Human pancreatic trypsinogens undergo post-translational sulfation on Tyr154, catalysed by the Golgi-resident enzyme tyrosylprotein sulfotransferase 2. Sequence alignments suggest that the sulfation of Tyr154 is facilitated by a unique sequence context which is characteristically found in primate trypsinogens. In the search for genetic variants that might alter this sulfation motif, we identified a single nucleotide polymorphism (c.457G>C) in the PRSS2 (serine protease 2, human anionic trypsinogen) gene, which changed Asp153 to a histidine residue (p.D153H). The p.D153H variant is common in subjects of African origin, with a minor allele frequency of 9.2%, whereas it is absent in subjects of European descent. We demonstrate that Asp153 is the main determinant of tyrosine sulfation in anionic trypsinogen, as both the natural p.D153H variation and the p.D153N mutation result in a complete loss of trypsinogen sulfation. In contrast, mutation of Asp156 and Glu157 only slightly decrease tyrosine sulfation, whereas mutation of Gly151 and Pro155 has no effect. With respect to the biological relevance of the p.D153H variant, we found that tyrosine sulfation had no significant effect on the activation of anionic trypsinogen or the catalytic activity and inhibitor sensitivity of anionic trypsin. Taken together with previous studies, the observations of the present study suggest that the primary role of trypsinogen sulfation in humans is to stimulate autoactivation of PRSS1 (serine protease 1, human cationic trypsinogen), whereas the sulfation of anionic trypsinogen is unimportant for normal digestive physiology. As a result, the p.D153H polymorphism which eliminates this modification could become widespread in a healthy population.
Trypsinogen is the most abundant digestive proenzyme produced by the pancreas. Trypsin, the active form of trypsinogen, plays a central role in digestive physiology as the universal activator of digestive proenzymes. Trypsin also facilitates the digestion of dietary proteins by cleaving arginyl and lysyl peptide bonds. In the human pancreas, trypsinogen is expressed as three isoforms, which exhibit close to 90% identity in their primary amino-acid sequences. The three isoenzymes are encoded by separate genes, PRSS1 (serine protease 1, human cationic trypsinogen) and PRSS2 (serine protease 2, human anionic trypsinogen) on chromosome 7 and PRSS3 (serine protease 3, human mesotrypsinogen) on chromosome 9. The common biochemical names reflect the relative isoelectric points of the zymogens. Cationic and anionic trypsinogen represent >90–97% of total trypsinogen present in the pancreatic juice .
The two major human trypsinogen isoforms undergo post-translational sulfation on Tyr154 . This modification is catalysed by the Golgi-resident enzyme TPST (tyrosylprotein sulfotransferase) 2, which is highly expressed in the human pancreas [3–5]. Trypsinogen sulfation seems to be facilitated by the unique sequence context of Tyr154, which is characteristically found in primate trypsinogens . Mesotrypsinogen is also likely to be sulfated, but experimental evidence for this is lacking. Trypsinogen sulfation was first described by Scheele et al.  who incubated human pancreatic slices with [35S]Na2SO4 and demonstrated 35S incorporation into trypsinogens by two-dimensional gel electrophoresis. Subsequently, the crystal structure of native human cationic trypsin indicated the presence of a modification on Tyr154, which was erroneously described as phosphorylation . Conclusive evidence that human trypsinogens are sulfated on Tyr154 came from our previous study, in which we isolated the modified tyrosine amino acid from hydrolysed pancreatic trypsinogens and showed that it was tyrosine sulfate and not tryrosine phosphate . Furthermore, we demonstrated incorporation of 35S from [35S]Na2SO4 into cationic trypsinogen expressed by transfected HEK-293T cells (human embryonic kidney cells expressing the large T-antigen of simian virus 40), and labelling was abolished by mutation of Tyr154 to a phenylalanine residue. We also found that tyrosine sulfation stimulated the autoactivation of human cationic trypsinogen. More recent studies using MS confirmed tyrosine sulfation of pancreatic trypsinogens and also showed that trypsinogen expressed by tumours is not sulfated . It is interesting to note that pancreatic trypsinogens appear to be completely sulfated, as judged by two-dimensional gel electrophoresis, native gel electrophoresis or MS [2,6,7,8].
