Abstract
Chronic pancreatitis is an inflammatory disorder of the pancreas. We analyzed CPA1 encoding carboxypeptidase A1 in subjects with non-alcoholic chronic pancreatitis and controls in a German discovery cohort and three replication cohorts. Functionally impaired variants were present in 29/944 (3.1%) German patients and in 5/3,938 (0.1%) controls (odds ratio [OR] = 24.9; P = 1.5 × 10-16). The association was strongest in subjects aged ≤10 years (9.7%; OR = 84.0; P = 4.1 × 10-24). In the replication cohorts, defective CPA1 variants were observed in 8/600 (1.3%) patients and in 9/2,432 (0.4%) controls from Europe (P = 0.01), in 5/230 (2.2%) patients and 0/264 controls from India (P = 0.02), and in 5/247 (2.0%) patients but 0/341 controls from Japan (P = 0.013). The mechanism of increased pancreatitis risk by CPA1 variants may involve misfolding-induced endoplasmic reticulum stress rather than elevated trypsin activity as seen with other genetic risk factors.
Chronic pancreatitis is an inflammatory condition characterized by abdominal pain and progressive damage to both exocrine and endocrine components of the pancreas resulting in insufficiency of the organ with maldigestion and diabetes. Although alcohol abuse has been long recognized as a major risk factor for chronic pancreatitis, genetic susceptibility has emerged during the last two decades as a strong determinant of disease risk, particularly in the pediatric population1.
Genetic studies performed to date suggest that development of intra-pancreatic trypsin activity plays a central role in disease pathogenesis. Thus, gain-of-function mutations in cationic trypsinogen (PRSS1, OMIM 276000) as well as loss-of-function variants in the pancreatic trypsin inhibitor (SPINK1, OMIM 167790) and the trypsinogen-degrading enzyme chymotrypsin C (CTRC, OMIM 601405) increase the risk for chronic pancreatitis2-8. Consistent with the proposed pathogenic role of trypsin, a rapidly auto-degrading variant of anionic trypsinogen (PRSS2, OMIM 601564) and a common PRSS1 promoter variant protect against chronic pancreatitis9,10.
Despite these recent advances, many patients do not carry mutations in any of the known susceptibility genes, suggesting the involvement of other yet unidentified genes. In the present study, we investigated the role of CPA1 encoding carboxypeptidase A1 in chronic pancreatitis. Digestive carboxypeptidases are pancreatic metalloproteases, which hydrolyze C-terminal peptide bonds in dietary polypeptide chains11. Three different isoforms have been described in human pancreatic juice. A-type carboxypeptidases (CPA1 and CPA2) act on aromatic and aliphatic amino acid residues exposed by the action of chymotrypsins and elastases, whereas the B-type carboxypeptidase (CPB1) hydrolyzes C-terminal Lys and Arg residues generated by tryptic cleavages11. The gene encoding human CPA1 (OMIM 114850) maps to 7q32.2, spans approximately 8 kb, and contains 10 exons. The inactive preproprotein comprises 419 amino acids, including a 16 amino-acid secretory signal peptide and a 94 amino-acid long propeptide. Activation of human proCPA1 to CPA1 is catalyzed by the sequential action of trypsin and CTRC, which cleave and degrade the propeptide12. After trypsinogens, proCPA1 is the second largest component of pancreatic juice, contributing more than 10% of the total protein13.
We performed direct DNA sequencing of all 10 CPA1 exons in 944 individuals with non-alcoholic chronic pancreatitis and in 3,938 control subjects of German origin. Considering variants in the coding regions and flanking splice sites, we identified 31 missense variants, 1 nonsense variant, 1 frame-shift variant, and 1 splice-site variant; and found that 3 variants were significantly enriched in patients (Table 1). Functional analysis demonstrated that 17/34 (50%) variants resulted in a marked (>80%) loss of apparent CPA1 activity, a term we use to describe the combined effects of variants on secretion, proteolytic stability and catalytic competence (Table 1, Supplementary Figure 1, Methods). The vast majority of these variants were located in exons 7, 8, and 10. Remarkably, 14 out of 17 (82%) functionally impaired variants were found exclusively in patients, including the c.768C>G (p.Asn256Lys) variant, which was detected in 7 patients. Thus, CPA1 variants with less than 20% apparent activity were significantly overrepresented in the chronic pancreatitis group (29/944; 3.1%) as compared to controls (5/3,938; 0.1%) (OR = 24.9; CI = 9.6-64.6; P = 1.5 × 10-16) (Table 1). No individual was compound heterozygous or homozygous for a defective CPA1 variant. Variants found in non-coding regions and synonymous variants in coding regions are listed in Supplementary Table 1.
Table 1.
