|Home | About | Journals | Submit | Contact Us | Français|
Ten years ago, the groundwork for the discovery of the genetic basis of chronic pancreatitis was laid by linkage analyses of large kindreds with autosomal dominant hereditary chronic pancreatitis. Subsequent candidate gene sequencing of the 7q35 chromosome region revealed a strong association of the c.365G>A (p.R122H) mutation of the PRSS1 gene encoding cationic trypsinogen with hereditary pancreatitis. In the following years, further mutations of this gene were discovered in patients with hereditary or idiopathic chronic pancreatitis. In vitro the mutations increase autocatalytic conversion of trypsinogen to active trypsin and thus probably cause premature, intrapancreatic trypsinogen activation in vivo. The clinical presentation is highly variable, but most affected mutation carriers have relatively mild disease. In this review, we summarize the current knowledge on trypsinogen mutations and their role in pancreatic diseases.
The human pancreas produces the digestive pro-enzyme trypsinogen in three highly similar isoforms. The isoenzymes are encoded by separate genes, PRSS1 (protease, serine, 1; OMIM +276000), PRSS2 (OMIM *601564) and PRSS3. On the basis of their relative isoelectric points and electrophoretic mobility, these are commonly referred to as cationic trypsinogen (PRSS1), anionic trypsinogen (PRSS2), and mesotrypsinogen (PRSS3). Normally, cationic trypsinogen represents circa 2/3 of total trypsinogen in the pancreatic juice, while anionic trypsinogen makes up approximately 1/3 (Guy et al., 1978; Rinderknecht et al., 1979, 1985). Mesotrypsinogen is a minor species, accounting for 2–10 % of human trypsinogens (Rinderknecht et al., 1984). Trypsinogens are synthesized as pre-pro-enzymes containing a signal peptide of 15 amino acids, followed by the 8 amino acid long pro-peptide, the trypsinogen activation peptide (TAP). The signal-peptide is removed upon entry into the endoplasmic reticulum lumen and the pro-enzymes (also called zymogens) are packaged into zymogen granules and eventually secreted into the pancreatic juice. Activation of trypsinogen to trypsin occurs in the duodenum by the brush-border localized protease, enteropeptidase (enterokinase), which removes the TAP. Trypsin also activates trypsinogen (autoactivation), which in the duodenum may have a physiological role in facilitating zymogen activation, whereas ectopic autoactivation in the pancreas may result in pancreatitis. Genetic variants of the cationic trypsinogen (PRSS1) have been identified in patients with hereditary pancreatitis (OMIM #167800), familial or sporadic chronic pancreatitis. In contrast, mutations of anionic trypsinogen (PRSS2) or mesotrypsinogen (PRSS3) have not been found in association with chronic pancreatitis to date (Chen et al., 1999a; Nemoda et al., 2005). Classic hereditary pancreatitis follows an autosomal dominant inheritance pattern with incomplete penetrance and highly variable disease expression (Comfort and Steinberg, 1952; Le Bodic et al., 1996a; Sossenheimer et al., 1997; Keim et al., 2001; Howes et al., 2004). The mutations are located in three clusters within the trypsinogen sequence: in the TAP, in the N-terminal part of trypsin or in the longest peptide segment not stabilized by disulfide bonds between Cys64 and Cys139 (fig. 1). Thus, all pancreatitis-associated mutations discovered to date seem to cluster in the N-terminal half of the molecule encoded by exons 2 and 3. It is important to note, however, that investigation of the PRSS1 gene in patients with suspected genetically determined chronic pancreatitis is restricted to these exons in most laboratories and possible C-terminal mutations may have been missed. The discovery of pancreatitis-associated cationic trypsinogen mutations in 1996 demonstrated that trypsinogen plays a central role in the pathogenesis of human pancreatitis (Whitcomb et al., 1996a). Here we review the wealth of genetic and biochemical data that has accumulated over the past ten years of research on the role of trypsinogen mutations in chronic pancreatitis.
