Chymotrypsin C (CTRC) mutations in chronic pancreatitis

Abstract

Chronic pancreatitis is a persistent inflammatory disorder characterized by destruction of the pancreatic parenchyma, maldigestion, and chronic pain. Mutations in the CTRC gene encoding the digestive enzyme chymotrypsin C have been shown to increase the risk of chronic pancreatitis in European and Asian populations. Here we review the biochemical properties and physiological functions of human CTRC; summarize the functional defects associated with CTRC mutations and discuss mechanistic models that might explain the increased disease risk in carriers.

Genetic risk factors in chronic pancreatitis

Chronic pancreatitis is a relapsing or continuing inflammatory disease of the pancreas characterized by progressive destruction of the pancreatic parenchyma which results in pancreatic fibrosis, acinar cell atrophy and duct irregularities with calcifications [1–3]. Clinical features include chronic abdominal pain, maldigestion and diabetes mellitus. The reported annual incidence of chronic pancreatitis is 3–10 per 100,000 population [1–3]. Chronic pancreatitis secondary to environmental or metabolic causes is mostly associated with chronic alcohol abuse, possibly smoking [4–6], and hypercalcemia due to hyperparathyroidism. Primary or idiopathic chronic pancreatitis is diagnosed in 15–30% of cases, and some of these patients have a positive family history (familial chronic pancreatitis). In a subgroup of families, inheritance of chronic pancreatitis follows an autosomal dominant pattern, and if the disease is present at least in two first-degree or three second-degree relatives in two or more generations, hereditary chronic pancreatitis is diagnosed [7]. Disease penetrance in classic hereditary pancreatitis is about 70–80%, but expressivity is highly variable, with most patients having mild disease [8].

Although the first description of hereditary chronic pancreatitis dates back to the 1950s [9], the underlying genetic defect remained obscure until 1996 when the genetic locus was mapped to chromosome 7q35 [10–12], and a missense mutation (p.R122H) in the serine protease 1 (PRSS1) gene encoding cationic trypsinogen was identified as a causative alteration [13]. Follow-up studies found additional mutations in the PRSS1 gene not only in patients with hereditary or familial but also in individuals with idiopathic chronic pancreatitis with no family history [14, 15 and references therein]. Triplication and duplication of the trypsinogen locus was also observed in idiopathic and hereditary chronic pancreatitis [16, 17]. Furthermore, mutations in the serine protease inhibitor Kazal type 1 (SPINK1) gene encoding pancreatic secretory trypsin inhibitor and variants of the cystic fibrosis transmembrane conductance regulator (CFTR) gene were also recognized as important risk factors for idiopathic chronic pancreatitis [18–22]. Finally, the p.G191R variant in the serine protease 2 (PRSS2) gene encoding anionic trypsinogen was shown to afford protection against chronic pancreatitis [23]. Taken together, the genetic studies indicate that chronic pancreatitis is a multigenic disease and the balance between risk and protective genetic factors determines susceptibility. The genetics of PRSS1, PRSS2, SPINK1 and CFTR mutations in chronic pancreatitis has been the subject of excellent reviews [2, 3, 14, 15, 24, 25].

The trypsin-dependent pathological pathway in chronic pancreatitis

Functional studies with mutant cationic trypsinogens demonstrated that the most frequently and consistently found phenotypic change was an increased propensity for trypsin-mediated trypsinogen activation, also referred to as autoactivation [26–30]. On the basis of these findings, we proposed that most PRSS1 variants are gain-of-function mutations which cause chronic pancreatitis by promoting premature trypsinogen activation in the pancreas. We and others showed that genetic variants in the SPINK1 gene are loss-of-function mutations which diminish expression of the inhibitor, either at the mRNA or at the protein level, thereby impairing its protective function [31–35]. Finally, in contrast to the pathogenic PRSS1 and SPINK1 mutations, we found that the p.G191R variant in PRSS2 results in rapid autodegradation of anionic trypsinogen and thereby affords protection against chronic pancreatitis [23]. Conceptually, the properties of p.G191R are noteworthy because they highlight the protective role of trypsinogen degradation against chronic pancreatitis. Taken together, the genetic and biochemical evidence defines a pathological pathway in which the imbalance between intrapancreatic trypsinogen activation, trypsinogen degradation and trypsin inhibition increases the risk for the development of chronic pancreatitis (Figure 1).

Figure 1.

The trypsin-dependent pathological pathway in chronic pancreatitis associated with genetic mutations. Activation of trypsinogen to active trypsin is mitigated by trypsinogen degradation and active trypsin is inhibited by pancreatic secretory trypsin inhibitor (SPINK1). Mutations in PRSS1 stimulate autoactivation of cationic trypsinogen. Loss-of-function mutations in SPINK1 reduce inhibitor expression and compromise trypsin inhibition. The p.G191R variant in PRSS2 stimulates trypsin-mediated degradation of anionic trypsinogen and thereby protects against chronic pancreatitis. Loss-of function mutations in CTRC reduce secretion or activity of chymotrypsin C and thus impair protective trypsinogen degradation.