In the present study, we characterized the sequence determinants that promote the sulfation of Tyr154 on human trypsinogens. Furthermore, we identified the c.457G>C (p.D153H) alteration in PRSS2 as a common polymorphic variant in subjects of African origin. We demonstrated that Asp153 is the main determinant of tyrosine sulfation in anionic trypsinogen, and the p.D153H variation causes a loss of trypsinogen sulfation without any appreciable effect on anionic trypsinogen function.
MATERIALS AND METHODS
The study was approved by the medical ethical review committee of the Charité University Hospital. All patients gave informed consent for genetic analysis. The study population included 1305 healthy German subjects and 198 German patients suffering from chronic pancreatitis with either idiopathic or hereditary etiology. The clinical diagnosis of chronic pancreatitis was based on criteria published previously . To determine the frequency of p.D153H in non-Europeans, we studied 201 Native Americans from Ecuador and 946 unrelated individuals of African descent originating from Benin (n=161, 52 females, 106 males, 3 with unknown gender, age 3–80 years, 88 of unknown age, median 31 years old; mean 31 years old); Cameroon (n=410, 303 females, 107 males, age 10–79 years, 1 of unknown age, median 28 years old, mean 32 years old), Ethiopia (n=155, 80 females, 75 males, age unknown) and Ecuador (n=220, 107 females, 113 males; age 6–80 years, 2 of unknown age, median 31 years old, mean 34 years old). All non-European subjects were recruited for genetic population studies between 1986–1990 (subjects originating from Cameroon) and 1990–1998 (subjects originating from Benin, Ethiopia and Ecuador).
We performed melting-curve analysis to detect the p.D153H variant using a pair of FRET (fluorescence resonance energy transfer) probes, and the LightCycler (Roche Diagnostics). Primers and PCR conditions for exon 4 of PRSS2 have been published previously . The p.D153H sensor probe was 5′-LC-TCTGGGTAGTGGGCTGTGAGGAT-ph (LC, LightCycler Red 705 attached to 5′-terminus; ph, 3′-phosphate) and the anchor probe was 5′-CAGCACAGGAGCATCCAGGCACTGCAGCTC-FL (FL, 5,6-carboxyfluorescein attached to 3′-O-ribose). The primers and probes were designed and synthesized by TIB MOLBIOL.
Plasmid construction and mutagenesis
The PRSS2 coding sequence (GenBank® Nucleotide Sequence Database accession number NM_002770.2) was PCR amplified from the GeneStorm clone H-M27602M (Invitrogen) using the Hu2-XbaI sense primer [5′-TTTTTTTCTAGACACACTCTACCACCATGAATCTACTTCTGATCCT-3′ (where the XbaI restriction site is underlined)] and the Hu2-BamHI antisense primer [5′-TTTTTTGGATCCGGACCAGGGGCTTTAGCTGTTGGCAGCTATG-3′ (where the BamHI restriction site is underlined)]. The PCR product was digested with XbaI and BamHI and ligated into the pcDNA 3.1(−) expression vector. DNA sequencing revealed that compared with the GenBank reference entry, our PRSS2 sequence contained a known polymorphic variant in Ala90 (c.270A>G) and an unexpected mutation in codon 91 (c.272C>T; p.A91V), which was also present in the GeneStorm clone used as the template for PCR amplification. This mistake was subsequently fixed by overlap extension PCR mutagenesis. Missense mutations in the sulfation motif of PRSS2 were generated by overlap extension PCR and were cloned into the pcDNA 3.1(−) expression plasmid.
