CATIONIC TRYPSINOGEN GENE MUTATIONS

The cationic trypsinogen gene (UniGene name: protease, serine 1; PRSS1) was identified as the first pancreatitis-specific susceptibility gene through genetic linkage studies in families with hereditary pancreatitis.⁹ Cationic trypsinogen is the major form of trypsinogen (≈65%) followed by anionic trypsinogen (PRSS2, ≈30%), and mesotrypsin (PRSS3, ≈5%). The trypsin molecule is formed by a single peptide that folds into an enzyme with an active site between two globular domains linked by a single connecting chain. An eight amino acid extension of the enzyme, called the trypsinogen activation peptide (TAP), maintains the enzyme as inactive trypsinogen until it is cleaved by enterokinase or another trypsin molecule (autoactivation). Cleavage of TAP allows a conformation change that activates trypsin. Trypsin is also susceptible to trypsin-mediated autolysis beginning at the arginine 122 (R122) site of the connecting chain. The connecting chain is further degraded by chymotrypsin C (CTRC),¹⁶ which was biochemically characterized as enzyme Y.¹⁷ Finally, the trypsinogen molecule also has two separate calcium binding pockets that play key roles in trypsin regulation by exposing the activation site and blocking the autolysis site, respectively.

Trypsin plays a critical role in pancreatic physiology as the activator of the other pancreatic zymogens, a process that normally occurs within the duodenum where the zymogen activation cascade is initiated by the enterokinase-stimulated conversion of trypsinogen to trypsin. Trypsinogen activation and trypsin inactivation are primarily controlled by trypsin (autoactivation and autolysis), and the ambient calcium concentration serves as the switch between on and off (Fig. 57-2).⁸ Calcium binding to the first binding pocket that is formed by four aspartic acids within the TAP portion of trypsinogen facilitates trypsinogen activation by trypsin. Calcium binding to the second calcium binding pocket formed by a peptide loop in both trypsinogen and trypsin located adjacent to the autolysis loop, prevents exposure of the trypsin-sensitive R122 autolysis site and thereby prevents autolysis. Thus, physiologic regulation of trypsin activity is determined by cellular calcium, with increased cellular calcium facilitating activation and preventing inactivation and low cellular calcium levels limiting activation and permitting autolysis.

Figure 57-2. X-ray crystallography-based model of cationic trypsinogen (PRSS1) and pancreatic secretory trypsin inhibitor (SPINK1). The trypsinogen molecule contains two globular domains (blue and yellow) joined by a connecting side chain (see top of drawing). Trypsinogen is activated to trypsin with cleavage of trypsinogen activation peptide (TAP), allowing a three-dimensional conformational change, opening of the specificity pocket (S), and high-efficiency enzyme activity at the active site (*). The location of the two major PRSS1 mutations (N29, R112) associated with hereditary pancreatitis are illustrated. Note the location of R112 in the side chain connecting the two (blue and yellow) globular domains. The SPINK1 molecular is shown bound to trypsin. The location of the major SPINK1 mutation associated with idiopathic and familial pancreatitis, N34, is illustrated.

(Figure courtesy of Drs. Andrew Brunskill and William Furey.)

The maintenance of low calcium concentrations within the acinar cells is critical to protecting them from premature trypsinogen activation. However, other protective mechanisms are used to limit trypsinogen activation within the high calcium concentrations of the pancreatic duct fluid. Acinar cell calcium can rise through neurohormonal hyperstimulation (which opens basolateral calcium channels and is linked to calcium tunnels transporting calcium to the acinar pole)^18,19; high extracellular calcium concentrations and submaximal pancreatic stimulation²⁰; bile acid reflux, which opens apical membrane calcium pathways²¹; and prolonged, high-dose alcohol consumption, which lowers the threshold for stimulation-induced acute pancreatitis,²² possibly through mitochondrial damage²³ and other factors that regulate intracellular calcium.²⁴ Any process that increases acinar cell calcium will predispose to acute pancreatitis through a calcium-dependent trypsinogen activation and stabilization mechanism.²⁴

