts showing that cathepsin B can be detected in the secretory compartment of the pancreas under physiological conditions. Using immunogold electron microscopy on healthy human pancreatic samples, cathepsin B was detected in the pancreatic ducts, acinar cell lumen and secretory vesicles of acinar cells, co-localized with trypsinogen [ 32 ]. Similarly, Van Acker and colleagues (2006) showed that while pretreatment with the CA-074me inhibited intrapancreatic trypsinogen activation and reduced pancreatitis severity, it did not alter co-localization of lysosomal enzymes with trypsinogen [ 22 ].

Taken together, the published observations suggest that autoactivation is the disease-relevant mechanism of intrapancreatic trypsinogen activation and cathepsin B-mediated intrapancreatic trypsin activity plays a lesser role in pancreatitis development. The critical questions remain: what determines pathogenicity of trypsin generated through autoactivation, and why does cathepsin B-elicited trypsinogen activation have limited impact on disease severity? As mentioned previously, protease activation can be routinely measured after supramaximal secretagouge stimulation both in vivo and ex vivo using isolated acinar cells. Immunofluorescence studies showed that treatment with supramaximal concentrations of CCK, cerulein and bombesin resulted in the intracellular generation of trypsinogen activation peptide [ 50 ]. Interestingly, however, acinar cell damage was observed only after CCK and cerulein stimulation, as judged by LDH release, trypan blue retention, mitochondrial swelling and vacuole formation. Bombesin treatment, in contrast, did not elicit acinar cell injury [ 50 ]. Furthermore, in vivo stimulation with bombesin resulted in similar edema levels to those seen with cerulein administration, although amylase levels only showed a moderate increase [ 51 ]. This can be explained by the fact that while supramaximal stimulation with CCK and cerulein lead to inhibition of acinar cell secretion, bombesin stimulation does not induce a secretion block. This suggests that acinar cell injury may depend not only on activation of digestive proteases but also on their retention within the cells. Therefore, the specific location of pathologically activated trypsin warrants further discussion. Numerous studies employing isolated acinar cells have demonstrated that trypsinogen can undergo intracellular activation [ 50 , 52 - 54 ]. Within the cell, the primary sites of trypsinogen activation have been localized to secretory vesicles, lysosomes and endosomes [ 17 , 23 , 55 ]. Inhibition of phosphatidylinositol 3-kinase (PI3K), a key regulator of vesicle trafficking, using wortmannin, was shown to reduce cathepsin B redistribution, decrease trypsinogen activation and attenuate pancreatitis severity in both secretagouge and duct-injection models [ 56 ]. Similarly, Messenger et al. (2015) reported that blocking early-to-late endosome maturation through inhibition of phosphatidylinositol 3-phosphate 5-kinase (PIKfyve) prevented intracellular trypsin accumulation and protected acinar cells from injury induced by CCK, bile acids and cigarette toxins [ 57 ]. Collectively these findings underscore the role of endosomal trafficking and vesicle maturation in regulating zymogen activation and pancreatitis pathogenesis. Furthermore, treatment with sodium arsenate induced the expression of heat shock protein (HSP) 70 expression, which inhibited cathepsin B redistribution, suppressed trypsinogen and NF-κB activation, and conferred protection against acute pancreatitis [ 58 ]. Autophagy has also emerged as a key intracellular regulator of trypsinogen activation and pancreatic inflammation [ 59 - 63 ]. Pancreas-specific deletion of Atg5 , a gene essential for autophagosome formation led to reduced trypsinogen activation both in isolated acinar cells and pancreatic homogenates, and milder acute pancreatitis phenotype [ 61 ]. Conversely, Diakopoulos et al. (2015) reported that mice with pancreatic Atg5 deficiency developed spontaneous chronic pancreatitis associated with increased endoplasmic reticulum stress and mitochondrial damage [ 62 , 63 ]. Malla et al. (2020) further argued that early trypsinogen activation does not occur in autophagosomes or autolysosomes but rather takes place in a different subcellular compartment of the secretory pathway [ 64 ]. However, they acknowledged that at later stages an autophagy-dependent wave of trypsinogen activation may occur. Insights from pH-modification studies added further complexity to our understanding of intracellular trypsinogen activation. Colocalization of digestive zymogens with cathepsin B occurs in large acidic cytoplasmic vacuoles, with