Non-synonymous CPA1 variants in German subjects with non-alcoholic chronic pancreatitis and healthy controls
| Exon | Nucleotide change | Amino acid change | Patients (n = 944) | Controls (n = 3938) | P value | OR | 95% CI | Apparent activity | Secretion level |
|---|---|---|---|---|---|---|---|---|---|
| 1 | c.5G>A | p.Arg2Gln | 0 (0%) | 2 (0.05%) | 1.0 | - | - | 103 | 92 |
| 2 | c.79C>T | p.Arg27Ter | 0 (0%) | 2 (0.05%) | 1.0 | - | - | 0 | 0 |
| 2 | c.101C>T | p.Ala34Val | 0 (0%) | 1 (0.03%) | 1.0 | - | - | 98 | 97 |
| 3 | c.197G>A | p.Arg66Gln | 0 (0%) | 2 (0.05%) | 1.0 | - | - | 60 | 55 |
| 3 | c.281A>G | p.Gln94Arg | 1 (0.1%) | 13 (0.3%) | 0.5 | - | - | 57 | 57 |
| 3 | c.321C>G | p.Phe107Leu | 0 (0%) | 1 (0.03%) | 1.0 | - | - | 112 | 100 |
| 3 | c.371C>T | p.Thr124Ile | 8 (0.9%) | 45 (1.1%) | 0.6 | - | - | 23 | 27 |
| 4 | c.410C>G | p.Ala137Gly | 1 (0.1%) | 0 (0%) | 0.2 | - | - | 52 | 56 |
| 5 | c.497G>A | p.Gly166Asp | 5 (0.5%) | 20 (0.5%) | 1.0 | - | - | 73 | 66 |
| 5 | c.542G>A | p.Arg181Gln | 0 (0%) | 1 (0.03%) | 1.0 | - | - | 1 | 39 |
| 6 | c.622G>A | p.Ala208Thr (het) | 71 (7.5%) | 266 (6.8%) | 0.4 | - | - | 81 | 73 |
| 6 | c.622G>A | p.Ala208Thr (hm) | 1 (0.1%) | 1 (0.03%) | 0.4 | - | - | - | - |
| 6 | c.622G>T | p.Ala208Ser | 0 (0%) | 1 (0.03%) | 1.0 | - | - | 91 | 83 |
| 6 | c.673G>A | p.Gly225Ser | 1 (0.1%) | 0 (0%) | 0.2 | - | - | 4 | 12 |
| 7 | c.710G>A | p.Arg237His | 0 (0%) | 2 (0.05%) | 1.0 | - | - | 0 | 81 |
| 7 | c.751G>A | p.Val251Met | 2 (0.2%) | 0 (0%) | 0.1 | - | - | 0 | 0 |
| 7 | c.758C>G | p.Pro253Arg | 1 (0.1%) | 0 (0%) | 0.2 | - | - | 0 | 0 |
| 7 | c.768C>G | p.Asn256Lys | 7 (0.7%) | 0 (0%) | 9.9 × 10-6 | nc | nc | 0 | 0 |
| 7 | c.775G>A | p.Ala259Thr | 0 (0%) | 1 (0.03%) | 1.0 | - | - | 85 | 82 |
| 8 | c.811T>C | p.Cys271Arg | 1 (0.1%) | 0 (0%) | 0.2 | - | - | 1 | 0 |
| 8 | c.829G>A | p.Gly277Ser | 1 (0.1%) | 0 (0%) | 0.2 | - | - | 0 | 0 |
| 8 | c.839C>A | p.Ala280Asp | 1 (0.1%) | 0 (0%) | 0.2 | - | - | 0 | 5 |
| 8 | c.847G>A | p.Glu283Lys | 2 (0.2%) | 0 (0%) | 0.1 | - | - | 0 | 0 |
| 8 | c.982G>A | p.Glu328Lys | 1 (0.1%) | 0 (0%) | 0.2 | - | - | 9 | 42 |
| 9 | c.1009G>C | p.Val337Leu | 0 (0%) | 1 (0.03%) | 1.0 | - | - | 64 | 61 |
| Intron 9 | c.1073-2A>G | p.Tyr358fs§ | 3 (0.3%) | 0 (0%) | 0.007 | nc | nc | 0 | 0 |
| 10 | c.1085G>A | p.Gly362Glu | 1 (0.1%) | 0 (0%) | 0.2 | - | - | 0 | 6 |
| 10 | c.1126T>C | p.Ser376Pro | 2 (0.2%) | 0 (0%) | 0.1 | - | - | 0 | 7 |
| 10 | c.1144C>T | p.Arg382Trp | 5 (0.5%) | 0 (0%) | 0.0003 | nc | nc | 0 | 31 |
| 10 | c.1157G>A | p.Arg386His | 0 (0%) | 1 (0.03%) | 1.0 | - | - | 92 | 97 |
| 10 | c.1193C>T | p.Pro398Leu | 0 (0%) | 1 (0.03%) | 1.0 | - | - | 42 | 64 |
| 10 | c.1217C>G | p.Ala406Gly | 0 (0%) | 1 (0.03%) | 1.0 | - | - | 137 | 114 |
| 10 | c.1247delA | p.Asn416fs | 1 (0.1%) | 0 (0%) | 0.2 | - | - | 11 | 15 |
| 10 | c.1251C>A | p.His417Gln | 0 (0%) | 1 (0.03%) | 1.0 | - | - | 62 | 54 |
| 10 | c.1253C>T | p.Pro418Leu | 0 (0%) | 1 (0.03%) | 1.0 | - | - | 91 | 99 |
| All variants with apparent activity <20% | 29 (3.1%) | 5 (0.1%) | 1.5 × 10-16 | 24.9 | 9.6-64.6 | - | - | ||
P values were determined by Fisher’s Exact Test. Apparent CPA1 activity and secretion level are expressed as percent of wild type. Apparent activity corresponds to the CPA1 activity measured in the conditioned medium of transfected cells after activation with trypsin and chymotrypsin C (CTRC) (See Methods). Thus, apparent activity reflects the combined effects of the variants on secretion, catalytic activity and degradation by trypsin and/or CTRC. Secretion level indicates the concentration of proCPA1 in the conditioned medium measured by SDS-PAGE and densitometry (See Methods). Alterations marked in bold indicate variants with less than 20% apparent activity.
Splice site variant modeled functionally as intron retention as described in Methods.
OR, odds ratio, CI confidence interval, het, heterozygous; hm, homozygous; nc, not calculated as variant was not detected in controls rendering OR infinite.