Construction of the pcDNA3.1(−) expression plasmid containing the PRSS1 gene has been described previously . The proteincoding portion of the TPST2 cDNA (NM_003595.3) was PCR amplified from IMAGE clone #4857366 (BC017509) using the TPST2 XhoI sense primer [5′-CCCTGCCTCGAGGCCACCATGGGCCTGTCGGTGCGGAGG-3′ (where the XhoI restriction site is underlined)] and the TPST2 BamHI antisense primer [(5′-GAGATCGGATCCTCACGAGCTTCCTAAGTGGGAGGAGGT-3′ (where the BamHI restriction site is underlined)]. The sense primer changes the 5′-upstream sequence (CCCAGC>GCCACC) and codon 2 (CGC>GGC; Arg2>Gly) to create an optimal Kozak sequence. The PCR product was digested with XhoI and BamHI and cloned into the pcDNA3.1(−) plasmid.
The coding region of the TPST1 cDNA (NM_003596.2) was PCR amplified from an Incyte full-length cDNA clone (clone ID LIFESEQ3598404) using the TPST1 XhoI sense primer [5′-CAAGATCTCGAGGCCACCATGGTTGGAAAGCTGAAGCAGAACTTACTA-3′ (where the XhoI restriction site is underlined)] and the TPST1 BamHI antisense primer [5′-CTCCTGGGATCCCTACTCCACTTGCTCAGTCTGTGG-3′ (where the BamHI restriction site is underlined)]. The sense primer changes the 5′-upstream sequence to an optimal Kozak sequence (ATCAAG>GCCACC). The PCR product was digested with XhoI and BamHI and cloned into the pcDNA3.1(−) plasmid.
Cell culture and transfection
To study trypsinogen sulfation, we used HEK-293T cells because this cell line does not express endogenous trypsinogens. HEK-293T cells were cultured in 6-well tissue-culture plates (106 cells per well) in DMEM (Dulbecco's modified Eagle's medium) supplemented with 10% (v/v) fetal bovine serum, 4 mM glutamine and 1% penicillin/streptomycin at 37 °C in a humidified atmosphere containing 5% CO2. Transfections were carried out in a total volume of 2 ml of DMEM using 10 μl of Lipofectamine™ 2000 (Invitrogen) and 2 μg of wild-type or mutant PRSS2 plasmids and 0.5 μg of TPST1 or TPST2 plasmids, unless indicated otherwise. After overnight incubation at 37 °C, the transfection medium was replaced with 2 ml of OptiMEM medium containing 1 mM benzamidine to prevent autoactivation of secreted trypsinogen.
Expression and purification of anionic trypsinogen
For larger scale expression of sulfated and non-sulfated anionic trypsinogen, HEK-293T cells were cultured in 75 cm2 flasks as described above and transfections were performed in 20 ml of DMEM using 75 μl of Lipofectamine™ 2000 with 24 μg of the PRSS2 plasmid with or without 6 μg of the TPST2 plasmid. Cells were incubated overnight at 37 °C, and the transfection medium was then replaced with 20 ml of OptiMEM medium containing 1 mM benzamidine. Conditioned medium was collected after 32 h, supplemented with Tris/HCl (pH 8.0) to a final concentration of 0.1 M and anionic trypsinogen was purified by ecotin affinity chromatography as described previously .
Expression and purification of cationic trypsinogen
The procedure to express and purify sulfated and non-sulfated human cationic trypsinogen was essentially the same as described above for anionic trypsinogen, with the following modifications. We found that HEK-293T cells express wild-type cationic trypsinogen at 5–10-fold lower levels than anionic trypsinogen. To increase expression, we used the p.K237D/p.N241D cationic trypsinogen variant, which is expressed almost 3-fold better than wild type (E. Kereszturi and M. Sahin-Tóth, unpublished work). In contrast with anionic trypsinogen, cationic trypsinogen was partly sulfated by HEK-293T cells, therefore we included 50 mM sodium chlorate in the growth medium when our aim was to purify non-sulfated cationic trypsinogen. Because cationic trypsinogen was autoactivated in the growth medium even in the presence of 1 mM benzamidine, we collected the conditioned medium every 8 h and replaced it with fresh OptiMEM medium.