the activation process undergoing at low pH. In vitro , administration of the weak base chloroquine increased the pH of acidic intracellular compartments, inhibited cathepsin B activity and consequent protease activation in cerulein-stimulated acinar cells [ 65 , 66 ]. Interestingly while intravenous administration of chloroquine significantly elevated intracellular pH, it failed to protect against either secretagogue- or diet-induced pancreatitis in vivo [ 67 ]. Lerch et al. (1993) concluded that while cathepsin B redistribution may be necessary for intracellular activation, it is not sufficient on its own to induce pancreatitis. They further proposed that either intracellular trypsinogen activation does not directly contribute to disease development, or that activation occurs through cathepsin B-independent mechanisms. Notably, although cathepsin B exhibits optimal activity at acidic pH, these findings suggest that it can also catalyze trypsinogen activation within a near-neutral pH range. Adding yet another layer, Sendler et al. (2018) discovered that infiltrating macrophages can act as secondary sites of trypsinogen activation [ 68 ]. They demonstrated that macrophages phagocytose zymogen-containing acinar vesicles, where trypsinogen becomes activated in a cathepsin B and pH-dependent manner, leading to NF-κB activation and the release of pro-inflammatory cytokines. This mechanism likely represents a secondary source of pancreatic trypsin activity that may impact inflammation. Investigation into subcellular kinetics of early trypsinogen activation provided additional insights. Mithofer and colleagues (1998) observed a significant increase in trypsinogen activation peptide levels predominantly in the zymogen fraction shortly after cerulein injection. Interestingly, three hours after cerulein stimulation, a proportionally greater elevation of the activation peptide was detected in the soluble postmicrosomal fraction, which the authors suggested corresponds to the interstitial space [ 16 ]. Furthermore, although cerulein injections alone induced mild edematous pancreatitis, combining cerulein stimulation with intraductal infusion of enterokinase resulted in a pronounced necro-hemorrhagic phenotype along with increased trypsinogen activation peptide levels in both blood and urine [ 69 ]. Fernández-del Castillo et al. (1994) explained that in cerulein-induced pancreatitis, a large pool of trypsinogen accumulates in the interstitial space, and its activation can lead to progression from mild to severe necrotizing pancreatitis [ 69 , 70 ]. The proposal that the interstitial space is a disease-relevant site of trypsinogen activation aligns with reports demonstrating pathologic basolateral exocytosis in acinar cells. While the apical plasma membrane accounts for only 10 % of the pancreatic acinar cell surface, it maintains a very rapid physiological secretory process. Upon disease-causing stimuli, however, the exocytotic machinery can be disrupted, resulting in basolateral secretion. Both physiological apical exocytosis and pathologic basolateral exocytosis are mediated by SNARE [soluble N-ethylmaleimide-sensitive factor (NSF) attachment protein (SNAP) receptors] complexes. Munc18c was identified as one of the accessory cytosolic proteins, which primes the SNARE complex and thereby regulates apical and basolateral exocytosis. A study by Gaisano et al. (2001) demonstrated that pathological basolateral secretion results from the phosphorylation of Munc18c and activation of the STX-4/SNAP23/VAMP8 SNARE complex on the basolateral membrane [ 71 , 72 ]. Pancreatic acini from heterozygous Munc18c-depleted mice (Munc18c +/− ) exhibited normal apical exocytosis of zymogen granules in response to physiological secretagogue stimulation. Interestingly when CCK-8 was administered in supramaximal doses, Munc18c +/− mouse acini showed reduced basolateral secretion compared to wild-type acini, likely due to a decreased formation of STX-4 SNARE complexes. Additionally, cerulein hyperstimulation in Munc18c +/− mice led to lower serum amylase and lipase levels, reduced pancreatic MPO activity, and diminished histopathological injury [ 72 , 73 ]. In addition to experimental pancreatitis models, Klöppel and colleagues (1986) examined pancreatic samples from patients with severe pancreatitis using immunocytochemistry and electron microscopy and found that acinar cells adjacent to fat necrotic areas released their granules through basolateral secretion [ 74 ]. Thus, both human and animal studies provide compelling evidence that upon pathological stimuli, acinar secretory granules fuse with the basolateral membrane and are secreted into the interstitial space.