After genome wide linkage analyses, three independent groups reported an association of the hereditary pancreatitis phenotype with the long arm of chromosome 7 (Le Bodic et al., 1996b; Pandya et al., 1996; Whitcomb et al., 1996b). Within the same year, the c.365G>A mutation (p.R122H; p.R117H in the chymotrypsin numbering system) of the cationic trypsinogen gene was identified in all hereditary pancreatitis affected individuals and obligate carriers from five kindreds, but not in individuals who married into the families nor in 140 unrelated individuals (Whitcomb et al., 1996a). In the early literature, the mutation nomenclature was based on the chymotrypsin numbering commonly used by biochemists and crystallographers. The genetic numbering system, which designates the initiator methionine as position 1, was widely adopted in 2000 and has been in use since (Chen and Férec, 2000a; Teich, Hoffmeister and Keim, 2000).
To explain why the R122H mutation might cause pancreatitis, David Whitcomb proposed that the Arg122-Val123 autolytic peptide bond in trypsin plays an important role in the degradation of prematurely activated trypsin in the pancreas. Destruction of this “failsafe mechanism” by the R122H mutation would increase intrapancreatic trypsin activity and disturb the protease-antiprotease equilibrium and eventually precipitate pancreatitis (Whitcomb et al., 1996a). Biochemical evidence supports the notion that Arg122 is important for autolysis of trypsin and mutations of this amino-acid result in increased trypsin stability (Higaki and Light 1986; Gaboriaud et al., 1996; Várallyai et al., 1998, Sahin-Tóth and Tóth, 2000). A study using cerulein-induced zymogen activation in isolated rat acini demonstrated that autodegradation of trypsin mitigates cathepsin B-mediated trypsinogen activation, suggesting that a failsafe mechanism might be indeed operational in the mammalian pancreas (Halangk et al., 2000). The Whitcomb model in its original form has remained very popular over the years; even though more detailed biochemical analysis indicated that the R122H mutation results not only in increased trypsin stability but also in increased zymogen stability and increased autoactivation (Sahin-Tóth and Tóth, 2000, Kukor et al., 2002b). A weak trypsin-inhibitory activity associated with the Arg122 site is also lost in the R122H mutant (Kukor et al., 2002b). Thus, the pleiotropic biochemical effect of R122H raises the possibility that the pathogenic alteration is unrelated to trypsin stability. More importantly, the model fails to explain how the other pancreatitis-associated PRSS1 mutations might work, as the majority of these do not affect trypsin stability.
A few months after the R122H mutation was reported, a second hereditary pancreatitis associated mutation was identified within the N-terminal region of the cationic trypsinogen molecule. The c.86A>T mutation (p.N29I; p.N21I according to chymotrypsin numbering) was initially found in 4 families, 2 from the USA and 2 from Germany (Gorry et al., 1997; Teich, Mössner and Keim, 1998). A different c.86A>C mutation affecting the same amino-acid (p.N29T) was also reported in 2002 (Pfützer et al., 2002). Biochemical characterization of the N29I mutation using recombinant trypsinogen found no effect on trypsin or trypsinogen stability. On the other hand, moderately increased autoactivation was observed by two independent laboratories in 4 published studies (Sahin-Tóth 2000; Sahin-Tóth and Tóth 2000; Szilágyi et al., 2001; Teich et al., 2005). The N29T mutant exhibited a phenotype similar to that of R122H, both increased trypsin stability and enhanced autoactivation were documented (Sahin-Tóth 2000). Because increased autoactivation was observed with the R122H, N29I and N29T mutations, whereas N29I had no effect on trypsin stability, the logical conclusion was put forth that enhanced autoactivation is the common pathogenic mechanism of hereditary pancreatitis associated PRSS1 mutations (Sahin-Tóth and Tóth, 2000).
The R122H and the N29I mutations are the most common PRSS1 mutations worldwide. They have been frequently reported from Europe, North America and Asia (Nishimori et al., 2001) and R122H was also recently found in a family of Aboriginal descent in Australia (McGaughran et al., 2004). Neither mutation was detected in two hereditary pancreatitis families from Brazil (Bernardino et al., 2002) and no hereditary pancreatitis cases have been reported from Africa. So far, no trypsinogen mutations have been found in animals. Curiously, there is a relatively high incidence of pancreatitis in miniature Schnauzer dogs. Bishop and colleagues sequenced all exons of the canine cationic trypsinogen gene in four Schnauzers but no mutation was found (Bishop et al., 2004).