CTRC: the latest susceptibility gene

In 2007, an international team of scientists reported that loss-of-function variants in the CTRC gene, which encodes the digestive proenzyme chymotrypsinogen C, are risk factors for chronic pancreatitis and this finding was replicated by an independent study published shortly thereafter [36, 37]. Screening of CTRC in subjects affected by chronic pancreatitis was stimulated by biochemical studies from our laboratory, which demonstrated that CTRC plays an important role in regulating trypsinogen activation and degradation. The initial genetic experiments took place at the University of Leipzig, Germany, where Niels Teich and Jonas Rosendahl used direct DNA sequencing to investigate 100 subjects with idiopathic and hereditary chronic pancreatitis and found variants in four subjects. The senior author of this review visited Leipzig in 2006 and still recalls the palpable excitement these initial observations elicited. The studies were later extended and a large number of collaborators became involved; now spearheaded by Jonas Rosendahl and Heiko Witt, who eventually convincingly verified the association of CTRC mutations with chronic pancreatitis [36]. Here we review the biochemical studies on the function of CTRC in digestive enzyme regulation; the genetic variants in CTRC; the functional effects of CTRC variants; and we discuss the potential mechanism of action of CTRC variants as risk factors for chronic pancreatitis.

Biochemical properties of CTRC

The CTRC gene (OMIM *601405) is located on chromosome 1p36.21 and comprises 8 exons spanning 8.2 kb. The human CTRC primary translation product (pre-chymotrypsinogen C) is composed of a secretory signal peptide of 16 amino acids, a propeptide (activation peptide) of 13 amino acids, and a chymotrypsin-like enzyme of 239 amino acids. CTRC is a digestive protease synthesized and secreted by the pancreatic acinar cells as an inactive proenzyme (zymogen), which becomes activated in the duodenum after tryptic cleavage of the Arg²⁹-Val³⁰ peptide bond at the C-terminal end of the propeptide. CTRC was first isolated from the pig pancreas and was found to cleave after Phe, Tyr, Leu, Met, Gln, and Asn amino acid residues, showing chymotrypsin-like substrate specificity with characteristically higher activity on leucyl peptide bonds both in synthetic and natural substrates [38–40]. On the other hand, human CTRC is found in the vicinity of the ELA2A and ELA2B genes and the extent of sequence identity between human CTRC and ELA2A is higher than between CTRC and CTRB1 or CTRB2.

In the pancreatic juice of ruminants, chymotrypsinogen C is found in ternary complex with procarboxypeptidase A (proCPA) and proproteinase E, or in binary complex with proCPA [41–43]. The crystal structure of the ternary complex has been determined [43]. The substrate specificity of the activated and chemically dissociated active bovine CTRC was similar to its porcine ortholog [44]. CTRC is almost certainly the same protein as caldecrin, a serum-calcium decreasing protein isolated from porcine and rat pancreas and later cloned from rat and human pancreas, although the identity of these two proteins has not been demonstrated formally [45–47]. Caldecrin was also found to inhibit osteoclast activation and bone resorption [48]. The protease activity of CTRC and its effect on calcium homeostasis appear to be distinct and unrelated functions, although both require activation by trypsin.

Physiological functions of CTRC in regulation of digestive enzyme activity

CTRC stimulates autoactivation of human cationic trypsinogen

The first indication that CTRC is not only a digestive enzyme but also plays a role in regulating the activity of other digestive enzymes came from the discovery that CTRC stimulates autoactivation of human cationic trypsinogen [49]. Activation of trypsinogen to trypsin involves the proteolytic removal of the trypsinogen activation peptide by cleavage of the Lys²³-Ile²⁴ peptide bond; a process physiologically catalyzed by the serine protease enteropeptidase in the duodenum. Trypsin can also cleave the trypsinogen activation peptide and this trypsin-mediated trypsinogen activation is termed autoactivation. Because human trypsinogens are prone to autoactivation and because hereditary pancreatitis-associated cationic trypsinogen mutations increase autoactivation, we proposed that autoactivation is a key pathological pathway in human chronic pancreatitis, the hereditary form in particular (Figure 1). We found that CTRC stimulates autoactivation of cationic trypsinogen through cleaving the Phe¹⁸-Asp¹⁹ peptide bond in the activation peptide, thereby excising the N-terminal tripeptide and processing the activation peptide to a shorter form (Figure 2). This action of CTRC is highly specific as other human pancreatic chymotrypsins (CTRB1, CTRB2, CTRL1) or elastases (ELA2A, ELA3A, ELA3B) do not digest the trypsinogen activation peptide. The shorter activation peptide is cleaved by trypsin more readily, resulting in approximately 3-fold increased autoactivation. The structural basis of this phenomenon lies in the disruption of an inhibitory interaction between cationic trypsin and the trypsinogen activation peptide [50]. Thus, Asp²¹⁸ on cationic trypsin participates in a repulsive electrostatic interaction with the negatively charged tetra-Asp motif of the activation peptide. This interaction inhibits autoactivation. Once the activation peptide is processed by CTRC, the inhibitory interaction with Asp²¹⁸ is partially relieved and autoactivation can proceed at a faster rate. Interestingly, Asp²¹⁸ is unique to human cationic trypsin, suggesting that a similar mechanism of autoactivation regulation does not exist in other vertebrates.