To investigate the effects of trypsinogen autoactivation in vivo , knock-in mouse models were generated by introducing human pancreatitis-associated mutations into the native mouse cationic (T7) trypsinogen ( Table 1 ). The T7K24R mice carry the p.K24R mutation, which is analogous to the hereditary pancreatitis-associated p.K23R mutation in humans. This variant replaces the trypsinogen activation site Lys with an Arg, thereby increasing autoactivation by fivefold. As a result, T7K24R mutant mice exhibited more severe pancreatitis when challenged with the secretagogue cerulein as evidenced by increased edema, inflammatory cell infiltration and elevated plasma amylase activity [ 44 ]. Intrapancreatic trypsin activity was also significantly higher in cerulein-induced T7K24R mice compared to wild-type controls. In addition, Jancsó et al. (2022) reported that after an episode of cerulein-induced acute pancreatitis, T7K24R mice developed spontaneously progressive chronic pancreatitis characterized by extensive acinar cell atrophy, persistent macrophage infiltration, and diffuse fibrosis [ 45 ]. Since T7K24R mice developed no spontaneous pancreatitis, another model with more strongly accelerated trypsinogen autoactivation was generated. The T7D23A mice harbored the p.D23A mutation corresponding to the human hereditary pancreatitis-associated p.D22G mutation. This variant disrupts the negatively charged penta-Asp motif within the mouse trypsinogen activation peptide thereby mitigating its inhibitory effect against tryptic cleavage. Ultimately, this results in a 50-fold increase in trypsinogen autoactivation. T7D23A mice developed spontaneous acute pancreatitis around three weeks of age presenting with increased pancreatic edema, inflammatory cell infiltration, and centrally localized lobular necrosis [ 46 ]. Disease progression to early chronic pancreatitis occurred rapidly between one and two months, and end-stage disease was reached between four and six months of age, characterized by extensive acinar cell loss, dilated ducts, limited fibrosis and extensive fatty replacement. Markedly elevated spontaneous trypsin activity was detected in pancreatic homogenates of mice of four weeks and two months of age, demonstrating that trypsinogen autoactivation not only initiates pancreatitis but also drives its progression. Furthermore, Demcsák and Sahin-Tóth (2022) engineered a trypsinogen double mutant (p.D22N,K24R) with a 13-fold increased autoactivation rate that lay between those observed with T7K24R and T7D23A mice [ 47 ]. Homozygous T7D22N,K24R mice displayed a spontaneous pancreatitis phenotype, though with a slightly delayed onset as compared to heterozygous T7D23A mice. At the same time, heterozygous T7D22N,K24R mice did not show any signs of spontaneous pancreatic damage but developed more severe pathology when subjected to repeated cerulein injections. Notably, none of the three described mutations that accelerated autoactivation had a similar effect on cathepsin B-mediated trypsinogen activation, which was either unchanged or slightly decreased ( Table 1 ). Findings from these T7 knock-in models not only confirmed that trypsinogen mutations that stimulate autoactivation are directly pathogenic, but also demonstrated a clear correlation between the extent of autoactivation and pancreatitis severity. Most recently, to assess whether global deletion of the cathepsin B gene impacts disease onset and severity in the T7D23A mice, we generated T7D23A × Ctsb-KO mice with homozygous CTSB deletion (Tran and Geisz-Fremy, unpublished). We found that cathepsin B deficiency slightly attenuated the early trypsin signal and parameters of acute pancreatitis. However, by 5 weeks of age and thereafter, both T7D23A and T7D23A × Ctsb-KO mice exhibited similar progression of chronic pancreatitis, characterized by acinar atrophy, fibrosis, inflammatory cell infiltration and adipose replacement. Remarkably, when spontaneous protease activation was measured in pancreatic homogenates, both the T7D23A , and T7D23A × Ctsb-KO strains exhibited significantly increased trypsin activity, whereas