In addition to the originally reported frequently found R122H mutation (c.365G>A), a rare double nucleotide substitution variant was also found (c.365–366 GC>AT), which also results in the R122H mutation at the amino-acid level. The routinely used Afl-III restriction digestion assay for the PRSS1 R122H mutation does not detect this variant, therefore, DNA-sequencing is recommended to screen patients from hereditary pancreatitis families (Chen et al., 2000b; Howes et al., 2001). The R122H double nucleotide substitution variant was the first convincing example of a possible gene conversion event leading to a pancreatitis-causing trypsinogen mutation (see below).
A third mutation of Arg122, c.364C>T resulting in p.R122C, was discovered in 2001 by three independent groups (LeMaréchal et al., 2001; Pfützer et al., 2002; Simon et al., 2002a) and was also found in one patient from our cohort. The recombinantly expressed R122C mutant showed a complex phenotype. PRSS1 contains 10 cysteine residues, which form 5 disulfide bridges in the correctly folded protease. The introduction of the extra unpaired cysteine in the R122C mutant results in partial misfolding during expression and purification of the recombinant protein. As a result, the R122C preparation exhibits reduced activity, autolysis and autoactivation. On the other hand, when the data are normalized to the active trypsinogen fraction, the phenotype becomes very similar to that of R122H, exhibiting reduced autolysis (i.e. increased trypsin stability) and increased autoactivation (Simon et al., 2002a). The low-activity phenotype of R122C spawned the theory that a loss of trypsin function caused by a mutation could cause pancreatitis by eliminating a protective trypsin-dependent mechanism. This model has been favored by Markus Lerch’s group and received additional support from their studies indicating that trypsin might play a protective role in cerulein-induced zymogen activation (Halangk et al., 2000). Although this model is thought-provoking, from a genetic point-of-view no evidence exists to support it; as so far no clear loss-of-function mutations (e.g. nonsense, splice-site or frameshift) have been found in association with chronic pancreatitis. Furthermore, results from the large majority of in vitro experiments using recombinantly expressed trypsinogen mutants favor enhanced trypsinogen activation as the key event in the pathogenesis of genetically determined chronic pancreatitis.
In addition to Arg122, mutations were also identified in the neighboring residues, Ala121 and Val123. The c.367G>A mutation, which results in p.V123M was found in a patient with chronic pancreatitis by Jian-Min Chen and Claude Férec and their co-workers (Chen et al., 2001). The c.361G>A mutation, which results in p.A121T was detected in association with hereditary pancreatitis in a German family (Felderbauer et al., 2005). Also in the vicinity of Arg122, the c.346C>T mutation resulting in p.R116C was reported by four research groups. Arg116 is unique to human cationic trypsinogen, however, its function is unclear. Unlike Arg122, it does not appear to be important for autolysis (M. S.-T., unpublished observation). The mechanism of the “Arg122-flanking” mutations described above has not been studied yet.
Taken together, the high frequency and variability of mutations in or close to Arg122 suggests that this sequence is particularly prone to mutations, for reasons that are not readily apparent. The “mutational hot-spot” idea is consistent with the worldwide distribution of R122H and and it is also supported by the direct identification of a “de novo” R122H mutation in a German patient (Simon et al., 2002b).