Figure 2.

Proteolytic regulation of activation and degradation of human cationic trypsinogen by chymotrypsin C (CTRC). The primary structure of cationic trypsinogen is shown with disulfide bonds indicated. CTRC stimulates autoactivation of cationic trypsinogen by cleaving the Phe¹⁸-Asp¹⁹ peptide bond and removing the N-terminal three amino-acids. The shortened activation peptide is more susceptible to trypsin-mediated cleavage (activation) which occurs at the Lys²³-Ile²⁴ peptide bond. CTRC cleaves the Leu⁸¹-Glu⁸² peptide bond in the calcium-binding loop of cationic trypsinogen and trypsin. Calcium binding to trypsin(ogen) protects against CTRC cleavage. Trypsin cleaves the Arg¹²²-Val¹²³ peptide bond, which allows dissociation of the yellow peptide segment not stabilized by disulfide bonds. CTRC cleavage sites are highlighted in red; trypsin cleavage sites are shown in blue and the catalytic residues are in green. Figure modified from Szmola and Sahin-Tóth (2007) [52].

CTRC-mediated stimulation of trypsinogen autoactivation may constitute a positive feed-back loop in the digestive enzyme activation cascade which facilitates full activation of trypsinogen in the gut. More importantly, the pancreatitis-associated cationic trypsinogen mutation p.A16V increases the rate of CTRC-mediated processing of the activation peptide 4-fold [51]. This observation suggests that p.A16V causes accelerated trypsinogen activation by this indirect mechanism, as opposed to other cationic trypsinogen mutations which directly stimulate autoactivation.

CTRC degrades trypsin: a physiological role in intestinal trypsin degradation

CTRC can trigger degradation of human cationic trypsin by selectively cleaving the Leu⁸¹-Glu⁸² peptide bond within the Ca²⁺ binding loop (Figure 2) [52]. Degradation and inactivation of cationic trypsin is then achieved through tryptic (autolytic) cleavage of the Arg¹²²-Val¹²³ peptide bond. The peptide segment between Glu⁸² and Arg¹²² is not stabilized by disulfide bonds and it becomes detached from the enzyme. Because the catalytically important Asp¹⁰⁷ amino-acid residue (Asp¹⁰² in classic chymotrypsin numbering) is located within this sequence, loss of trypsin activity can be explained by disruption of the catalytic triad. CTRC-induced inactivation of cationic trypsin is highly specific, as other human chymotrypsins (CTRB1, CTRB2, CTRL1) or elastases (ELA2A, ELA3A, ELA3B) do not cleave the Leu⁸¹-Glu⁸² peptide bond to a significant extent. Calcium protects cationic trypsin against CTRC-mediated degradation in a concentration-dependent manner, with a half maximal protective Ca²⁺ concentration of 40 µM. Since the relevant cleavage sites for CTRC-mediated trypsin degradation are conserved in human anionic trypsin and human mesotrypsin, as well as in the majority of mammalian trypsins, CTRC probably degrades these isoforms by a similar mechanism, but experimental confirmation of this is lacking.

A number of studies in humans have demonstrated that trypsin becomes inactivated during its intestinal transit and in the terminal ileum only about 20 % of the duodenal trypsin activity is detectable [53–55]. On the basis of in vitro experiments, a theory was put forth that digestive enzymes are generally resistant to each other and degradation only occurs via autolysis [56]. However, human cationic trypsin is highly resistant to autolytic inactivation because tryptic (autolytic) cleavage of the Arg¹²²-Val¹²³ peptide bond does not result in degradation or inactivation. Instead, due to trypsin-mediated re-synthesis of the peptide bond, a dynamic equilibrium is reached between the single-chain (intact) and double-chain (cleaved) forms, which are functionally equivalent [57]. The CTRC-dependent mechanism of trypsin degradation resolves the apparent contradiction between the in vivo documented intestinal trypsin degradation and the in vitro observed resistance of human cationic trypsin against autolysis, and strongly suggests that CTRC is responsible for the elimination of trypsin activity in the lower small intestine. In the duodenum and upper small intestine the millimolar calcium concentrations coming from the pancreatic juice and dietary intake should inhibit CTRC-mediated trypsin cleavage and normal digestion can proceed. As the Ca²⁺ concentration falls below millimolar in the lower intestine, trypsin degradation may prevail. Although intestinal Ca²⁺ absorption has been studied extensively [58], reliable data on the ionized Ca²⁺ concentrations along the small intestine is lacking. It is noteworthy that ionized Ca²⁺ concentrations in the gut are largely determined by luminal pH and insoluble complex formation, which becomes more significant at the alkaline pH of the lower intestine, where trypsin degradation has been shown to occur [59].