Trypsin activity assay
Aliquots (10 μl) of conditioned medium were mixed with 1 μl of 1 M Tris/HCl (pH 8.0), 1 μl of 10 mM CaCl2 and 1 μl of human enteropeptidase (1.4 μg/ml; R&D Systems) and incubated at 37 °C for 1 h. Trypsin activity was then measured using the synthetic chromogenic substrate, N-Cbz (benzyloxycarbonyl)–Gly–Pro–Arg–p-nitroanilide at a final concentration of 0.14 mM. Time courses (1 min) of p-nitroaniline release were followed at 405 nm in 0.1 M Tris/HCl (pH 8.0) and 1 mM CaCl2 at room temperature (22 °C) in a Spectramax Plus 384 microplate reader (Molecular Devices).
Native gel electrophoresis
To analyse trypsinogens by PAGE under non-denaturing conditions, 400 μl of conditioned medium was precipitated with 10% (w/v) trichloroacetic acid. Samples were centrifuged at 16000 g for 10 min in an Eppendorf microcentrifuge and pellets were resuspended in native loading buffer devoid of SDS or reducing agents. Samples were then electrophoresed on 13% Tris/glycine gels. The gel apparatus was placed on ice and the gels were run for 2–2.5 h at 35 mA constant current. Proteins were visualized by Coomassie Blue staining.
SDS/PAGE and Western blotting
Conditioned medium was precipitated with 10% (w/v) trichloroacetic acid (final concentration), resuspended in Laemmli sample buffer with 100 mM dithiothreitol, heat denatured by incubating at 95 °C for 5 min and resolved by SDS/PAGE (15% gels). The gels were stained with Coomassie Blue. For immunodetection of sulfated trypsinogens, 30 μl of conditioned medium was precipitated and resolved by SDS/PAGE (15% gels). The proteins were then transferred on to an Immobilon-P membrane (Millipore) at 300 mA for 1.5 h. The membrane was blocked overnight at 4 °C with 5% (w/v) non-fat dried skimmed milk powder dissolved in PBST (PBS with 0.1% Tween 20) and then incubated in the same solution for 1 h at room temperature with an anti-(tyrosine-sulfate) IgG [12,13] added at a dilution of 1:2000. After three 10 min washes with PBST, the membrane was blocked again for 1 h at room temperature and the HRP (horseradish peroxidase)-conjugated rabbit polyclonal secondary antibody against human IgG (Abcam, catalogue number ab6759) was applied at a dilution of 1:10000 for 1 h. After three 10 min PBST washes, HRP was detected using the SuperSignal West Pico Chemiluminescent Substrate (Pierce).
The p.D153H variant alters the sulfation motif in human anionic trypsinogen
Recognition of the target tyrosine amino acid by sulfotransferases is aided by the sequence context, commonly referred to as the sulfation motif. Although there is no consensus sequence that describes this motif, certain amino acids are strongly preferred or excluded, as summarized in Figure 1 [14–17]. Thus an acidic amino acid (aspartic acid or glutamic acid) is frequently found adjacent to the sulfated tyrosine residue, and at least three acidic residues are usually present within the five amino acids flanking the tyrosine residue on either side. Turn-forming amino acids proline and glycine are often seen in sulfation motifs. Cysteine and basic amino acids are generally excluded, and hydrophobic amino acids are rare. With the exception of the cysteine residue at position + 6, the trypsinogen sulfation sequence conforms to these requirements. In a search of the NCBI database (http://www.ncbi.nlm.nih.gov/sites/entrez) for genetic variants that might alter the trypsinogen sulfation motif, we identified the c.457G>C (rs1804564a) single nucleotide polymorphism in the PRSS2 gene, which changes Asp153 to a histidine residue (p.D153H). To determine the prevalence of this variant, we screened 1305 healthy German subjects and 198 German patients with chronic pancreatitis. Surprisingly, we found the p.D153H variant only in a single healthy individual whose mother was German and father was of African descent. Next, we analysed 946 subjects of African origin from four different countries and 201 Native Americans from Ecuador, and found the p.D153H variant with an average minor allele frequency of 9.2% in subjects of African origin, but not in Native Americans (Table 1). Thus p.D153H is a common African polymorphism. The p.D153H variant changes the aspartic acid residue closest to Tyr154 within the putative sulfation motif, and thus it is expected to have a significant impact on tyrosine sulfation of anionic trypsinogen.