Gene conversion – the substitution of genetic material from another gene – is recognized as the underlying cause of a growing number of genetic diseases, as Gaucher disease (OMIM #230800), von Willebrand disease (OMIM #193400), polycystic kidney disease (OMIM #601313) and Shwachman-Diamond syndrome (OMIM #260400). In the majority of cases, the donor gene is a duplicated pseudogene, which accumulated mutations over time and recombination of the mutated sequences with the normal gene results in the pathogenic genotype. In contrast, there is growing evidence, that gene conversion between two functional paralogous trypsinogen genes can occur and cause genetically determined chronic pancreatitis. Trypsinogen genes are tandemly repeated within the T-cell-receptor beta (TCR-beta) locus (Rowen, Koop and Hood, 1996). This is a hot spot for gene conversion events to generate a broad variety of TCR-beta genes (Flajnik, 2002). Therefore, conversion mutations within the interpolated trypsinogen gene family are very likely. This mechanism is evident in case of the R122H mutation associated with the c.365–366 GC>AT di-nucleotide substitution (c.365_366conNM_002770.2:c.365_366), but remains speculative for the single-nucleotide substitutions causing mutations R122H (c.365G>A), N29I (c.86A>T) and A16V (c.47C>T) (Chen and Férec, 2000b; Chen and Férec, 2000c; Chen et al., 2000). Recently, we described the clinical, genetic and biochemical properties of a large gene conversion event between the anionic trypsinogen gene (donor gene) and the cationic trypsinogen gene (acceptor gene) in a six-year old girl with chronic pancreatitis. In this patient, the conversion created 22 heterozygous nucleotide changes affecting mostly exon 2 of the PRSS1 gene (c.41-34_c.200+236conNM_002770.2:c.41-34_c.200+236). This resulted in the N29I and N54S mutations at the amino acid level. Biochemical characterization of the N29I–N54S double mutant revealed no difference from the previously investigated N29I mutant, indicating that the N54S mutation has no additional deleterious effect. This finding provided proof for the concept, that gene conversion is a possible mechanism for the development of pancreatitis associated PRSS1 mutations (Teich et al., 2005).
In 1999, the c.47C>T mutation, which results in p.A16V was reported in 4 out of 44 pediatric patients with chronic pancreatitis (Witt, Luck and Becker, 1999). This transition gained particular interest, as it was the first indication, that trypsinogen mutations may be associated with idiopathic chronic pancreatitis. Initially, the mutation was reported in 10 percent of chronic pancreatitis patients without a family history. In subsequent European and North American studies, only a minority of patients with idiopathic chronic pancreatitis carried the PRSS1-A16V mutation (Chen et al., 1999b; Pfützer and Whitcomb, 1999; Chen et al., 2001; Howes et al., 2001; Truninger et al., 2001; Teich et al., 2002; Howes et al., 2004). This discrepancy might be explained by the strict selection of pediatric patients in the initial report, which decreased the effect of environmental risk factors of chronic pancreatitis to a large extent (Witt, Luck and Becker, 1999). Even though it has been reported in less than 20 families worldwide, the A16V mutation represents the 3rd most common cationic trypsinogen gene mutation. In contrast to the considerably more prevalent N29I and R122H mutations, A16V was almost exclusively detected in patients without a family history of chronic pancreatitis. Unfortunately, thorough clinical and functional characterization of this mutation is lacking so far. The A16V mutation affects the first amino acid of the activation peptide and thus forms part of the signal peptide cleavage site. It was proposed that enhanced trypsinogen activation or altered intracellular transport might be the relevant pathogenic phenotype. The reduced cellular viability of AR42J cells transfected with A16V-trypsinogen might support either model (Gaiser et al., 2005). On the other hand, a recent study using recombinant human cationic trypsinogen failed to demonstrate increased autoactivation in the A16V mutant (Király et al., 2006). Furthermore, secretion is typically not influenced by the amino-acid distal to the signal-peptide cleavage site and valine is frequently found in this position in other secretory proteins, including human mesotrypsinogen.
Other genetic variants of the PRSS1 gene have been found in families with suspected hereditary pancreatitis, or in chronic pancreatitis patients without a family history (table 1). Some have been reported only once (−28delTCC, D19A, D22G, K23R, N29I+N54S, P36R, V39A, G83E, K92N, D100H, L104P, A121T, V123M, C139F) or in only a few families (N29T, E79K, R116C, R122C). Despite their limited frequency, the biochemical properties caused by some of these mutations offered interesting insights into the pathophysiology of genetically-determined chronic pancreatitis. The most interesting findings will be reviewed here.