CTRC degrades trypsinogen: a defensive mechanism against pancreatitis

CTRC cleaves the Leu⁸¹-Glu⁸² peptide bond much faster in cationic trypsinogen than in cationic trypsin [52]. As described above for the inactivation of cationic trypsin, this cleavage per se does not result in trypsinogen degradation, which requires at least an additional cleavage by trypsin after Arg¹²². Cationic trypsinogen cleaved at the Leu⁸¹-Glu⁸² bond may be further digested by CTRC at a slow rate at the Leu⁴¹-Asn⁴² peptide bond. In contrast to cationic trypsinogen, CTRC cleaves human anionic trypsinogen and human mesotrypsinogen at multiple sites [52].

In 1986 and 1988 Heinrich Rinderknecht described an unidentified serine protease in human pancreatic juice with trypsinogen degrading activity, which he named enzyme Y [60, 61]. Rinderknecht initially alleged that mesotrypsin can degrade trypsinogens, but later he withdrew this claim and attributed the trypsinogen degrading activity to enzyme Y [60]. Rinderknecht believed that enzyme Y was probably a degradation fragment of cationic trypsin [61], perhaps complexed with pancreatic secretory trypsin inhibitor [62], although he acknowledged the possibility of contamination with an unknown protease [61]. Our findings indicated that CTRC is in fact enzyme Y, and, following Rinderknecht's theory, CTRC protects the pancreas by decreasing trypsinogen concentrations during inappropriate zymogen activation (Figure 1).

CTRC-mediated trypsinogen degradation is more likely to play a protective role against pancreatitis than CTRC-mediated trypsin inactivation, because the former reaction is much faster and trypsinogen concentrations are much higher. When trypsinogen autoactivation is measured in the presence of CTRC, the ultimate trypsin levels generated are lower and depend on the trypsinogen/CTRC ratio and the calcium concentration (unpublished observations). Thus, in essence, CTRC regulates trypsinogen autoactivation through degradation. The cationic trypsinogen mutation p.R122H, which causes hereditary chronic pancreatitis [13], eliminates the Arg¹²² autolytic cleavage site and thereby blocks CTRC-mediated trypsin and trypsinogen degradation. Autoactivation of p.R122H mutant trypsinogen in the presence of CTRC results in higher trypsin levels than those observed with wild-type cationic trypsinogen (unpublished observations), suggesting that mutation p.R122H exerts its pathogenic effect by interfering with the CTRC-dependent defense mechanism of trypsinogen degradation.

CTRC is a physiological co-activator of human procarboxypeptidases A1 and A2

Human digestive carboxypeptidases CPA1, CPA2 and CPB1 are secreted by the pancreas as inactive proenzymes containing a 94–96 amino-acid long propeptide [63]. Activation of procarboxypeptidases is initiated by proteolytic cleavage at the C-terminal end of the propeptide by trypsin. The trypsin-cleaved propeptide is still inhibitory, suppressing about 90% of carboxypeptidase activity. Recently, we demonstrated that CTRC induces a nearly 10-fold increase in the activity of trypsin-activated CPA1 and CPA2 [64]. CTRC exerts its effect by proteolyzing the α-helical connecting segment of the propeptide which becomes accessible only after tryptic activation. As a result, the propeptide dissociates and becomes completely degraded, resulting in full carboxypeptidase activity. CTRC cleaves the connecting segment in the propeptide at multiple sites but the critical cleavage appears to be at the conserved Leu⁹⁶-Leu⁹⁷ peptide bond. Other human pancreatic chymotrypsins (CTRB1, CTRB2, CTRL1) or elastases (ELA2A, ELA3A, ELA3B) are inactive or markedly less effective at promoting procarboxypeptidase activation. Taken together, these observations indicate that CTRC is an essential co-activator of proCPA1 and proCPA2.