The sulfation motif in human anionic trypsinogen
|Country||N||Homozygote wild-type (%)||Heterozygote (%)||Homozygote p.D153H (%)|
|Country||N||Homozygote wild-type (%)||Heterozygote (%)||Homozygote p.D153H (%)|
Tyrosine sulfation of human anionic trypsinogen by TPST2 in HEK-293T cells
To study tyrosine sulfation of anionic trypsinogen by TPST2 in a quantitative manner, we established a cellular system in which trypsinogen and sulfotransferase are co-expressed from separate plasmids, thus allowing independent manipulation of either protein. HEK-293T cells were co-transfected with 2 μg of the PRSS2 plasmid and increasing amounts (0, 2.5, 8, 25, 80 and 250 ng) of the TPST2 plasmid. Conditioned medium containing secreted anionic trypsinogen was analysed by native PAGE. Because tyrosine sulfation introduces an extra negative charge, sulfated anionic trypsinogen migrates faster on native gels than its non-sulfated counterpart. As Figure 2(A) demonstrates, with increasing amounts of the TPST2 plasmid, the slower mobility non-sulfated anionic trypsinogen form is gradually converted into the higher mobility sulfated form. Complete sulfation was observed when 250 ng of the TPST2 plasmid was used for co-transfection. The experiment also demonstrates that endogenous TPST activity in HEK-293T cells is low, as essentially no tyrosine sulfation occurs in the absence of added TPST2 plasmid. To confirm these findings with an independent method, Western blot analysis was carried out using an antibody that specifically recognizes tyrosine sulfate [12,13]. Figure 2(B) demonstrates that in the absence of the TPST2 plasmid no immunosignal was detected, whereas the inclusion of increasing amounts of the TPST2 plasmid resulted in increasing levels of immunoreactivity, which perfectly correlated with the amount of sulfated anionic trypsinogen observed on native gels. Co-transfection with higher amounts (≥1 μg) of the TPST2 plasmid resulted in a marked suppression of the secretion of anionic trypsinogen (results not shown). This phenomenon was unrelated to the sulfation of anionic trypsinogen, as production of β-galactosidase was also suppressed by co-transfection with TPST2 under similar conditions. On the basis of these results, in our subsequent experiments, we chose to use 0.5 μg of the TPST2 plasmid, which provided complete sulfation without affecting protein secretion.