only basal trypsin levels were detected in the Ctsb-KO and wild-type C57BL/6N controls. Trypsin activity was initially higher in T7D23A mice than in T7D23A × Ctsb-KO. These findings suggest that while cathepsin B may contribute to early trypsinogen activation and acute pathology, it is dispensable for the spontaneous intrapancreatic trypsinogen activation and overall course of chronic pancreatitis in the T7D23A model. Finally, Gui et al. (2020) published results on transgenic mice expressing the human PRSS1 gene with p.R122H mutation. These mice developed chronic pancreatitis after an episode of cerulein-induced acute pancreatitis [ 48 ]. From the same laboratory, Wang et al. (2022) demonstrated that introducing the human PRSS2 transgene to this model resulted in spontaneous, progressive pancreatitis in PRSS1 -p.R122H- PRSS2 double-transgenic mice [ 49 ]. The advantage of these models is the use of human trypsinogen isoforms, however, a limitation is the difficulty to control for gene dosage effects and the potentially random nature of genomic integration. Although the role of cathepsin B in these mouse models has not been studied, the known propensity of human trypsinogens for rapid autoactivation suggest that in these transgenic mice intrapancreatic trypsin activity and pancreatitis development were driven by trypsinogen autoactivation.

The potential impact of altered pancreatic acinar secretion on pancreatitis has also been highlighted in studies investigating the role of protease activated receptor 2 (PAR2). PAR2 is a G protein-coupled receptor activated by serine protease-mediated cleavage of its extracellular N-terminal domain. Genetic deletion of PAR2 was found to exacerbate cerulein-induced acute pancreatitis, suggesting a protective function in pancreatic injury [ 75 , 76 ]. In another report by Namkung et al. (2004), PAR2 demonstrated a protective effect locally in acinar and duct cells, whereas excessive PAR2 stimulation lead to increased vascular permeability and leukocyte infiltration, thereby contributing to systemic complications [ 77 ]. Notably, PAR2 was also shown to localize on the basolateral membrane of pancreatic ductal cells, a domain typically inaccessible to trypsin under physiological conditions. However, during acute pancreatitis, pathological disruptions such as tight junction breakdown and basolateral exocytosis from acinar cells could compromise the epithelial barrier and facilitate the aberrant exposure of ductal PAR2 to intrapancreatic trypsin [ 78 , 79 ]. Thus, while PAR2 appears to have both protective and pathological roles in pancreatitis, its precise function remains unclear. Beyond directly activating PAR2, trypsin initiates a cascade of downstream proteolytic events by cleaving and activating other zymogens. One key example is the conversion of pro-elastase to active elastase ( Figure 1A ). Once prematurely activated, elastase may amplify tissue injury through degradation of extracellular matrix proteins such as elastin, fibronectin, and collagen, thereby weakening structural integrity and promoting interstitial edema. Importantly, pancreatic pro-ealstase emerged as a risk gene when Moore and colleagues (2019) performed whole exome sequencing from patients with pancreatitis of previously unknown etiology [ 80 ]. They identified a rare missense mutation in the gene encoding pancreas-specific protease elastase 3B (CELA3B) that segregated with disease. Functional analysis revealed that the variant caused translational upregulation of CELA3B, which, they proposed that upon activation by trypsin, caused uncontrolled proteolysis and lead to chronic pancreatitis damage in mice homozygous for the Cela3b R89C and R89L mutations. While this aligns with the classic autodigestion concept, Tóth et al (2022) reported a splice-site variant resulting in diminished CELA3B expression in patients with chronic pancreatitis [ 81 ]. Collectively, these findings highlight that both increased and decreased elastase function may influence disease risk, and further studies are required to clarify the precise role of CELA3B in pancreatitis pathogenesis.