We observed that patients bearing a defective CPA1 variant were younger than those without a CPA1 alteration. In the German chronic pancreatitis group, the majority of CPA1 variants with less than 20% apparent activity were observed in patients at or below 20 years of age (27/586 [4.6%]; OR = 38.0; CI = 14.6-99.1; P = 6.8 × 10-20]. This becomes even more significant in a subgroup of patients at or below 10 years of age. In this group, 22/228 (9.7%) carried an impaired CPA1 variant (OR = 84.0; CI = 31.5-224.1; P = 4.1 × 10-24) (patients ≤ 10 yrs. vs. patients ≤ 20 years, P = 0.007; patients ≤ 10 yrs. vs. all patients, P = 7.6 × 10-5) (Table 2).
Table 2.
Distribution of functionally impaired CPA1 variants in different age groups of German subjects with non-alcoholic chronic pancreatitis
| Age of Patients | Patients | Controls | P value | OR | 95% CI |
|---|---|---|---|---|---|
| All | 29/944 (3.1%) | 5/3938 (0.1%) | 1.5 × 10-16 | 24.9 | 9.6-64.6 |
| age > 20 years | 2/358 (0.6%) | 5/3938 (0.1%) | 0.2 | - | - |
| age ≤ 20 years | 27/586 (4.6%) | 5/3938 (0.1%) | 6.8 × 10-20 | 38.0 | 14.6-99.1 |
| age ≤ 10 years | 22/228 (9.7%) | 5/3938 (0.1%) | 4.1 × 10-24 | 84.0 | 31.5-224.1 |
P values were determined by Fisher’s Exact Test. Alterations with less than 20% apparent activity were included.
We also investigated all CPA1 exons in 465 German patients with alcohol-related chronic pancreatitis. Only 2/465 (0.4%) patients were heterozygous for a defective CPA1 variant: c.954_955delCA (p.Tyr318Ter) and c.811T>C (p.Cys271Arg), respectively. This indicates that loss-of-function CPA1 alterations play a minor role in alcoholic pancreatitis.
To confirm the association of non-alcoholic chronic pancreatitis and CPA1 in an independent European cohort, we sequenced all CPA1 exons in 600 patients with non-alcoholic chronic pancreatitis and 2,432 control subjects originating from France, the Czech Republic and Poland. Again, variants with less than 20% apparent activity were significantly overrepresented in chronic pancreatitis patients (8/600; 1.3%) versus ethnically matched controls (9/2,432; 0.4%) (OR = 3.6; CI = 1.4-9.5; P = 0.01) (Table 3). One subject with chronic pancreatitis was homozygous for the c.1115G>A (p.Gly372Asp) variant.
Table 3.
Non-synonymous CPA1 variants in the European replication study
| Exon | Nucleotide change | Amino acid change | Patients (n = 600) | Controls (n = 2432) | P value | OR | 95% CI | Apparent activity | Secretion level |
|---|---|---|---|---|---|---|---|---|---|
| 1 | c.5G>A | p.Arg2Gln | 0 (0%) | 6 (0.2%) | 0.6 | - | - | 103 | 92 |
| 2 | c.79C>T | p.Arg27Ter | 1 (0.2%) | 5 (0.2%) | 0.7 | - | - | 0 | 0 |
| 2 | c.80G>C | p.Arg27Pro | 0 (0%) | 1 (0.04%) | 1.0 | - | - | 0 | 0 |
| Intron 2 | c.148-1G>A | p.Leu50_Glu127del§ | 0 (0%) | 1 (0.04%) | 1.0 | - | - | 0 | 0 |
| 3 | c.197G>A | p.Arg66Gln | 0 (0%) | 1 (0.04%) | 1.0 | - | - | 60 | 55 |
| 3 | c.241T>C | p.Ser81Pro | 0 (0%) | 1 (0.04%) | 1.0 | - | - | 53 | 57 |
| 3 | c.281A>G | p.Gln94Arg | 0 (0%) | 9 (0.4%) | 1.0 | - | - | 57 | 57 |
| 3 | c.313T>C | p.Phe105Leu | 1 (0.2%) | 0 (0%) | 0.2 | - | - | 109 | 99 |
| 3 | c.334C>T | p.Arg112Cys | 1 (0.2%) | 0 (0%) | 0.2 | - | - | 69 | 78 |
| 3 | c.371C>T | p.Thr124Ile | 3 (0.5%) | 14 (0.6%) | 1.0 | - | - | 23 | 27 |
| 4 | c.389A>C | p.Asp130Ala | 0 (0%) | 1 (0.04%) | 1.0 | - | - | 77 | 68 |
| 5 | c.497G>A | p.Gly166Asp | 1 (0.2%) | 11 (0.5%) | 0.5 | - | - | 73 | 66 |
| 6 | c.604C>A | p.Gln202Lys | 1 (0.2%) | 0 (0%) | 0.2 | - | - | 114 | 104 |
| 6 | c.622G>A | p.Ala208Thr | 45 (7.5%)* | 143 (5.9%)* | 0.1 | - | - | 81 | 73 |
| 6 | c.686C>T | p.Thr229Met | 0 (0%) | 1 (0.04%) | 1.0 | - | - | 0 | 0 |
| 6 | c.695C>T | p.Thr232Met | 1 (0.2%) | 0 (0%) | 0.2 | - | - | 87 | 80 |
| 7 | c.751G>A | p.Val251Met | 1 (0.2%) | 0 (0%) | 0.2 | - | - | 0 | 0 |
| 8 | c.809C>G | p.Pro270Arg | 1 (0.2%) | 0 (0%) | 0.2 | - | - | 9 | 14 |
| 8 | c.941A>G | p.Tyr314Cys | 1 (0.2%) | 0 (0%) | 0.2 | - | - | 0 | 23 |
| 8 | c.954_955delCA | p.Tyr318Ter | 2 (0.3%) | 0 (0%) | 0.04 | nc | nc | 0 | 0 |
| 9 | c.1010T>C | p.Val337Ala | 0 (0%) | 1 (0.04%) | 1.0 | - | - | 63 | 90 |