More than 20 mutations have been identified in PRSS1 that increase susceptibility to recurrent acute pancreatitis,²⁵ although the R122H and N29I mutations are most common. The locations of the R122H and N29I (N21I) mutations are shown in relationship to the active site in Figure 57-2 and in a mechanistic model in Figure 57-3. The mutations are clustered in regions associated with calcium-dependent trypsin regulation and may confer “gain-of-function” features by facilitating trypsinogen activation or retarding trypsinogen inactivation independent of cell calcium. Gain-of-function mutations often result in an autosomal dominant inheritance pattern; only one of the two trypsinogen alleles must code for a super-functional trypsin in order to prematurely trigger the zymogen activation cascade and cause pancreatitis, thus manifesting the phenotype. Other trypsinogen mutations that are unrelated to calcium-dependent trypsin regulation may predispose to recurrent pancreatitis by altering the activation or inactivation process normally regulated by pH or through interaction with other molecules,²⁶ but the clinical relevance of these potential types of trypsinogen variants remains an area of investigation. The fact that the trypsinogen molecule has two calcium “switch” sites may explain why pancreatitis occurs only intermittently. The clinically important trypsinogen mutations and polymorphisms are discussed following.

Figure 57-3. Diagram of trypsin control mechanisms in the pancreas (green arrows are positive and red lines are inhibitory influences). Trypsinogen (far left) is protected from activation within the zymogen granules. Trypsinogen activation (arrow to trypsin) is supported by elevated calcium (Ca⁺⁺) and trypsin autoactivation (green dashed arrow from trypsin to the left). Trypsin degradation is inhibited by calcium, but facilitated by trypsin (solid green arrow from trypsin to the right) and chymotrypsin-C (CTRC). Mutations in the calcium-sensing receptor (CASR) or CTRC lead to excess trypsin and eventually pancreatic injury (yellow). After pancreatic injury and an acute inflammatory response (AIR), there is a dramatic upregulation of SPINK1, which inhibits trypsin’s effects on autoactivation and further damage. If trypsinogen (and trypsin) are secreted into the duodenum, then the pancreas is protected by removal of trypsin from the pancreas by a CFTR-dependent mechanism. Blue letters indicate that mutations in the corresponding genes are associated with an increased risk of pancreatitis. Dashed lines indicate a pathway contingent on the absence of inhibition. CFTR, cystic fibrosis transmembrane conductance regulator; PRSS1, protease serine 1; SPINK1, serine protease inhibitor, Kazal type 1.

The importance of maintaining a tight regulation of cationic trypsinogen is further highlighted by additional genetic findings. First, the R122H and N29I mutations appear to be gene conversions, in which a segment of deoxyribonucleic acid (DNA) from another similar gene replaces a similar segment of the gene of interest. There is now strong evidence that the “H” of PRSS1 R122H is a conversion mutation from Tryp 6,²⁷ and the “I” of N29I is from PRSS2.^28,29 These findings highlight the critical importance of the high-fidelity regulatory mechanism of PRSS1 in preventing recurrent, premature trypsinogen activation or trypsin survival in low-calcium environments. Secondly, it has been reported that gene copy number variants are also associated with a risk of chronic pancreatitis in France.^30,31 In this case a segment of the genome containing the trypsinogen genes is duplicated on one chromosome, and when combined with the expression of trypsinogen for the opposite allele gives an extra “dose” of trypsinogen—enough to increase the risk of pancreatitis.

ANIONIC TRYPSINOGEN GENE MUTATIONS

Anionic trypsinogen (UniGene name: protease, serine 2; PRSS2) is a form of pancreatic trypsinogen that is usually expressed at about half the amount as cationic trypsinogen, although this ratio may change in some cases.^32,33 To date, no gain-of-function mutations have been identified. However, a loss-of-function mutation, PRSS2 G191R, is associated with protection from pancreatitis.^34,35 The mutation introduces an arginine (“R”) into a surface loop of PRSS2, making it a target for trypsin-mediated degradation.