Loss-of-function CTRC variants are risk factors for chronic pancreatitis

Primary chronic pancreatitis (idiopathic with or without family history)

As part of a large international collaboration we analyzed CTRC for potential mutations by direct DNA sequencing in 901 German patients with idiopathic or hereditary chronic pancreatitis, as well as in 2804 control individuals of German origin with no pancreatic morbidity [36]. We identified 11 missense and 2 deletion variants. The two most frequent variants, where disease association reached statistical significance, were c.760C>T (p.R254W) and c.738_761del24 (p.K247_R254del), both located in exon 7. The effect sizes of these mutations, as measured by the odds ratio (OR), were 3.3 and 11.5, respectively. The frequency of these variants in the patient population was 2.1% and 1.2%, respectively, indicating that these genetic risk factors contribute to the development of chronic pancreatitis in only a small fraction of cases. The 11 other rare CTRC variants were present in affected subjects and healthy controls with a total frequency of 1.3% and 0.82%, respectively. Because information is lacking about which variants might be pathogenic and which are just innocuous variations, an estimate cannot be drawn as to the risk conferred by rare CTRC variants.

A follow-up study by Masson et al. (2008) also found p.R254W and p.K247_R254del mutations in 5/287 (1.7%) and 2/287 (0.7%) French subjects affected by idiopathic, familial or hereditary chronic pancreatitis [37]. All carriers were detected within the 216 idiopathic cases and none in the 42 familial or 29 hereditary pancreatitis patients. The same variants were found among 350 healthy French subjects, each with a frequency of 0.3%. Disease association was statistically significant for the p.R254W variant (OR 6.1). The absence of these variants in the familial and hereditary groups stands in contrast to our study, where subgroup analysis did not show a significant difference between idiopathic and hereditary groups. In addition to these two variants, the study by Masson et al. (2008) found 17 other rare CTRC variants, including 8 missense mutations, one nonsense mutation, one promoter variant, five intronic variants and two variants in the 3' flanking region. These variants were identified almost exclusively in the patient group and their combined frequency was 7.7%. The high frequency of rare CTRC variants in chronic pancreatitis patients and their conspicuous absence among healthy controls differs from our own observations described above.

For the first time, Masson et al. (2008) also described two common synonymous CTRC polymorphisms, c.180C>T (p.G60=) and c.285C>T (p.D95=) with minor allele frequencies in the French control population of 11.9% and 4.3%, respectively [37]. Remarkably, a positive association was observed between the genotype CT of the c.180C>T variation and familial chronic pancreatitis (OR 2.5, relative to the CC genotype).

Secondary chronic pancreatitis associated with alcohol or hyperparathyroidism

The exon-7 p.R254W variant also showed statistically significant enrichment (OR 5.1) in 96 German subjects affected with alcohol-related chronic pancreatitis relative to 432 German subjects with alcohol-induced liver cirrhosis but without pancreatic disease [36]. Mutation p.K247_R254del was found in two subjects within the alcoholic pancreatitis group (0.6%), and in one control (0.2%). Four other rare CTRC variants were detected both in patients and controls with similar combined frequency (0.9%).

Felderbauer et al. (2010) analyzed an interesting subgroup of chronic pancreatitis patients, those with primary hyperparathyroidism, for CTRC mutations [65]. They found the p.R254W mutation in 2/31 (6.5%) German pancreatitis patients, while none of the 100 German controls with hyperparathyroidism but without pancreatitis carried the variant. Although the difference did not reach statistical significance, the finding is still strongly suggestive and is in agreement with previous studies documenting disease association for the p.R254W mutation. No other CTRC variants were reported in this study.

The observation that CTRC mutations contribute to the risk for secondary chronic pancreatitis is interesting, as other genetic risk factors are either absent (e.g. PRSS1 mutations) or exhibit much lower effects (e.g. SPINK1 mutations) in these diseases relative to primary chronic pancreatitis.

Table 1 demonstrates the combined dataset from the three studies in European populations. For this analysis, different disease etiologies were also pooled. Only two variants show statistically significant association, p.K247_R254del and p.R254W, with OR values of 7.1 and 3.9, respectively.

Table 1.

CTRC variants in subjects with chronic pancreatitis and healthy controls of European origin. The table compiles the results of three studies (Rosendahl et al. 2008; Masson et al. 2008; and Felderbauer et al. 2010). Rosendahl et al. sequenced all 8 exons in 621 patients with idiopathic chronic pancreatitis and 614 controls; exons 2, 3 and 7 in 280 patients with idiopathic chronic pancreatitis and exons 2 and 3 in 2075 controls and exon 7 in 2190 controls. Rosendahl et al. also sequenced all 8 exons in 96 patients with alcoholic chronic pancreatitis; and exons 2, 3 and 7 in 252 patients with alcoholic chronic pancreatitis and 432 controls with alcoholic liver disease. Note that Rosendahl et al. did not report promoter, intronic or 3' flanking region variants. Masson et al. sequenced all 8 exons in 287 patients and used direct sequencing or DHPLC in 350 controls. Felderbauer et al. only screened for mutations p.R254W and p.K247_R254del in 31 patients and 100 controls. CP, chronic pancreatitis, including idiopathic, hereditary, alcoholic and secondary to hyperparathyroidism forms; 3'-FR, 3' flanking region; OR, odds ratio, CI, confidence interval. The P value was calculated with two-tailed Fisher's exact test. Note that the common polymorphic variants c.180C>T (p.G60=) and c.285C>T (p.D95=) are not listed. Variants showing a statistically significant association with CP are highlighted in gray.