| Intron 9 | c.1072+1G>T | p.Asp330fs§ | 0 (0%) | 1 (0.04%) | 1.0 | - | - | 0 | 0 |
| Intron 9 | c.1073-2A>G | p.Tyr358fs§ | 1 (0.2%) | 0 (0%) | 0.2 | - | - | 0 | 0 |
| 10 | c.1115G>A | p.Gly372Asp | 1& (0.2%) | 0 (0%) | 0.2 | - | - | 25 | 34 |
| 10 | c.1203G>C | p.Lys401Asn | 1 (0.2%) | 0 (0%) | 0.2 | - | - | 115 | 103 |
| 10 | c.1217C>T | p.Ala406Val | 1 (0.2%) | 0 (0%) | 0.2 | - | - | 0 | 87 |
|
| |||||||||
| All variants with apparent activity <20% | 8 (1.3%) | 9 (0.4%) | 0.01 | 3.6 | 1.4-9.5 | - | - | ||
P values were determined by Fisher’s Exact Test. Apparent CPA1 activity and secretion level were measured as described in Table 1 and expressed as percent of wild type. Alterations marked in bold indicate variants with less than 20% apparent activity.
Functional effects of splice site variants c.148-1G>A, c.1072+1G>T, and c.1073-2A>G were modeled as skipping of exon 3, skipping of exon 9 and retention of intron 9, respectively, as described in Methods.
One individual was homozygous for p.Ala208Thr.
Individual was homozygous for this variant.
OR, odds ratio, CI confidence interval, nc, not calculated as variant was not detected in controls rendering OR infinite.
In order to investigate the significance of CPA1 variants in subjects of non-European descent, we sequenced all 10 exons in 230 individuals with non-alcoholic chronic pancreatitis and 264 controls of Indian origin and in 247 patients and 341 controls from Japan. Overall, 2.2% (5/230) of Indian patients but none of the controls carried a defective CPA1 variant (P = 0.02) (Table 4). In the Japanese cohort, 2.0% (5/247) of patients but none of controls carried an impaired CPA1 variant (P = 0.013) (Table 5). No individual from India or Japan was compound heterozygous or homozygous for a defective CPA1 variant.
Table 4.
Non-synonymous CPA1 variants in Indian subjects with non-alcoholic chronic pancreatitis and healthy controls
| Exon | Nucleotide change | Amino acid change | Patients (n = 230) | Controls (n = 264) | P value | OR | 95% CI | Apparent activity | Secretion level |
|---|---|---|---|---|---|---|---|---|---|
| 2 | c.94G>C | p.Asp32His | 1 (0.4%) | 0 (0%) | 0.5 | - | - | 79 | 75 |
| 5 | c.506G>A | p.Arg169His | 4 (1.7%) | 0 (0%) | 0.046 | nc | nc | 24 | 23 |
| 6 | c.622G>A | p.Ala208Thr | 6 (2.6%) | 7 (2.7%) | 1.0 | - | - | 81 | 73 |
| 8 | c.922T>C | p.Tyr308His | 5 (2.2%) | 0 (0%) | 0.02 | nc | nc | 3 | 17 |
|
| |||||||||
| All variants with apparent activity <20% | 5 (2.2%) | 0 (0%) | 0.02 | nc | nc | - | - | ||
P values were determined by Fisher’s Exact Test. Apparent CPA1 activity and secretion level were measured as described in Table 1 and expressed as percent of wild type. Alterations marked in bold indicate variants with less than 20% apparent activity.
OR, odds ratio, CI confidence interval, nc, not calculated as variant was not detected in controls rendering OR infinite.
Table 5.
Non-synonymous CPA1 variants in Japanese subjects with non-alcoholic chronic pancreatitis and healthy controls
| Exon | Nucleotide change | Amino acid change | Patients (n = 247) | Controls (n = 341) | P value | OR | 95% CI | Apparent activity | Secretion level |
|---|---|---|---|---|---|---|---|---|---|
| 4 | c.410C>G | p.Ala137Gly | 1 (0.4%) | 0 (0%) | 0.42 | - | - | 52 | 56 |
| 7 | c.713A>T | p.Lys238Met | 1 (0.4%) | 0 (0%) | 0.42 | - | - | 0 | 3 |
| 7 | c.751G>A | p.Val251Met | 2 (0.8%) | 0 (0%) | 0.18 | - | - | 0 | 0 |
| 7 | c.764G>T | p.Arg255Met | 1 (0.4%) | 0 (0%) | 0.42 | - | - | 0 | 86 |
| 9 | c.1021G>A | p.Ala341Thr | 37 (15.0%) | 53 (15.5%) | 1.0 | - | - | 99 | 85 |
| 10 | c.1079-27_1111dup60 | p.Thr368_Tyr369ins20§ | 1 (0.4%) | 0 (0%) | 0.42 | - | - | 0 | 49 |
|
| |||||||||
| All variants with apparent activity <20% | 5 (2.0%) | 0 (0%) | 0.013 | nc | nc | - | - | ||
P values were determined by Fisher’s Exact Test. Apparent CPA1 activity and secretion level were measured as described in Table 1 and expressed as percent of wild type. Alterations marked in bold indicate variants with less than 20% apparent activity.