Chiari introduced the concept of pancreatic autodigestion in 1896 based on his observations on post-mortem autolysis of the pancreas [ 6 ]. Subsequently, the autodigestion hypothesis was extended to pancreatitis, positing that trypsinogen and other pancreatic protease zymogens become activated prematurely in the pancreas early in the course of pancreatitis, ultimately leading to the organ's self-digestion. Compelling evidence for ectopic zymogen activation emerged nearly a century later, when Geokas and Rinderknecht detected active trypsin in the pancreatic juice of patients with acute pancreatitis [ 7 ]. Under physiological conditions, trypsinogen is secreted from the pancreas as an inactive precursor and is converted to active trypsin in the small intestine by the brush-border-localized protease called enteropeptidase (legacy name enterokinase). Once activated, trypsin then proteolytically processes other pancreatic zymogens to their active forms, including chymotrypsinogens, pro-elastases and pro-carboxypeptidases ( Figure 1A ). Activation of trypsinogen by enteropeptidase occurs through limited proteolysis of the Lys-Ile peptide-bond at the C-terminal end of the trypsinogen activation peptide. Under pathological conditions, this same peptide bond can be cleaved either by trypsin or by the lysosomal cysteine protease cathepsin B leading to intrapancreatic trypsinogen activation [ 8 ] ( Figure 1B ). Cathepsins are ubiquitously expressed enzymes with diverse pathophysiological functions. Their proteolytic activity contributes to various cellular processes including antigen processing, protein activation and degradation. Consequently, cathepsins play pivotal roles in inflammatory, cardiovascular and bone disorders as well as in cancer [ 9 , 10 ]. In the pancreas, cathepsin B was shown to mediate trypsinogen activation [ 11 ], whereas cathepsin L was identified as a trypsinogen-degrading enzyme [ 12 ]. The role of cathepsin B in intrapancreatic trypsinogen activation has been predominantly studied in rodent models of experimentally-induced pancreatitis [ 5 , 13 - 15 ]. In these studies, pancreatitis is often elicited by supramaximal stimulation with the secretagogue cerulein, which leads to the intracellular colocalization of cathepsin B and trypsinogen. This occurs either through a missorting process that redistributes lysosomal proteases into a zymogen granule-enriched subcellular compartment or through the fusion of lysosomes with zymogen granules ( Figure 2 ). Subcellular fractionation studies revealed that cathepsin B redistributes from the lysosomes to the zymogen granules within 15 minutes of cerulein administration, with activation of trypsinogen occurring around the same time [ 16 ]. Immunolocalization experiments showed that during the early stages of pancreatitis, cathepsin B and the cleaved trypsinogen activation peptide were contained in the same cytoplasmic vacuoles and subsequent to its activation, trypsin was released to the cytosol [ 17 ]. Importantly, the colocalization phenomenon was observed not only in the secretagogue but also in the diet-induced, duct obstruction, duodenal loop, and the taurocholate models of experimental pancreatitis. In each of these, co-localization of lysosomal cathepsins and digestive zymogens occurred prior to the appearance of organ injury [ 18 - 20 ]. Further indications that cathepsin B-mediated trypsin activity may contribute to pancreatitis severity came from studies where cathepsin B was inhibited either by pharmacological agents or by genetic modification. Using the cell permeable cathepsin B inhibitor E-64d in vitro , intraacinar trypsinogen activation was prevented after stimulation with cerulein [ 21 ]. In vivo , CA-074me, a similar inhibitor with a prolonged plasma half-life, was found to be effective not only in decreasing intrapancreatic trypsinogen activation but also in alleviating pancreatitis severity by markedly reducing inflammatory cell infiltration, acinar cell necrosis, pancreatic edema and plasma amylase levels. Notably, CA-074me was tested in both secretagogue and duct-infusion models of pancreatitis [ 22 ]. Global genetic deletion of cathepsin B led to a significant decrease in pancreatic trypsin activity and lower levels of trypsinogen activation peptide in pancreatic homogenates from cerulein-treated mice as compared to wild-type controls [ 23 ]. When challenged with repeated cerulein injections, cathepsin B deficient ( Ctsb-KO ) mice showed decreased serum amylase and lipase levels and less necrosis, while apoptosis remained unaffected. The authors interpreted these results as direct evidence of reduced pancreatic injury, however the effect size was modest, and Ctsb-KO mice still developed significant acute pancreatitis. More recently, a study by Modenbach et al. (2025) found that cystatin 3 (CTS3) is a critical regulator of cathepsin B and L activity during cerulein-induced pancreatitis, and genetic deletion of CTS3 in mice led to increased trypsinogen activation and more severe acute pancreatitis in response to cerulein administration [ 24 ].