Location	Nucleotide change	Amino-acid change	CP	controls	OR	95% CI	P value
promoter	c.−59C>T		6/287 (2.09%)	2/350 (0.57%)
intron-1	c.40+24g>A		1/287 (0.35%)	0/350
intron-1	c.40+66g>A		1/287 (0.35%)	0/350
intron-1	c.41−50g>A		1/287 (0.35%)	0/350
exon-2	c.103g>C	p.D35H	0/1536	1/3471 (0.03%)
exon-2	c.103g>A	p.D35N	0/1536	1/3471 (0.03%)
exon-2	c.110g>A	p.R37Q	6/1536 (0.39%)	13/3471 (0.37%)
intron-2	c.133−19C>g		1/287 (0.35%)	0/350
exon-3	c.143A>g	p.Q48R	2/1536 (0.13%)	1/3471 (0.03%)
exon-3	c.164g>A	p.W55X	1/1536 (0.07%)	0/3471
exon-3	c.217g>A	p.A73T	1/1536 (0.07%)	0/3471
exon-4	c.308delg	p.G103VfsX31	1/1004 (0.10%)	0/964
exon-5	c.464g>A	p.C155Y	1/1004 (0.10%)	0/964
exon-5	c.485g>A	p.R162H	1/1004 (0.10%)	0/964
intron-5	c.494−10C>T		1/287 (0.35%)	0/350
exon-6	c.514A>g	p.K172E	1/1004 (0.10%)	1/964 (0.10%)
exon-6	c.598A>g	p.M200V	1/1004 (0.10%)	0/964
exon-7	c.649g>A	p.G217S	3/1536 (0.20%)	1/3586 (0.03%)
exon-7	c.649g>C	p.G217R	2/1536 (0.13%)	0/3586
exon-7	c.652g>A	p.G218S	0/1536	1/3586 (0.03%)
exon-7	c.659T>g	p.L220R	0/1536	1/3586 (0.03%)
exon-7	c.674A>C	p.E225A	0/1536	1/3586 (0.03%)
exon-7	c.703g>A	p.V235I	3/1536 (0.20%)	2/3586 (0.06%)
exon-7	c.738_761del24	p.K247_R254del	15/1567 (0.96%)	5/3686 (0.14%)	7.1	2.6–19.6	<0.0001
exon-7	c.746C>T	p.P249L	1/1536 (0.07%)	0/3586
exon-7	c.760C>T	p.R254W	34/1567 (2.17%)	21/3686 (0.57%)	3.9	2.2–6.7	<0.0001
3'-FR	c.807+83T>C		1/287 (0.35%)	0/350
3'-FR	c.807+86A>G		1/287 (0.35%)	0/350

Tropical chronic pancreatitis

In a pilot study, we demonstrated that CTRC variants were found in 71 individuals of Indian origin affected with tropical pancreatitis with much higher overall frequency (14.1%) than in 84 controls (1.2%) of Indian origin [36]. The relatively frequent c.217G>A (p.A73T) missense alteration was absent in 901 German patients and found only once among 287 French patients. Similarly, the c.190_193delATTG (p.I64LfsX69) frame-shift deletion was only observed in this Indian cohort. On the other hand, the p.K247_R254del variant was not found in the Indian population, and the enrichment of the p.R254W variant in subjects with tropical pancreatitis did not reach statistical significance.

In a follow up study by Derikx et al. (2009), 150 patients affected with tropical pancreatitis and 150 controls of Indian origin were investigated for CTRC mutations [66]. These authors also reported that the common polymorphic variant c.180C>T (p.G60=) was associated with chronic pancreatitis in Indians (OR 2.1), as described by Masson et al (2008) in a French population. Five other rare CTRC variants were found in patients and controls with an overall frequency of 6.8% and 4.1%, respectively. Among these, the p.A73T variant was found in 4/146 patients (2.7%) and in 1/144 (0.7%) controls, but the difference did not reach statistical significance.

The combined dataset from these two studies is shown in Table 2. Only the p.A73T variant shows statistically significant disease association, with an OR value of 8.7.

Table 2.