CALCIUM-SENSING RECEPTOR GENE POLYMORPHISMS

Regulation of intra-acinar cell calcium is critical for the prevention of pancreatic injury.^24,36 The calcium-sensing receptor (CASR) is a membrane-bound member of the G-protein–coupled receptor superfamily.³⁷ CASR plays an important role in calcium homeostasis, as is reflected in its expression by cells of the parathyroid gland and renal tubules that are involved in the calcium metabolism. CASR has been identified in human pancreatic acinar and ductal cells, as well as in various nonexocrine tissues,³⁸ although its functional significance in the pancreas has not yet been determined. More than 100 functional mutations (40 activating and 72 inactivating) have been described in the CASR mutation database related to familial hypocalcuric hypercalcemia (FHH), neonatal severe primary hyperparathyroidism (NSPHT), autosomal dominant hypocalcaemia (ADH) and related hypercalcemic or hypocalcemic disorders.³⁹

In 2003 Felderbauer⁴⁰ investigated a kindred with familial pancreatitis and a serine protease inhibitor Kazal type 1 (SPINK1) gene mutation (see following for discussion of SPINK1 mutations). However, only two of these family members had chronic pancreatitis, and both were found to have a novel CASR 518T>C mutation that was linked to hypercalcemia. An association between additional CASR variants, with or without SPINK1 mutations, was subsequently identified in patients in India with TP,⁴¹ as well as in the United States in sporadic and alcoholic chronic pancreatitis in which the CASR mutation doubles and triples the relative risk, respectively.⁴² The finding of different CASR polymorphisms in different populations is intriguing, but it appears that the presumed mild hypercalcemia is a cofactor rather than an independent risk factor, as seen in animal models of hypercalcemia.^20,43 At the present time, CASR mutations analysis is not offered for clinical decision making.

CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR GENE MUTATIONS

The cystic fibrosis transmembrane conductance regulator (CFTR) gene is the most important molecule for regulating pancreatic duct cell function. This includes generating the bicarbonate-rich pancreatic juice that helps neutralize the acidic chyme coming from the stomach and flushing digestive enzymes out of the pancreas and into the duodenum. Both functions are accomplished by anion secretion through the CFTR located on the apical side of the pancreatic duct cells (see Chapter 56). The CFTR molecule becomes relevant to pancreatic diseases when its function, or regulation of its function, is altered by various gene mutations.

The CFTR molecule forms a regulated ion channel expressed on epithelial cells in the respiratory system, sweat glands, the digestive tract mucosa, biliary epithelium, pancreatic duct cells, and other locations. The primary anions conducted through CFTR under physiologic conditions are chloride and, under some conditions, bicarbonate. The CFTR gene contains more than 4300 nucleotides, with 24 exons and three splice variants that code for a single protein of 1480 amino acids.⁴⁴ The CFTR molecule has 12 membrane spanning domains, 2 nucleotide-binding domains (NBD1 and NBD2), and a regulatory domain (R domain) with multiple phosphorylation sites (Fig. 57-4).

Figure 57-4. CFTR structural domains. The CFTR molecule is a single peptide that forms a regulated anion channel through the apical cell membrane of the pancreatic duct cell. CFTR exists in at least two conformations (single channel and double channel). The molecule is positioned in the cell membrane by 12 transmembrane domains (numbered 1 through 12). There are at least three major regulatory domains including nucleotide binding domain 1 and 2 (NBD1, NBD2) and a regulatory domain (R domain). Several second-messenger systems interact directly with the regulatory domain, including ATP and PKA. Calcium, intracellular glutamate, and other second messenger systems or factors also regulate various aspects of CFTR (not shown). ATP, adenosine triphosphate; CFTR, cystic fibrosis transmembrane conductance regulator; PKA, protein kinase A.