Functional effect of variant c.1079-27_1111dup60 was modeled as insertion of 20 amino acids between Thr368 and Tyr369, as described in Methods.
OR, odds ratio, CI confidence interval, nc, not calculated as variant was not detected in controls rendering OR infinite.
Chronic pancreatitis is a complex multi-genic disease and affected individuals often carry mutations in several disease-associated genes. To elucidate the relationship between CPA1 alterations and PRSS1, SPINK1, CTRC, and CFTR variants, we investigated all German subjects with chronic pancreatitis for mutations in PRSS1 (p.Ala16Val, p.Asn29Ile, and p.Arg122His), in SPINK1 (p.Asn34Ser and c.194+2T>C), in CTRC (p.Arg254Trp and p.Lys247_Arg254del), and in CFTR (p.Phe508del). In total, 50/944 (5.3%) individuals carried a heterozygous PRSS1 variant, 147/944 (15.6%) were positive for p.Asn34Ser (121 heterozygotes, 18 homozygotes) and c.194+2T>C (20 heterozygotes; 12/20 compound heterozygous with p.Asn34Ser), 28/944 (3.0%) were positive for a CTRC variant (21 instances of p.Arg254Trp and 7 occurrences of p.Lys247_Arg254del), and 42/944 (4.5%) were positive for CFTR p.Phe508del. Altogether, 273/944 (28.9%) of patients showed at least one of the above-mentioned genetic alterations and 24/944 (2.5%) were trans-heterozygous. However, only 1/29 (3.6%) patients with a defective CPA1 variant was trans-heterozygous; this subject carried the CPA1 c.1073-2A>G alteration (inherited from the mother) and the SPINK1 p.Asn34Ser variant (inherited from the father). This suggests limited interaction of CPA1 variants with variants in other susceptibility genes and stands in contrast with the high number of trans-heterozygotes for SPINK1 and CTRC and/or CFTR variants, as described recently14.
The mechanism by which loss-of-function CPA1 variants predispose to chronic pancreatitis is not intuitively apparent. We found no detectable effect of CPA1 on trypsinogen activation, trypsin activity or trypsinogen degradation by CTRC (Supplementary Figure 2), indicating that CPA1 mutations do not exert their effect via increasing intrapancreatic trypsin activity. On the other hand, the low apparent activity of most defective variants was due to markedly reduced secretion (Tables 1-5, Supplementary Figure 1 and 3), raising the possibility that CPA1 mutants misfold in the endoplasmic reticulum (ER) and cause ER stress, as demonstrated previously for some PRSS1 and CTRC mutants15,16. Indeed, expression of the most frequently found p.Asn256Lys variant in AR42J rat acinar cells resulted in ER stress, as evidenced by increased splicing of XBP1 and elevated mRNA levels of the chaperones BiP and calreticulin (Figure 1). Considering that CPA1 is one of the most abundant proteins synthesized by the pancreas, misfolding induced ER stress seems a plausible mechanism to explain the clinical effect of heterozygous CPA1 variants.
Figure 1.
Endoplasmic reticulum (ER) stress induced by the p.Asn256Lys CPA1 variant. (A) AR42J rat acinar cells were transfected with the indicated wild-type, mutant or empty adenovirus vectors for 24 h using 4×107 pfu per mL virus concentration. Conditioned media (200 μL) were precipitated with trichloroacetic acid (10% final concentration) and analyzed by SDS-PAGE and Coomassie blue staining. Note the complete lack of secretion of the p. Asn256Lys mutant. The faint band at 40 kDa represents an endogenous protein also found in the medium from cells infected with empty virus. The asterisk indicates the characteristically strong amylase band. See Methods for experimental details. A representative gel of three independent transfections is shown. (B) XBP1 splicing was assessed by RT-PCR and agarose gel electrophoresis with ethidium bromide staining. A representative gel of three independent experiments is shown. (C) Levels for spliced, unspliced and total XBP1 mRNA were measured by quantitative real-time PCR as described in Methods and expressed as fold changes relative to levels measured in cells transfected with empty adenovirus. (D) Quantitative real-time PCR measurement of BiP and calreticulin mRNA was performed as described in Methods and expressed as fold changes relative to levels measured in cells transfected with empty adenovirus. Error bars represent standard deviation (n = 3).
In summary, loss-of-function CPA1 variants are strongly associated with non-alcoholic chronic pancreatitis, especially with early-onset disease. Although there was evidence of mutational heterogeneity, identification of functionally impaired CPA1 variants in both the European and non-European cohorts establishes its global role in the pathogenesis of chronic pancreatitis.
ONLINE METHODS
Study population
The medical ethical review committees of all participating study centers approved this study. All study subjects gave informed consent. We enrolled 944 unrelated German individuals with the diagnosis of non-alcoholic chronic pancreatitis and 465 patients with alcohol-related chronic pancreatitis. In the replication study, we investigated 600 unrelated non-alcoholic chronic pancreatitis patients originating from France (n = 456), the Czech Republic (n = 21), and Poland (n = 123). In addition, we also investigated unrelated subjects affected with non-alcoholic chronic pancreatitis from India (n = 230) and Japan (n = 247). The diagnosis of chronic pancreatitis was based on two or more of the following findings: presence of a typical history of recurrent pancreatitis, pancreatic calcifications and/or pancreatic ductal irregularities revealed by endoscopic retrograde pancreaticography or by magnetic resonance imaging of the pancreas and/or pathological sonographic findings. Alcoholic chronic pancreatitis was diagnosed in patients who consumed more than 60 g (females) or 80 g (males) of ethanol per day for more than two years. Control subjects were recruited from Germany (n = 3,938), France (n = 2,000), the Czech Republic (n = 235), Poland (n = 197), India (n = 264), and Japan (n = 341).