CTRC variants in subjects with tropical pancreatitis and healthy controls of Indian origin. The table compiles the results of two studies (Rosendahl et al., 2008; and Derikx et al. 2009). In both studies, all 8 exons were directly sequenced. CP, chronic pancreatitis; OR, odds ratio, CI, confidence interval. The P value was calculated with two-tailed Fisher's exact test. Note that the common polymorphic variants c.180C>T (p.G60=) and c.285C>T (p.D95=) are not listed. The p.A73T variant showing a statistically significant association with CP is highlighted in gray.

Location	Nucleotide change	Amino-acid change	CP	controls	OR	95% CI	P value
exon-1	c.−1C>T		1/199 (0.50%)	0/139
exon-3	c.181g>A	p.G61R	1/217 (0.46%)	0/228
exon-3	c.190_193delATTg	p.I64LfsX69	2/217 (0.92%)	0/228
exon-3	c.217g>A	p.A73T	8/217 (3.69%)	1/228 (0.44%)	8.7	1.1–70.1	0.018
exon-6	c.514A>g	p.K172E	1/220 (0.45%)	2/230 (0.87%)
exon-7	c.679g>A	p.G227S	0/219	1/234 (0.43%)
exon-7	c.703g>A	p.V235I	5/219 (2.28%)	2/234 (0.85%)
exon-7	c.760C>T	p.R254W	2/219 (0.91%)	1/234 (0.43%)
exon-7	c.778g>A	p.D260N	1/219 (0.46%)	0/234

Chronic pancreatitis in Taiwan

An analysis of 126 Chinese patients with chronic pancreatitis and 90 controls was reported by Chang et al. (2009) [67]. All study subjects were from Taiwan. Although this is a potentially important study to obtain insight into the role of CTRC variations in a different population, the experimental data showing very large enrichment of so far unknown CTRC variants in the patient population stands in stark contrast with all other published studies. In order to clarify the credibility of this extraordinary finding, we urge the authors to re-examine their data, and, if discrepancies are found, to publish a revised dataset.

Compound heterozygosity of CTRC variants with the SPINK1 p.N34S mutation

Chronic pancreatitis is a complex, multigenic disease, and affected individuals often carry mutations in several disease-associated genes. We found that among 30 German subjects with idiopathic or hereditary pancreatitis carrying a disease-associated CTRC variant, 9 also carried a heterozygous SPINK1 p.N34S mutation [36]. Interestingly, none of the subjects homozygous for SPINK1 p.N34S carried a CTRC variant. Compound heterozygosity was not detected in the control group. In the alcoholic pancreatitis group, one subject was compound heterozygous for CTRC p.K247_R254del and SPINK1 p.N34S mutations. Masson et al. (2008) described five patients with a CTRC variant and the p.N34S SPINK1 mutation [37]. One of these patients was also trans-heterozygous for the c.1A>T SPINK1 mutation, while another was homozygous for SPINK1 p.N34S. Felderbauer et al (2010) reported that between the two carriers of the p.R254W CTRC mutation with primary hyperparathyroidism, one also carried a heterozygous SPINK1 p.N34S mutation [65].

In our tropical pancreatitis cohort, 6 subjects were found to carry a CTRC variant and the p.N34S SPINK1 mutation [36]. In one case, trans-heterozygosity for two CTRC variants (p.A73T and p.D260N) together with the p.N34S SPINK1 mutation was observed. Again, homozygosity for SPINK1 p.N34S was never associated with a CTRC variant, and no CTRC-SPINK1 compound heterozygosity was detected in control subjects. Derikx et al. (2009) found that among the 10 subjects affected with tropical pancreatitis that carried a rare CTRC variant, two (one with p.G61R, and one with p.A73T CTRC mutation) were also heterozygous for SPINK1 p.N34S [66].

Copy number variations of the CTRC gene

Masson et al. (2008) found no copy number variations of the CTRC gene in 287 French patients with chronic pancreatitis [37].

Functional Effects of CTRC Mutations

Impaired secretion

We found that secretion of the p.K247_R254del and p.A73T mutants from transiently transfected human embryonic kidney (HEK) 293T cells was diminished (~5%) relative to wild-type CTRC, whereas cells expressing the p.R254W and p.Q48R variants exhibited reduced secretion at about 40% and 30% of wild type levels, respectively [36]. Derikx et al. (2009) reported that the p.G61R mutant was not secreted to a measurable extent from transfected HEK 293T cells [66]. The secretion defect caused by the p.A73T mutation was also observed in the AR42J rat acinar cell line transfected with recombinant adenovirus [68]. The frameshift mutations p.I64LfsX69 and p.G103VfsX31, as well as the nonsense mutation p.W55X, are also expected to result in complete loss of CTRC secretion, although direct experimental demonstration of this is lacking.