Published findings indicate that while cathepsin B-mediated trypsinogen activation may contribute to intrapancreatic trypsin activity in experimental models, its impact on pancreatitis development, severity and progression appears limited. In contrast, extensive genetic, biochemical and mouse model evidence strongly supports trypsinogen autoactivation as the more pathologically relevant mechanism. Emerging research also suggests that pathogenic autoactivation of trypsinogen may occur extracellularly, generating harmful trypsin in the interstitial space and thereby initiating and/or exacerbating the inflammatory response ( Figure 2 ). Future studies should focus on defining the proteolytic target(s) of extracellular trypsin and their link to inflammation. Our improved understating of trypsinogen activation mechanisms and their role in pancreatitis pathogenesis open new avenues for therapeutic intervention. With appropriate animal models of trypsin-dependent pancreatitis now available, it is time to revisit protease inhibitors that were tested in earlier studies. Multiple clinical and experimental efforts have explored protease inhibitor therapies aimed at mitigating trypsin activation and reducing pancreatitis severity [ 82 - 88 ]. While many of these compounds showed protective effects in rodent models, particularly in cerulein-induced pancreatitis, their clinical efficacy has been inconsistent. Aprotinin, a bovine trypsin inhibitor was among the first agents tested, and although early experimental studies suggested reduced morbidity and mortality, clinical trials failed to reproduce these benefits [ 82 ]. Synthetic serine protease inhibitors such as nafamostat and gabexate were subsequently developed with improved tissue distribution and demonstrated benefits in preclinical studies, but translation to patients produced variable outcomes [ 83 - 88 ]. Several limitations may explain this discrepancy. First, most early preclinical studies relied on cerulein-induced models, which do not faithfully recapitulate trypsin-dependent disease mechanisms. Second, these compounds suffer from unfavorable pharmacokinetics, including short plasma half-lives, low bioavailability, and limited pancreatic accumulation following systemic administration. Third, the lack of specificity complicates interpretation: aprotinin, nafamostat, gabexate and camostat can inhibit multiple proteases beyond trypsin, including thrombin, kallikrein, and complement factors. Despite these challenges, some inhibitors remain of interest. Camostat and dabigatran, which exhibit modest activity against trypsin, have shown benefit in trypsin-dependent mouse models, and camostat continues to be used clinically in Japan [ 48 , 88 ]. In addition to focusing solely on intracellular trypsin inhibition, future efforts should explore strategies to target extracellular trypsin, which may be potentially accessible to larger molecules, such as antibodies or proteinaceous trypsin inhibitors capable of neutralizing trypsin activity. Gene therapy using adeno-associated viral (AAV) vectors has emerged as a potentially safe and efficient method for delivering the therapeutic genes to specific organs and cell types [ 89 ]. With respect to pancreatitis, a recent study demonstrated that a single intraperitoneal injection of AAV8 carrying the human trypsin inhibitor SPINK1 successfully targeted the pancreas and reduced pancreatitis severity in experimental mouse models [ 90 ]. These findings not only provide proof-of-concept that sustained expression of trypsin inhibitors can alleviate pancreatic inflammation but also highlight the potential of AAV-based approaches to overcome many of the pharmacokinetic and specificity limitations that hinder conventional protease inhibitors.