Mutation screening
We designed primers complementary to intronic sequences flanking CPA1 exons based on the published nucleotide sequence (GenBank # NT_007933.15) (Supplementary Table 2). After PCR amplification, the entire coding region and the exon-intron boundaries were sequenced. All mutations were confirmed with a second independent PCR reaction. In the German laboratories, we performed PCR using 0.75 U AmpliTaq Gold polymerase (Perkin Elmer, Rodgau, Germany), 400 μmol/L deoxynucleoside triphosphates and 0.1 μmol/L primers in a total volume of 25 μL. Cycle conditions were as follows: initial denaturation for 12 min at 95°C; 48 cycles of 20 s denaturation at 95°C, 40 s annealing at 64°C and 90 s primer extension at 72°C; and a final extension step for 2 min at 72°C. PCR products were digested with Antarctic phosphatase (New England Biolabs, Ipswich MA) or shrimp alkaline phosphatase (USB, Santa Clara, CA) and exonuclease I (New England Biolabs, Ipswich MA). Cycle sequencing was performed using BigDye terminator mix (Applied Biosystems, Darmstadt, Germany) with 56° annealing temperature. The reaction products were purified with ethanol precipitation and loaded onto an ABI 3730 or an ABI 3100-Avant fluorescence sequencer (Applied Biosystems).
Functional characterization of CPA1 variants
We investigated the functional consequences of CPA1 alterations by transient transfection of HEK 293T cells (#Q401, GenHunter, Nashville, TN) with wild-type and mutant constructs and analyzing the conditioned medium for the amount of proCPA1 protein constitutively secreted using densitometry of stained gels and CPA1 activity after activation with trypsin and CTRC.
Expression plasmids, mutagenesis, adenovirus
Construction of the pcDNA3.1(-) human CPA1 expression plasmid has been reported previously12. The coding DNA in this plasmid was derived from IMAGE clone #3949850 (GenBank accession BC005279), which contains a c.827A>G (p. H276R) alteration. This mistake was corrected by back-mutating Arg276 to His. CPA1 mutants were created by PCR mutagenesis and ligated into the pcDNA3.1(-) expression plasmid. Recombinant adenovirus carrying wild-type CPA1 or the p.N256K mutant was generated by Viraquest (North Liberty, Iowa).
Details regarding the construction of the CPA1 splice-site and duplication mutant expression plasmids are provided in the Supplementary Note.
Cell culture and transfection
HEK 293T cells were cultured in 6-well tissue culture plates (1.5×106 cells per well) in Dulbecco’s Modified Eagle Medium (DMEM) (Invitrogen, Carlsbad CA) supplemented with 10% fetal bovine serum, 4 mM glutamine and 1% penicillin/streptomycin at 37°C. Transfections were carried out at 90% confluence, using 10 μL Lipofectamine 2000 (Invitrogen) and 4 μg expression plasmid in 2 mL DMEM final volume. After overnight incubation, cells were washed and the transfection media was replaced with 2 mL OPTI-MEM I Reduced Serum Medium (Invitrogen). The conditioned OPTI-MEM media were harvested after 48 h incubation. AR42J rat pancreatic acinar cells (American Type Culture Collection #CRL-1492) were maintained in DMEM supplemented with 20% fetal bovine serum, 4 mM glutamine and 1% penicillin/streptomycin at 37°C. Prior to transfection, cells were plated in 6-well plates (106 cells per well) and were grown in the presence of 100 nM concentration of dexamethasone for 48 h to induce differentiation. Infections with adenovirus were performed using 4×107 plaque forming units (pfu) per mL final adenovirus concentrations in a total volume of 1 mL OPTI-MEM in the presence of dexamethasone (100 nM final concentration).
CPA1 activity assay
Enzymatic activity of CPA1 was determined after activation with trypsin and chymotrypsin C (CTRC) using the N-[4-methoxyphenylazoformyl]-L-phenylalanine substrate17, with minor modifications of our previously published conditions12. The CPA1 activity measured in the conditioned medium of transfected cells is referred to as “apparent activity” and reflects the combined effects of the variants on secreted proCPA1 levels, proteolytic degradation during activation and catalytic activity of the activated CPA1. To activate proCPA1, an aliquot (20 μL) of conditioned medium was supplemented with 0.1 M Tris-HCl (pH 8.0), 1 mM CaCl2, 0.05% Tween 20, 100 nM human cationic trypsin and 50 nM human CTRC (final concentrations in 40 μL final volume) and was incubated at 37 °C for 60 min. CPA1 activity was then measured by adding 50 μL assay buffer (0.1 M Tris-HCl (pH 8.0), 1 mM CaCl2, 0.05% Tween 20) and 10 μL substrate (60 μM final concentration) to the activation mix. The decrease in absorbance was followed at 350 nm for 2 min. Rates of substrate cleavage were calculated from fits to the initial linear portion of the curves and were expressed as percent of the wild-type rate, which was set to 100%. The wild-type activity corresponded to 116 ± 34 mOD·min-1 (average ± S.D.), which equals to 262 ± 77 nM·s-1 substrate cleavage rate.