Loss of catalytic activity

Mutants p.K247_R254del and p.G217S were found catalytically inactive, whereas mutants p.Q48R and p.A73T exhibited measurable but decreased protease activity [36].

Increased propensity to elicit endoplasmic reticulum (ER) stress

We hypothesized that the reduced secretion of CTRC mutants occurred because of intracellular retention and degradation in the ER due to mutation-induced misfolding. If this were the case, the CTRC mutants might cause ER stress and trigger the unfolded protein response, a signal transduction pathway aimed at alleviating ER protein burden and increasing ER folding capacity [69]. Potentially harmful consequences of this signaling process are the activation of the inflammatory transcription factor nuclear factor kappa B (NF-κB) and the induction of apoptotic cell death. To test this hypothesis, we transfected dexamethasone-differentiated AR42J pancreatic acinar cells and freshly isolated mouse acini with recombinant adenovirus carrying the p.A73T CTRC mutant [68]. We found that the CTRC mutant p.A73T was intracellularly retained and degraded; and markers of ER stress (BiP expression and XBP1 splicing) were significantly elevated in cells expressing the p.A73T CTRC mutant relative to cells transfected with wild-type CTRC or a control adenovirus. Furthermore, we observed that AR42J cells underwent apoptotic cell death as a result of expressing the p.A73T CTRC mutant, whereas NF-κB activation was not detectable. Apoptosis was related to ER stress, as evidenced by increased expression of the pro-apoptotic transcription factor C/EBP-homologous protein (CHOP). These above experiments indicate that certain mutations, p.A73T in this case, can increase the ability of CTRC to cause ER stress and subsequent cell death by a mechanism which is unrelated to the trypsin-degrading activity of CTRC but involves mutation-induced misfolding. Extension of these studies to other CTRC mutants is necessary to test the general applicability of this mechanism.

How do CTRC mutations increase the risk of chronic pancreatitis?

Loss-of-function models

Taking into consideration the biochemical activities of CTRC and the functional properties of CTRC mutants, there are at least three, mutually not exclusive models that might explain why CTRC mutations increase the risk of chronic pancreatitis. These putative pathomechanistic pathways involve (i) impaired trypsinogen and/or trypsin degradation; (ii) impaired activation of A-type carboxypeptidases, and (iii) induction of ER stress. While the first two models consider loss of CTRC function as the disease-relevant phenotypic change, the ER stress model is actually a gain-of-function model, as discussed below.

Intuitively, the most plausible mechanism of action of CTRC mutations is through the “trypsin-dependent pathological pathway”, whereby loss of CTRC activity would impair the protective trypsinogen and/or trypsin-degrading activity of CTRC (Figure 1) [52]. Since trypsinogen degradation proceeds at a much higher rate than trypsin degradation, it appears more likely that CTRC curtails trypsinogen activation rather than clearing away active trypsin. The significance of trypsinogen degradation in protecting the pancreas against pancreatitis is also underscored by the protective effect of the p.G191R anionic trypsinogen (PRSS2) variant, which undergoes trypsin-induced degradation [23].

The physiological role of CTRC in promoting activation of proCPA1 and proCPA2 raises the possibility that loss of CTRC function increases pancreatitis risk through impaired carboxypeptidase activation [64]. This model would predict that loss-of-function mutations in the CPA1 or CPA2 genes should be also risk factors for chronic pancreatitis. Surprisingly, this seems to be the case, as newer, yet unpublished, studies from Heiko Witt’s laboratory indicate that CPA1 is a susceptibility gene for chronic pancreatitis, and loss of CPA1 function increases disease risk (personal communication). However, the mechanism through which reduced carboxypeptidase activity would promote pancreatitis development is not readily apparent yet.

ER stress: a gain-of-function model

The p.A73T mutation increases the propensity of CTRC to elicit ER stress, possibly through mutation-induced misfolding [68]. ER stress induced apoptosis can accelerate the loss of functional acini and contribute to exocrine insufficiency, a hallmark of chronic pancreatitis. These effects of the p.A73T mutant can be considered as gain of function, because the mutant CTRC protein triggers cellular signal transduction processes that result in acinar cell damage and increased risk of chronic pancreatitis. There are two caveats to this attractive model. First, more research is needed to clarify whether all disease-associated CTRC mutants can elicit ER stress, or whether this is a unique property of the p.A73T mutant. Second, it remains unclear whether CTRC expression levels in the human pancreas are high enough for mutant CTRC proteins to induce ER stress. Nevertheless, ER stress emerges as a potentially new paradigm for the mechanism of genetic risk in chronic pancreatitis [70].

The three mechanistic models described above reflect our current, rapidly expanding understanding of CTRC function and mutational effects thereon. The wealth of new information in this respect is a testimony to one of the fundamental benefits of human genetics; the stimulation of investigations into novel physiological functions and pathological pathways.