Although the regulation of CFTR is complex, many of the components that are relevant to the pancreas are now understood. CFTR-associated secretion is stimulated when the duct cell is stimulated by secretin or vasoactive intestinal peptide (VIP) acting on receptors that increase intracellular cyclic adenosine monophosphate (cAMP). The cAMP activates protein kinase A–mediated phosphorylation of various sites in the R domain, followed by increased anion conductance (e.g., chloride, bicarbonate) through the CFTR channel. The function associated with the individual phosphorylation sites on the R domain differs.⁴⁵ Possible consequences of R domain phosphorylation include movement and insertion into the apical membrane, increased or decreased channel activity or specificity, or stabilization of other parts of the molecule such as NBD1. Duct cell stimulation by cholinergic agents or other agonists that increase intracellular calcium also potentiate anion secretion. Recent studies demonstrate that chloride conductance is regulated by cytoplasmic glutamate, while bicarbonate conductance is regulated by binding of ATP to NBD1 and NBD2,⁴⁶ which may form a heterodimer to maximize adenosine triphosphatase (ATPase) activity.⁴⁷ Structural studies of the CFTR protein suggest that the molecule exists in two different conformations depending on the presence or absence of ATP binding to NBD1 and NBD2.⁴⁸ If the basolateral membrane of the duct cell is nearly impermeable to chloride during bicarbonate secretion and if bicarbonate conductance through CFTR is limited because of an unfavorable CFTR conformation, then net ion transport across the basolateral and apical membranes would markedly be reduced⁴⁹ and the duct cells would not be able to help flush digestive enzymes out of the duct. This would put the pancreas at risk of recurrent acute pancreatitis because prematurely activated digestive enzymes would not be removed from the pancreas.

Major mutations in both CFTR alleles result in loss of CFTR function. The consequences include inability to adequately hydrate mucus and other macromolecules, leading to accumulation of viscid material and inspissated glands. This condition results in progressive organ destruction of the pancreas and respiratory system, and dysfunction of the liver, intestine, sweat glands, and other sites where epithelial cell secretion plays an important physiologic role. The pancreas incurs a double risk because much of its proteins are zymogens and trypsin activation will lead to recurrent injury and eventually destruction of the pancreas through progressive fibrosis. Trypsin-mediated injury and destruction of the pancreas in children with CF is consistent with this model because the pancreatic pathology in CF is pseudocyst formation and fibrosis rather than atrophy (as expected with duct obstruction).⁵⁰ It appears that pancreatic gland injury in CF children roughly parallels the expression of trypsinogen in the developing acinar cells, which begins at 16 weeks’ gestation and gradually increases in concentration until birth and through the first six months of life when levels markedly rise.^51,52 The resulting histology has many of the features of end-stage chronic pancreatitis that develops in children and adults, but also has striking expanded ducts that appear as multiple protein filled cysts (Fig. 57-5).

Figure 57-5. Histology of the pancreas in cystic fibrosis (CF). This case is from an autopsy of a child with severe features of cystic fibrosis. There are no residual normal ducts or acini. Instead, dilated ducts and “cysts” with inspissated material dominate the image. Other cases of CF span the spectrum between this image and chronic pancreatitis seen with other forms of pancreatitis, with acinar atrophy, fibrosis, and chronic inflammation. Arrows demonstrate residual islets.

(Image courtesy of Alysa Krasinska, MD.) (From Whitcomb DC. Cystic fibrosis–associated pancreatitis. In: Beger HG, Warshaw A, Buchler MW, et al, editors. The Pancreas: An Integrated Textbook of Basic Science, Medicine and Surgery. Oxford, England: Blackwell; 2008.)

The overall clinical picture in an individual case depends on the nature of the combined CFTR mutations, the genetic background in which the defective genes operates (e.g., modifier genes), and environmental factors.^50,53 About 70% of white patients with CF have a three-base pair deletion of the phenylalanine-coding codon 508 (?F508), although 1600 other mutations have been reported. Distinct mutations are common to certain ethnic groups, including African Americans who carry the 3120+1G>A mutation at a frequency of 12.3%⁵⁴ and the R334W mutation in Hispanics, which is associated with CF with pancreatic sufficiency (PS) but recurrent acute pancreatitis.⁵⁵ Patients with one severe CFTR mutation and one mild CFTR mutation (e.g., R117H or R334W) often have CF with pancreatic sufficiency.^56,57 The reason may be that mutations such as R117H markedly reduce chloride conductance without affecting bicarbonate secretion.^46,58 Reduction of chloride conductance would affect all of the epithelial cells in organs that use CFTR to transport chloride, but would have much less effect on the pancreatic duct because it uses CFTR to transport bicarbonate.⁴⁹ On the other hand, CFTR mutations that specifically inhibit bicarbonate secretion put the pancreas at risk of recurrent pancreatitis without affecting the organs that use CFTR to transport chloride.