Measurement of proCPA1 secretion
Secreted proCPA1 protein levels in the conditioned medium were determined by SDS-PAGE and densitometry. An aliquot (200 μL) of the medium was precipitated with trichloroacetic acid (10% final concentration), the precipitate was recovered by centrifugation, dissolved in 20 μL Laemmli sample buffer containing 100 mM DTT (final concentration), and heat-denatured at 95 °C for 5 min. Electrophoretic separation was performed on 15% SDS-PAGE mini gels in standard Tris-glycine buffer and gels were stained with Brilliant Blue R-250. Quantitation of bands was carried out with the GelDocXR+ gel documentation system and Image Lab 3.0 software (Bio-Rad, Hercules, CA).
Measurement of ER stress
To study ER stress, we generated recombinant adenovirus carrying either wild-type proCPA1 or the p.Asn256Lys mutant, infected AR42J rat pancreatic acinar cells (#CRL-1492, American Type Culture Collection [ATCC], Manassas, VA) and measured ER stress markers as described below.
Reverse transcriptase (RT)-PCR analysis and real-time PCR
Total RNA was extracted from AR42J cell lysates using RNeasy mini kit (Qiagen, Valencia, CA). RNA was reverse-transcribed using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Carlsbad, CA). X-box binding protein 1 (XBP1) splicing was studied by PCR using a primer set that flanked the spliced region and amplified both spliced and unspliced forms (Supplementary Table 3). PCR was carried out using the Taq DNA Polymerase kit (Qiagen) with the following conditions: 10 min initial denaturation at 95°C followed by 35 cycles of 30 sec denaturation at 95°C, 30 sec annealing at 52°C, 30 sec extension at 72°C and a final extension at 72°C for 5 min. The PCR products were resolved on 2% agarose gels and stained with ethidium bromide. Quantification of mRNA expression was performed by real time PCR (7500 Real Time PCR System, Applied Biosystems). XBP1 expression was measured with SYBR Green (PCR Master Mix, Applied Biosystems) using different primer sets for the spliced, unspliced and total mRNA (Supplementary Table 3). Levels of immunoglobulin-binding protein (BiP) and calreticulin mRNA were determined using TaqMan primers (rat BiP, Rn00565250_m1; rat calreticulin, Rn00574451_m1) with TaqMan Universal PCR Mastermix (Applied Biosystems, Carslbad, CA). Real time PCR conditions were as follows: 2 min equilibration at 50°C, 10 min denaturation and enzyme activation at 95°C followed by 40 two-step cycles of 15 sec at 95°C and 60 sec at 60°C. Gene expression was quantitated using the comparative CT method (ΔΔCT method). Threshold cycle (CT) values were determined using the 7500 System Sequence Detection Software 1.3. Expression levels of target genes were first normalized to the GAPDH internal control gene (ΔCT) and then to expression levels measured in cells infected with empty adenovirus (ΔΔCT). Results were expressed as fold changes calculated with the formula 2-ΔΔCT.
Statistics
The significance of the differences between mutation frequencies in affected individuals and controls were tested by two-tailed Fisher’s Exact Test. Additional odds ratios were calculated using SAS/STAT software (v 9.1) and GraphPad Prism (v 4.03).
Supplementary Material
Acknowledgments
The authors thank all study participants and the members of the Gesellschaft für Pädiatrische Gastroenterologie und Ernährung (GPGE) for providing clinical data and blood samples. The authors also thank Claudia Ruffert (Leipzig), Knut Krohn, Kathleen Schön and Birgit Oelzner (IZKF core unit DNA technologies, Leipzig), Vera Sahin-Tóth (Boston) and K Radha Mani (CCMB, Hyderabad) for excellent technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft (DFG) grants Wi 2036/2-2 & Wi 2036/2-3 (to H.W.) and RO 3929/1-1 & RO 3939/2-1 (to J.R.), the Else Kröner-Fresenius-Foundation (EKFS) (to H.W.), by a grant of the Colora Stiftung gGmbH (to J.R.), NIH grants R01DK058088, R01DK082412; R01DK082412-S2 and R01DK095753 (to M.S.-T.), fellowships from the Rosztoczy Foundation (to M. Bence and A. Schnúr), the Bolyai postdoctoral fellowship from the Hungarian Academy of Sciences (to R.S.), the Institut National de la Santé et de la Recherche Médicale (INSERM), the Programme Hospitalier de Recherche Clinique (PHRC R 08-04), the French Association des Pancréatites Chroniques Héréditaires and her President Mme Meslet, by the Czech Ministry of Health - Conceptual development project of research organization University Hospital Motol in Prague (00064203) and CZ.2.16/3.1.00/24022OPPK (to M.M.), by the Council of Scientific and Industrial Research (CSIR), Ministry of Science and Technology, Govt of India, India grant GENESIS (to G.R.C.), and by Grant-in-Aid from Japan Society for the Promotion of Science (#23591008 to A.M.) and and by the Research Committee of Intractable Pancreatic Diseases provided by the Ministry of Health, Labour, and Welfare of Japan (to A. M. and T.S).
Dedicated to Rudolf Ammann in commemoration of his 87th anniversary.
Footnotes
AUTHOR CONTRIBUTIONS
H.W. and M.S.-T. conceived, designed and directed the study. G.R.C., J.-M.C., J.R., A.M. and H.W. designed, performed and interpreted genetic analyses with significant contributions from D.B., F.B., M.B. (Frankfurt), S.B., C.D., D.L., E.M., S.P., S.S., A.S.-T, K.K., E.N., Y.K., T.S., J.T. and, A.Sc., A.Sz., S.B. (Boston), M.B. (Boston), R.S. and M.S.-T. carried out functional characterization of CPA1 variants. H.W., M.S.-T. and S.B. (Boston) wrote the manuscript with significant contributions from G.R.C., J.-M.C., J.R. and A.M. O.L. provided oligonucleotides. All other co-authors recruited study subjects, collected clinical data and provided genomic DNA samples. All authors approved the final manuscript and contributed critical revisions to its intellectual content.