Tyrosine sulfation of human anionic trypsinogen in HEK-293T cells by TPST2
Asp153 is the critical determinant of tyrosine sulfation in anionic trypsinogen
To investigate the significance of the p.D153H variant and the amino acids neighbouring Tyr154 in tyrosine sulfation, we subjected this region to mutational analysis. We mutated Tyr154 to a phenylalanine residue (p.Y154F), Asp153 to a histidine residue (p.D153H) and an asparagine residue (p.D153N), Asp156 to an asparagine residue (p.D156N) and Glu157 to a leucine residue (p.E157L), because in the majority of mammalian trypsinogens Leu is found at this position. Furthermore, Gly151 and Pro155 were mutated to alanine residues (p.G151A and p.P155A). HEK-293T cells were transfected with wild-type and mutant PRSS2 plasmids and co-transfected with the TPST2 plasmid. Conditioned medium was analysed by reducing SDS/PAGE, trypsin-activity assays and Western blotting. The Coomassie-Blue-stained gel in Figure 3(A) demonstrates that wild-type and mutant trypsinogens were secreted at similar levels. Note that the sulfated and non-sulfated forms are not separated by SDS/PAGE due to the denaturing conditions. Trypsin activity of wild-type and mutant trypsinogens was also comparable when measured after activation with enteropeptidase, indicating that the mutations have no detrimental effect on the catalytic properties of the enzyme (Figure 3B). On the other hand, Western blot analysis revealed that the trypsinogen mutants were sulfated to an extent that varied considerably among the mutants (Figure 3C). As expected, removal of the phenolic hydroxyl group from Tyr154 by the p.Y154F mutation abolished sulfation. A striking loss of sulfation was also observed when the negative charge on the adjacent Asp153 residue was eliminated by p.D153N or p.D153H mutations. p.D156N and p.E157L mutants exhibited decreased sulfation, indicating that although Asp156 and Glu157 are not essential, the two acidic residues increase the efficiency of this modification. On the other hand, the sulfation of p.G151A and p.P155A mutants was unchanged relative to wild-type anionic trypsinogen.
Mutational analysis of the sulfation motif in human anionic trypsinogen
To confirm that the changes detected in immunoreactivity reflect different sulfation levels, we analysed tyrosine sulfation of the trypsinogen mutants on native gels. In this experiment, PRSS2 mutants were expressed in HEK-293T cells with or without co-transfection with the TPST2 plasmid. Conditioned medium was then electrophoresed under non-denaturing conditions on Tris/glycine gels. As shown in Figure 3(D), the results are in perfect agreement with the Western blots shown in Figure 3(C). Thus wild-type trypsinogen and p.G151A and p.P155A mutants were completely sulfated and p.D156N and p.E157L mutants were partially sulfated, whereas no sulfation was detectable on p.D153N, p.D153H and p.Y154F mutants.
Although TPST2 is the sulfotransferase isoform responsible for trypsinogen sufation in the pancreas [3,5], we wondered how well TPST1 would catalyse the sulfation of the trypsinogen mutants. As judged by Western blot analysis, co-transfection with the TPST1 plasmid resulted in the sulfation of wild-type anionic trypsinogen, which was abolished by p.D153N, p.D153H and p.Y154F mutations, and moderately decreased by p.D156N and p.E157L mutations (see Supplementary Figure S1 at http://www.BiochemJ.org/bj/418/bj4180155add.htm). The results indicate that determinants of the sulfation motif on anionic trypsinogen are recognized by TPST1 and TPST2 in a similar manner, suggesting that natural selection of TPST2 overexpression in the human pancreas was not driven by substrate specificity.
Biochemical characteristics of sulfated and non-sulfated PRSS2
We demonstrated previously that native sulfated cationic trypsinogen purified from human pancreatic juice exhibited faster autoactivation than its non-sulfated recombinant counterpart purified from Escherichia coli . These observations suggested that tyrosine sulfation can facilitate trypsinogen activation. To confirm and extend these experiments, we purified sulfated and non-sulfated anionic and cationic trypsinogen from conditioned medium from HEK-293T cells transfected with PRSS1 and PRSS2 plasmids and with or without the TPST2 plasmid. Autoactivation experiments were carried out at pH 8.0, in the presence of 1 mM and 10 mM CaCl2 (Figure 4). As reported previously, autoactivation of cationic trypsinogen was stimulated by tyrosine sulfation, particularly in the high calcium milieu . In contrast, autoactivation of anionic trypsinogen was slightly inhibited by tyrosine sulfation.