The trypsinogen activation peptide (TAP) is hydrolyzed as the first step of the trypsinogen activation process. In vertebrates, this typically 8–10 amino-acid long peptide contains an unusual and highly conserved tetra-aspartate sequence, which precedes the scissile peptide bond Lys23-Ile24 (figure 1). This acidic motif serves as a specific recognition site for enteropeptidase, the physiological activator of trypsinogen. In addition, the acidic stretch was shown to contribute to the suppression of autoactivation. Interestingly, a recent study found that in human cationic trypsinogen the tetra-aspartate motif per se is not required for enteropeptidase recognition, whereas it is essential for maximal inhibition of autoactivation (Nemoda and Sahin-Tóth, 2005). In addition to the A16V mutation (see above), 3 pancreatitis-associated TAP mutations have been detected to date. The c.68A>G and c.65A>G mutations (p.K23R and p.D22G, respectively) were identified in families with putative hereditary pancreatitis, while the c.56A>C (p.D19A) mutation was found in a patient with idiopathic chronic pancreatitis (Férec et al., 1999; Teich et al., 2000; Chen et al., 2003a). When analyzed using model peptides or recombinant trypsinogens, all 3 activation peptide mutations significantly increased autoactivation of cationic trypsinogen; particularly the D22G and K23R mutations, which lie closer to the trypsinogen activation site Lys23-Ile24 (Teich et al., 2000; Chen et al., 2003a). These data are consistent with earlier peptide studies showing that neutralization of any of the 4 aspartate residues in the activation peptide increases autoactivation of trypsinogen (Delaage et al., 1967; Abita, Delaage and Lazdunski, 1969; Radhakrishnan, Walsh and Neurath, 1969). This notion was also confirmed recently by Ala-scanning mutagenesis of the tetra-Asp motif in human cationic trypsinogen (Nemoda and Sahin-Tóth, 2005). The markedly increased autoactivation of the K23R mutant is explained by the preference of trypsin to cleave after Arg rather than Lys at the P1 position of substrates.
Conceptually, the activation peptide mutations are important, because they do not affect trypsin structure or function, indicating that pancreatitis-associated mutations exert their effect through altering the properties of the pro-enzyme trypsinogen and not the active enzyme trypsin. In this context, properties of the characterized mutants confirmed that increased autoactivation of cationic trypsinogen is a relevant pathomechanism in hereditary pancreatitis.
In one German family with putative hereditary pancreatitis and two patients with sporadic chronic pancreatitis from France (Chen et al., 2001) and Belgium, we detected the c.235G>A mutation, which results in p.E79K (Teich et al., 2004a). This mutation was also reported in a patient with alcoholic chronic pancreatitis and one healthy control from Brasil as well as 2 healthy subjects from France (Chen et al., 2001; Bernardino et al., 2003). The affected individuals in the German family exhibited late-onset chronic pancreatitis, suggesting that this mutation might have a relatively low penetrance (Teich et al., 2004a). The wild type and E79K mutant PRSS1 alleles are transcribed in the pancreas to comparable levels. In vitro analysis of recombinant cationic trypsinogen carrying the E79K mutation revealed unaltered catalytic activity, autolysis, and inhibition by pancreatic secretory trypsin inhibitor. In contrast to previously studied mutations that increased autoactivation of cationic trypsinogen, autoactivation of E79K-trypsinogen was markedly decreased. Instead, E79K-trypsin proved to be an at least 2-fold better activator of anionic trypsinogen than wild type cationic trypsin. Thus, the E79K mutation appears to conform to the common mechanism of pancreatitis-associated mutations insofar as it leads to increased trypsinogen activation. What sets E79K apart from the other mutations is that increased trypsinogen activation is the result of transactivation of anionic trypsinogen instead of autoactivation (Teich et al., 2004a). However, it remains controversial whether the decreased autoactivation or the increased transactivation is the disease-relevant phenotype of E79K.
Chen et al described the heterozygous nonsense mutation c.111C>A (p.Y37X) and the splicing site mutation IVS2+1G>A in the PRSS1 gene of alcoholics without the development of chronic pancreatitis, but not in alcoholics with chronic pancreatitis and patients with hereditary or idiopathic chronic pancreatitis. They proposed that such "loss of function" mutations might protect against chronic pancreatitis in persons with increased pancreatitis risk (Chen et al., 2003b).