The functional consequence of CFTR mutations depends on the combined effects of both CFTR alleles with the severity of the phenotype dependent on the mildest mutation.⁵⁹ The most common CFTR mutations have been organized according to the effect on clinical phenotype (severe, mild-variable, borderline, benign) and the effect on CFTR protein structure and function (classes 1 to 5).^50,53 (Examples are given in Table 57-1.) Class 1 to 3 mutations result in no functional protein and therefore are associated with a severe phenotype when combined with a second severe CFTR mutation. Class 4 and 5 CFTR mutations result in CFTR proteins with altered, but residual function, and are associated with mild-variable or borderline phenotypes. There is current debate as to whether additional classes are justified. If a class 4 or 5 mutation associates with a class 1 to 3 mutation, the phenotype is mild CF, with only a subset of organs affected.⁶⁰ The resulting conditions are often called atypical CF.⁵⁹ If a class 1 to 3 or some class 4 mutations are combined with wild-type CFTR or a benign polymorphism, the overall function of CFTR is reduced by up to 50%, but the phenotype is usually normal because more than 90% of overall CFTR function must be lost before clinical features of atypical CF are seen.⁵⁹ Recurrent acute pancreatitis requires, at minimum, a partially functioning pancreas and is therefore seen in some cases of CF with pancreatic sufficiency and atypical CF.

Many of the features of CF cannot be explained by variations in CFTR sequence. Instead, these features are caused by specific environmental factors or modifier genes.^50,61 Environmental factors, such as bacterial colonization of the respiratory system, tobacco smoke,⁶² and poor nutritional status contribute to the severity of lung disease.⁶³ The risk of liver disease appears to be independent of CFTR genotype; it is associated with meconium ileus.⁶⁴ Careful consideration must be given to patient with either classic CF symptoms, or atypical presentations resulting from less common combinations of genetic and environmental factors.

In 1998 two groups^65,66 demonstrated that CFTR mutations were also very common in idiopathic and alcoholic chronic pancreatitis, which indicates that some of the more than 1600 known CFTR gene sequence variants cause milder disease (atypical CF⁵⁹), pancreas-specific injury or that CFTR gene mutations may be part of a more complex trait.¹ In some cases, recurrent acute pancreatitis and chronic pancreatitis appeared to be associated with heterozygous CFTR genotypes. Because heterozygous CFTR mutations and polymorphisms are common in the general European and American population, and because the parents of CF children (obligate CFTR mutation carriers without CF) do not have an increased incidence of acute or chronic pancreatitis compared with the normal population,⁶⁷ it is likely that a second factor that specifically targets the pancreas is required.¹ In early onset idiopathic pancreatitis this second factor may be a SPINK1 mutation (discussed later), a potent environmental factor,⁶⁸ or other complex genetic combinations.^69–73

SHWACHMAN-BODIAN-DIAMOND GENE MUTATIONS

The Shwachman-Bodian-Diamond syndrome gene (SBDS) is a gene of unknown function that is mutated in most cases of Shwachman-Diamond syndrome (SDS).^74,75 This gene has five exons and encodes a predicted protein of 250 amino acids. The genetic defect in most cases of SDS is caused by gene conversion between the normal SBDS gene with a nonfunctional pseudogene designated SBDSP.⁷⁴ The SBDSP pseudogene DNA code is 97% identical to SBDS gene code. The differences between the SBDS gene and SBDSP pseudogene are critical nucleotide deletions and nucleotide changes that render the SBDS gene product nonfunctional. Fourteen distinct mutations were initially identified in these kindreds, with the most common being the conversion mutations 183-184TA→CT and 258+2T→C. Interestingly, most patients had compound heterozygous mutations, and no patient was homozygous for the common 183-184TA→CT mutation. Although the function of the gene is unknown, there is significant homology with genes in other species that regulate or facilitate mRNA utilization or metabolism.⁷⁴ Recent studies suggest that the gene encodes a protein that interacts with ribosomal RNA and helps regulate the maturation of the 60s ribosomal subunit and may also stabilize the mitotic spindle.^76–78 The genetic defect results in an acinar cell–specific defect with markedly reduced zymogen synthesis and pancreatic insufficiency rather than susceptibility to pancreatitis. The other clinical features of the SDS are discussed later.