Autoactivation of sulfated and non-sulfated forms of human anionic and cationic trypsinogens
We also compared the activation, catalytic activity, inhibitor sensitivity and degradation of sulfated and non-sulfated anionic trypsin(ogen) in a variety of biochemical assays, which are detailed and illustrated in Supplementary Table S1(at http://www.BiochemJ.org/bj/418/bj4180155add.htm) and Supplementary Figures S1–S6 (at http://www.BiochemJ.org/bj/418/bj4180155add.htm). Small differences were observed in inhibitor sensitivity and degradation by chymotrypsin C; however, none of these appeared to be biologically significant.
Tyrosine-O-sulfation is a ubiquitous post-translational modification of secretory and membrane proteins [18,19]. It is catalysed by two isoforms of TPST, TPST1 and TPST2, which are localized in the trans-Golgi system [3,4,20]. The amino-acid sequences of the two TPST enzymes are 64% identical and their substrate specificity overlaps, although their affinities for some synthetic substrates are slightly different . Both isoforms are expressed in all tissues at varying levels; in humans, the highest TPST1 mRNA expression was measured in the testis, whereas the TPST2 mRNA transcript is most abundant in the pancreas [3,5,20,21]. To address the biological role of tyrosine sulfation, mouse strains deficient in TPST1, TPST2 or both isoenzymes were engineered [22–24]. Selective disruption of the TPST1 gene resulted in reduced body mass and increased post-implantation fetal death, whereas TPST2-knockout mice were infertile, hypothyroid and exhibited delayed growth. Double knockout mice died in the early postnatal period with signs of cardiopulmonary insufficiency. The undersulfated target proteins responsible for the phenotypic changes in the knockout mice have not yet been identified. Hypothyroidism was also observed in a dwarf mouse strain which carried the natural p.H266Q mutation in the TPST2 gene . The authors proposed that defective sulfation of the thyroid-stimulating hormone receptor was the direct cause of the phenotype.
Biochemical studies on a handful of tyrosine-sulfated proteins led to the current consensus that tyrosine sulfation modulates extracellular protein–protein interactions. For example, the sulfation of P-selectin glycoprotein ligand-1 present on leucocytes is necessary for binding to P-selectin on the endothelial cells during adhesion . Similarly, tyrosine sulfation of endoglycan on endothelial cells is required for binding to L-selectin on lymphocytes during homing . The sulfation of the three tyrosine residues on platelet glycoprotein Ibα promotes its interaction with von Willebrand factor, which is essential for the initiation of haemostasis . The sulfation of Tyr1680 in coagulation factor VIII is important for binding to von Willebrand factor, and the p.Y1680F missense mutation (Y1699F when counted from the initiator methionine residue) causes a mild form of haemophilia A [29,30]. Finally, the sulfation of tyrosine residues in the N-terminus of CC chemokine receptor 5 promotes chemokine binding and facilitates HIV-1 entry [31,32].
In the present study, we characterized the amino-acid requirements for tyrosine sulfation of human trypsinogens. On the basis of sequence alignments, we suggested previously that human trypsinogens are sulfated because of a special sequence motif around Tyr154 . Mutational analysis of this sulfation motif confirmed that Asp153 is the critical determinant, but Asp156 and Glu157 are also important for efficient trypsinogen sulfation. In contrast, the turn-forming residues Gly151 and Pro155 within the sulfation motif are dispensable. Our findings are in agreement with studies published previously demonstrating that acidic residues, particularly in the immediate vicinity of the target tyrosine residue, are important for tyrosine sulfation (see Figure 1) [14–17]. Since Asp153 is absent in most vertebrate trypsinogens, our results confirm that trypsinogen sulfation is largely primate specific. Interestingly, trypsinogen genes in the recently sequenced genome of the laboratory opossum (Monodelphis domestica; XM_001362358.1 and XM_001362438.1) contain amino acids that correspond to Asp153, Tyr154, and Asp156, although Glu157 is missing. Furthermore, a newly sequenced platypus (Ornithorhynchus anatinus) trypsinogen gene (XM_001518173.1) also contains amino acids which correspond to Asp153 and Tyr154, which can be viewed as the minimal requirement for trypsinogen sulfation. In all likelihood, the presence of the sulfation motif in the trypsinogen genes of monotremes, marsupials and primates is the result of convergent evolution. Alternatively, trypsinogen sulfation may have begun to evolve in their common ancestor and was subsequently lost in all placental mammals with the exception of primates.