Autosomal dominant hereditary chronic pancreatitis was first described after long-term follow-up of 36 members from 4 generations of a family, in which 6 members were affected (Comfort and Steinberg, 1952). The investigation of such large families was the most important prerequisite for successful genetic linkage studies, which have localized the hereditary pancreatitis gene on chromosome 7 (Le Bodic et al., 1996b; Pandya et al., 1996; Whitcomb et al., 1996b). However, in the daily clinical setting establishing such extensive pedigrees is not feasible and the inheritance pattern cannot be determined in some cases. Consequently, the criteria of hereditary pancreatitis have been changing over the years and may be somewhat different in the various clinical centers. In the recently published Europac study, the diagnosis of hereditary pancreatitis was made on the basis of two first degree relatives or three or more second degree relatives, in two or more generations with recurrent acute pancreatitis, and/or chronic pancreatitis for which there were no precipitating factors. Cases in which these strict criteria were not met, but more than one affected family members were identified, mostly within the same generation, were classified as familial chronic pancreatitis. The diagnostic value of this classification is questionable, however, and at our center we diagnose hereditary pancreatitis if the patient has no other detectable cause of chronic pancreatitis and if he/she has one first or second degree relative with proven chronic pancreatitis. Furthermore, because the post-hoc analysis reveals a positive or putative family history of chronic pancreatitis in almost all "idiopathic chronic pancreatitis" patients with the common R122H or N29I mutations (Creighton et al., 1999, Teich, Hoffmeister and Keim, 1999), we classify all chronic pancreatitis patients with these mutations as highly susceptive of hereditary pancreatitis, even if a family history is not readily available. Based on Jane Creighton’s data, that a susceptive family history for chronic pancreatitis is found post-hoc in almost all patients with typical hereditary pancreatitis-associated PRSS1 mutations, one may assume that the presence of a mutation compels clinicians to obtain a full family history post-hoc, whereas in the absence of such a mutation no post-hoc analysis of the family history is done. This may misclassify chronic pancreatitis in patients with hereditary pancreatitis as “idiopathic”. Therefore, exact evaluation of anamnestic and clinical data has to precede genetic testing to avoid misclassification and unnecessary molecular investigations. In most patients with a rare or unique PRSS1 mutation, the family size is small or not reported (Chen et al., 2001; Teich et al., 2002) and the inheritance pattern of chronic pancreatitis associated with these mutations remains uncertain (table 1).
Idiopathic chronic pancreatitis has been a traditional clinical diagnosis describing the lack of an identifiable cause behind the disease. Clearly, as more and more genes (PRSS1, SPINK1, CFTR) and/or mutations are identified that cause or predispose to chronic pancreatitis, the number of patients with “true idiopathic” or “familial idiopathic” (Threadgold et al., 2002) disease keeps shrinking. As recent data suggest that chronic pancreatitis may be inherited in an autosomal dominant, autosomal recessive or multigenic fashion, the differentiation between hereditary and idiopathic chronic pancreatitis also becomes obsolete.
Hereditary pancreatitis is associated with PRSS1 mutations in more than 50 percent of affected families. However, in hereditary pancreatitis families with the PRSS1 mutations N29I or R122H, only 80% of the mutation carriers suffer from chronic pancreatitis. The cause of this incomplete penetrance is uncertain. In addition to further genetic factors, smoking, alcohol consumption or the lack of antioxidants were assumed as manifestation factors (Sibert 1978; Sossenheimer et al., 1997; Amann et al., 2001; Keim et al., 2001). On the basis of a study of 60 affected and 25 non-affected persons carrying the PRSS1 mutations N29I or R122H, we suggested that the TNF-238A promotor variant could be a relevant manifestation factor for chronic pancreatitis in hereditary pancreatitis-families (Beranek et al., 2003).