PANCREATIC SECRETORY TRYPSIN INHIBITOR GENE MUTATIONS

Pancreatic secretory trypsin inhibitor (PSTI, UniGene name: serine protease inhibitor, Kazal type 1; SPINK1) is a 56 amino acid peptide that specifically inhibits trypsin by physically blocking its active site (see Fig. 57-2). SPINK1 is synthesized by pancreatic acinar cells along with trypsinogen and it co-localizes with trypsinogen in the zymogen granules. In the mechanistic models of pancreatic acinar cell protection, SPINK1 acts as the first line of defense against prematurely activated trypsinogen in the acinar cell.^9,32,79 SPINK1 is an acute phase reactant and concentrations in serum rise markedly with inflammation.^80,81 Under normal conditions there is a great excess of potential trypsin compared with SPINK1 so that the inhibitory capacity of SPINK1 is limited, but with inflammation the expression of SPINK1 is increased dramatically.⁸²

A number of SPINK1 mutations have been identified, but a high-risk haplotype (five polymorphisms that are inherited together) defined by SPINK1 N34S is by far the most common, being present in 1% to 4% of most populations throughout the world.^79,83,84 Several other variants of the SPINK1 gene also have been described.²⁵ Pancreatitis associated with SPINK1 gene mutations is associated with early onset recurrent acute and chronic pancreatitis in children,⁷⁹ familial pancreatitis,⁸³ and TP,⁸⁵ and is often a feature of polygenic pancreatitis–associated genotype.

Because SPINK1 is a specific trypsin inhibitor, and because expression of SPINK1 is normally very low,⁸² it follows that SPINK1 cannot be a significant inhibitor of prematurely activated trypsin until the acute inflammatory process has become well established. Thus, the effect of loss-of-function SPINK1 polymorphisms would only become apparent late in the course of acute pancreatitis, or in the case of recurrent premature trypsinogen activation, as is seen with PRSS1 or CFTR mutations. This hypothesis was tested using multiple meta-analyses on data that were classified by proximal etiology.⁸⁶ This analysis suggested that the strongest effect of SPINK1 N34S was in pancreatitis etiologies that were linked with recurrent trypsin activation, but the effect was low in other etiologies (e.g., alcohol-associated pancreatitis). If this is mechanistically true, then up-regulation of mutant SPINK1 will fail to prevent trypsin-associated recurrent pancreatic injury. Furthermore, suppression of recurrent proinflammatory responses with macrophage-derived anti-inflammatory cytokines (e.g., TGF-β1) will drive the pancreatic stellate cells to produce fibrosis.^7,14,87 The implication is that the discovery of a SPINK1 mutation in an unaffected individual is of minimal importance, whereas the effect of a SPINK1 mutation in driving fibrosis in someone with PRSS1 or CFTR mutations is very strong. A summary of the mechanisms for trypsin activation and inactivation is illustrated in Figure 57-3.

ETIOLOGY OF ACUTE PANCREATITIS

The primary causes of acquired pancreatitis in children are listed in Table 57-2. In a recent review of 1276 children with acute pancreatitis compiled from five studies, the most common causes were idiopathic (22.2%), association with systemic disease (20.8%), trauma (18.6%), structural (e.g., pancreas divisum) (10.6%), and medications (10.2%); gallstones, post–endoscopic retrograde cholangiopancreatography (ERCP), familial, hypercalcemia, hyperlipidemia, diabetic ketoacidosis, and “other” causes made up the remaining etiologies.⁸⁸ Growing evidence suggests that some of these cases occur in children with high-risk genetic alterations, especially pancreatic-specific combinations of SPINK1 and CFTR mutations. Genetic testing, discussed later in this chapter, is usually performed after recurrent episodes and when other common causes have been excluded.

Table 57-2 Reported Causes of Acquired Pancreatitis in Children

AIDS, acquired immunodeficiency syndrome; ERCP, endoscopic retrograde cholangiopancreatography; HIV, human immunodeficiency virus.

* Class 1 drug: pancreatitis occurs with rechallenge(see Chapter 58).

† Probable cause of pancreatitis.