We identified a common polymorphism in the PRSS2 gene that altered Asp153 and abolished the sulfation of anionic trypsinogen. The p.D153H variant was found in subjects of African origin with an average minor allele frequency of 9.2%, whereas it was absent in subjects of European descent or in Native Americans originating from Ecuador. Variants of the PRSS1 gene were shown previously to cause hereditary chronic pancreatitis, and a variant of the PRSS2 gene was shown to afford protection against chronic pancreatitis [9,33]. Thus variants in human trypsinogen genes can play important pathological roles in defining the risk for chronic pancreatitis. Owing to the unavailability of African subjects with chronic pancreatitis, we could not study the potential disease association of the p.D153H variant, but the high frequency of the alteration suggests that p.D153H cannot be a significant risk factor for chronic pancreatitis, which is a relatively rare disease with ∼0.01–0.1% prevalence.
The functional significance of post-translational sulfation in human trypsinogens was first addressed previously, where we showed that autoactivation of human cationic trypsinogen was stimulated by tyrosine sulfation . These observations have been confirmed in the present study. Surprisingly, however, the stimulatory effect of sulfation on autoactivation proved to be isoform specific, as we found that autoactivation of anionic trypsinogen was inhibited slightly by this modification. Furthermore, the catalytic activity and inhibitor sensitivity of anionic trypsin were also unaffected by tyrosine sulfation. Taken together, we conclude that the functional role of trypsinogen sulfation in humans is to stimulate the autoactivation of cationic trypsinogen, whereas the sulfation of anionic trypsinogen appears to be inconsequential, at least with respect to autoactivation and other enzymatic properties examined in the present study. The prevalence of the p.D153H variant in subjects of African origin also seems to support the contention that tyrosine sulfation of anionic trypsinogen is redundant; therefore a polymorphism which abolishes this modification could become widespread in a healthy population.
We thank Éva Kereszturi (Department of Molecular and Cell Biology, Boston University Goldman School of Dental Medicine, Boston, MA, U.S.A.) for the PRSS1 expression plasmid carrying the p.K237D/p.N241D mutations, Zoltán Kukor (Department of Molecular and Cell Biology, Boston University Goldman School of Dental Medicine, Boston, MA, U.S.A.) for initiating functional experiments on the p.D153H variant, Richard Szmola (Department of Molecular and Cell Biology, Boston University Goldman School of Dental Medicine, Boston, MA, U.S.A.) for providing recombinant pro-carboxypeptidase B1 and chymotrypsin C, Olfert Landt (TIB MOLBIOL, Berlin, Germany) for designing FRET probes, and Markus Braun and Claudia Güldner (Department of Hepatology and Gastroenterology, Charité, Berlin, Germany) for excellent technical assistance. M. S.-T. thanks Kevin L. Moore (Oklahoma Medical Research Foundation, Oklahoma City, OK, U.S.A.) for helpful discussions.
Dulbecco's modified Eagle's medium
fluorescence resonance energy transfer
cell, human embryonic kidney cell expressing the large T-antigen of simian virus 40
PBS with 0.1% Tween 20
serine protease 1, human cationic trypsinogen
serine protease 2, human anionic trypsinogen
serine protease 3, human mesotrypsinogen
This work was supported by the National Institutes of Health [grant numbers AA014544, DK058088] (to M. S.-T.); and the Deutsche Forschungsgemeinschaft [grant number Wi2036/2-2] (to H. W.).
Present address: Department of Pediatrics and Else Kröner-Fresenius-Zentrum (EKFZ), Technische Universität München (TUM), 80804 Munich, Germany.