As to the clinical picture, both alcoholic and PRSS1 mutation-associated chronic pancreatitis exhibit essentially identical clinical laboratory results, histopathology or morphological changes in imaging studies. On the other hand, PRSS1 mutation associated pancreatitis manifests typically at an earlier age and pancreatic calcification and diabetes are less frequent complications (Keim et al., 2003). In the first large investigation of clinical criteria in hereditary pancreatitis, the mean age of onset was 12.9 years (Sossenheimer et al., 1997). Our own investigations revealed no difference in the age of onset between carriers of the N29I or R122H PRSS1 mutations. The median was 13 years in each group, respectively. Only 4% of our patients had severe chronic pancreatitis with exocrine and endocrine insufficiency, pancreatic calcification and duct dilation as well as hospitalizations due to pancreatitis. In general, half of the mutation carriers had little or no complaints or complications (Keim et al., 2001). A recent European study revealed a mean age of onset of 10 years and 14.5 years for affected carriers of the PRSS1 mutations R122H and N29I, respectively (Howes et al., 2004), but showed no mutation-dependent differences in complications such as exocrine or endocrine insufficiency or increased pancreatic cancer risk.
As shown in an investigation of 8 patients with pancreatic cancer in a cohort of 246 hereditary pancreatitis patients, the lifetime risk of pancreatic cancer is about 50-fold higher than in the control population and corresponds with 1 per 1066 person-years. It is only 20-fold elevated in patients with chronic alcoholic pancreatitis (Lowenfels et al., 1997; Malka et al., 2002). In our cohort of 101 hereditary pancreatitis patients, pancreatic cancer was diagnosed in 3 patients with the R122H mutation with a median of 23 years after the onset of pancreatitis. This corresponds to a rate of about 1 per 1200 person-years among affected R122H-carriers (Keim et al., 2001). However, the data basis for the estimation of the pancreatic cancer risk in patients with PRSS1-associated hereditary pancreatitis is small. The largest clinical investigation in the pre-genetic era, however, revealed no pancreatic cancer in 72 patients from 7 families (Sibert 1978). Taken together, the pancreatic cancer risk in hereditary pancreatitis patients with a PRSS1 mutation seems to be elevated - with an uncertain relative risk increase. Today, there is no generally accepted protocol for screening hereditary pancreatitis patients for early pancreatic cancer. However, affected mutation carriers should be strongly advised to stop smoking, as it is an additional risk factor for pancreatic cancer (Lowenfels et al., 2000).
Since the first description of a trypsinogen mutation in hereditary pancreatitis (Whitcomb et al., 1996a), the experimental and clinical information on genetic alterations in chronic pancreatitis has been rapidly growing resulting in a more and more complex data set. To address this issue, in early 2001 we implemented a continuously updated database, which contains all genetic variants of the PRSS1 and SPINK1 genes (www.uni-leipzig.de/pancreasmutation). In addition to exact genetic data, this database contains PubMed links to the clinical characterization of patients with different mutations and to in vitro studies with mutant molecules. Since 2002, this database has been part of the integrated database project of the Human Genome Variation Society (HGVS). This ambitious project fosters the discovery and characterization of genomic variations including population distribution and phenotypic associations. Backed by the HGVS, the novel WayStation has been inaugurated recently (www.centralmutations.org, Teich et al., 2004b). This is a web-accessible centralized structure for the submission, peer review, publication and release of genetic variation data.
The large majority of PRSS1 variants associated with chronic pancreatitis have been catalogued and the likelihood of finding novel PRSS1 mutations that affect a significant number of patients is diminishing. Nonetheless, as most laboratories have focused their studies on exons 2 and 3, it is still possible that new variants will be identified in exons 1, 3 and 5, or in the intronic and promoter regions. From a genetic aspect, a more interesting quest in the near future will be for modifier genes that might explain the penetrance problem, i.e. why some carriers of PRSS1 mutations remain healthy, whereas their relatives with the same mutation exhibit severe disease. Despite intensive research, the disease mechanism remains poorly understood. The biochemical alterations caused by the mutations have been mostly clarified, but the significance of these phenotypic changes in the disease remains to be determined. In this respect, future development of cellular and animal models of hereditary pancreatitis will be particularly valuable.
The authors thank Cèdric Le Maréchal for providing the nationalities of patients with rare PRSS1 mutations, as reported in table 1.