Competing interests statement:
The authors declare that they have no competing financial interests.
Accession codes. Entrez nucleotide: carboxypeptidase A1 (CPA1): NT_007933.15 (Homo sapiens chromosome 7 genomic contig, GRCh37.p5); NM_001868.2 (human CPA1 mRNA sequence).
References
- 1.Witt H, Apte MV, Keim V, Wilson JS. Chronic pancreatitis: challenges and advances in pathogenesis, genetics, diagnosis, and therapy. Gastroenterology. 2007;132:1557–1573. doi: 10.1053/j.gastro.2007.03.001. [DOI] [PubMed] [Google Scholar]
- 2.Whitcomb DC, et al. Hereditary pancreatitis is caused by a mutation in the cationic trypsinogen gene. Nat Genet. 1996;14:141–145. doi: 10.1038/ng1096-141. [DOI] [PubMed] [Google Scholar]
- 3.Witt H, Luck W, Becker M. A signal peptide cleavage site mutation in the cationic trypsinogen gene is strongly associated with chronic pancreatitis. Gastroenterology. 1999;117:7–10. doi: 10.1016/s0016-5085(99)70543-3. [DOI] [PubMed] [Google Scholar]
- 4.Le Maréchal C, et al. Hereditary pancreatitis caused by triplication of the trypsinogen locus. Nat Genet. 2006;38:1372–1374. doi: 10.1038/ng1904. [DOI] [PubMed] [Google Scholar]
- 5.Witt H, et al. Mutations in the gene encoding the serine protease inhibitor, Kazal type 1 are associated with chronic pancreatitis. Nat Genet. 2000;25:213–216. doi: 10.1038/76088. [DOI] [PubMed] [Google Scholar]
- 6.Chandak GR, Idris MM, Reddy DN, Bhaskar S, Sriram PV, Singh L. Mutations in the pancreatic secretory trypsin inhibitor gene (PSTI/SPINK1) rather than the cationic trypsinogen gene (PRSS1) are significantly associated with tropical calcific pancreatitis. J Med Genet. 2002;39:347–351. doi: 10.1136/jmg.39.5.347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Rosendahl J, et al. Chymotrypsin C (CTRC) variants that diminish activity or secretion are associated with chronic pancreatitis. Nat Genet. 2008;40:78–82. doi: 10.1038/ng.2007.44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Masson E, Chen JM, Scotet V, Le Maréchal C, Férec C. Association of rare chymotrypsinogen C (CTRC) gene variations in patients with idiopathic chronic pancreatitis. Hum Genet. 2008;123:83–91. doi: 10.1007/s00439-007-0459-3. [DOI] [PubMed] [Google Scholar]
- 9.Witt H, et al. A degradation-sensitive anionic trypsinogen (PRSS2) variant protects against chronic pancreatitis. Nat Genet. 2006;38:668–673. doi: 10.1038/ng1797. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Whitcomb DC, et al. Common genetic variants in the CLDN2 and PRSS1-PRSS2 loci alter risk for alcohol-related and sporadic pancreatitis. Nat Genet. 2012;44:1349–1354. doi: 10.1038/ng.2466. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Vendrell J, Querol E, Avilés FX. Metallocarboxypeptidases and their protein inhibitors. Structure, function and biomedical properties. Biochim Biophys Acta. 2000;1477:284–298. doi: 10.1016/s0167-4838(99)00280-0. [DOI] [PubMed] [Google Scholar]
- 12.Szmola R, Bence M, Carpentieri A, Szabó A, Costello CE, Samuelson J, Sahin-Tóth M. Chymotrypsin C is a co-activator of human pancreatic procarboxypeptidases A1 and A2. J Biol Chem. 2011;286:1819–1827. doi: 10.1074/jbc.M110.187369. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Scheele G, Bartelt D, Bieger W. Characterization of human exocrine pancreatic proteins by two-dimensional isoelectric focusing/sodium dodecyl sulfate gel electrophoresis. Gastroenterology. 1981;80:461–473. [PubMed] [Google Scholar]
- 14.Rosendahl J, et al. CFTR, SPINK1, CTRC and PRSS1 variants in chronic pancreatitis: is the role of mutated CFTR overestimated? Gut. 2013;62:582–592. doi: 10.1136/gutjnl-2011-300645. [DOI] [PubMed] [Google Scholar]
- 15.Kereszturi E, Szmola R, Kukor Z, Simon P, Weiss FU, Lerch MM, Sahin-Tóth M. Hereditary pancreatitis caused by mutation-induced misfolding of human cationic trypsinogen: a novel disease mechanism. Hum Mutat. 2009;30:575–582. doi: 10.1002/humu.20853. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Beer S, Zhou J, Szabó A, Keiles S, Chandak GR, Witt H, Sahin-Tóth M. Comprehensive functional analysis of chymotrypsin C (CTRC) variants reveals distinct loss-of-function mechanisms associated with pancreatitis risk. Gut. 2013;62:1616–1624. doi: 10.1136/gutjnl-2012-303090. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Mock WL, Liu Y, Stanford DJ. Arazoformyl peptide surrogates as spectrophotometric kinetic assay substrates for carboxypeptidase A. Anal Biochem. 1996;239:218–222. doi: 10.1006/abio.1996.0318. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.