Cellular Origin, Complementary Deoxyribonucleic Acid ... - CiteSeerX

BIOLOGY OF REPRODUCTION 53, 227-233 (1995)

Cellular Origin, Complementary Deoxyribonucleic Acid and N-Terminal Amino Acid Sequences of Human Seminal Progastricsin' Pal Bela Szecsi,2 3, 4, Hanne Halgreen, 3 Ricky N.S. Wong 3 Torben Kjxer,5 and Jordan Tang3

Protein Studies Research Program,3 Oklahoma Medical Research Foundation, Oklahoma City, OK Departmentof ClinicalChemistry,4 University of Lund, Malmo GeneralHospital, Sweden Department of Surgery,5 University of Aarhus, Denmark ABSTRACT The aspartic protease progastricsin (EC 3.4.23.3) is found in all parts of the mammalian stomach and has also been found extragastrically. In humans and monkeys, seminal fluid usually contains high concentrations of progastricsin. Using immunohistochemistry and in situ hybridization, we determined in this investigation the origin of seminal progastricsin to be the epithelia of both the prostatic gland and the seminal vesicles. In addition, Northern (RNA) blotting showed the presence of a 1.8-kb transcript in both tissues. Seminal progastricsin clones from two human prostatic gland cDNA libraries were isolated and sequenced. The combined sequence manifested only six nucleotide differences from the published genomic and gastric cDNA sequence. One conservative base substitution was present in both libraries. N-Terminal amino acid sequencing of all 43 residues of the seminal proenzyme and the first 34 residues of the mature enzyme yielded sequences identical to those deduced from cDNAs derived from both gastric and prostatic origin. The results obtained indicate that gastric and seminal progastricsin are products of the same gene and that the observed molecular differences between the zymogen from the two sources are probably due to posttranslational modifications.

INTRODUCTION Human and monkey seminal fluids usually contain high concentrations of the aspartic protease progastricsin (EC 3.4.23.3) (also known as gastric chief cell-synthesized pepsinogen C) [1-6]. The inactive zymogen progastricsin is converted into the active form gastricsin (or pepsin C) by an autocatalytic process at pH below 5.0 [7, 8]. The acidic environment in the vagina is capable of activating seminal progastricsin within an average of 3 h (range 2-7 h) after deposition of semen. This peptic activity persists in the vagina for at least 24 h after coitus [6]. Progastricsin-mediated proteolysis of most seminal fluid proteins occurred after 30 min of incubation at 37°C at pH levels found in the vagina [9]. Prostate-specific antigen, prostate acidic phosphatase, and Zn-a 2-glycoprotein are quickly degraded; albumin, transferrin, and lactoferrin are degraded more slowly. The low-molecular weight fragments of semenogelin I and II and especially -microseminoprotein are somewhat resistant to proteolysis. This suggests that the function of gastricsin is to degrade seminal fluid proteins deposited in the vagina, which in turn may decrease the antigenic load in the vagina and prevent immunoinfertility. The cellular origin of human seminal progastricsin has not been definitely determined. The epithelia of prostatic Accepted March 9, 1995. Received December 31, 1994. 'Financial support was provided by the Danish Medical Research Foundation; the National Institute of Health; the Medical Faculty, University of Lund; and the Alfred Osterlund Foundation. 2Correspondence and current address: Department of Clinical Chemistry, University of Lund, Malm6 General Hospital, 214 01 Malmo, Sweden. FAX: +46, 40 929023.

glands has been suggested to be the sole source of seminal progastricsin [10, 11]. Other researchers have reported progastricsin immunohistochemical staining of both seminal vesicle and prostate epithelia [12]. We have previously found progastricsin activity in extracts of both tissues [13, 14]. By agar gel electrophoresis at pH 6.0, three components of human progastricsin can be found in the gastric mucosa [13]. These components differ from each other by approximately one charge unit and are all converted to gastricsin (pepsin C) with the same electrophoretic mobility. However, only the two electrophoretically faster-moving of the three gastric progastricsin components are observed in semen. A monoclonal antibody is able to discriminate between seminal and gastric progastricsin, indicating structural differences [15]. The human gastric and seminal progastricsins are indistinguishable from each other in amino acid composition and in the 28 N-terminal amino acid residues [15]. The human gastric progastricsin cDNA and amino acid sequences are known from the cloning of its cDNA and gene [16-19]. Several genomic restriction fragments of human progastricsin have been detected by Southern blotting with rat progastricsin cDNA used as probe [16]. A restriction fragment-length polymorphism in the progastricsin gene involving an insertion or deletion of 100 bp of intron sequence has been located between exons 7 and 8 [18, 19]. However, only one gene on chromosome 6p21.1-pter has been localized [19]. In order to clarify the structure and origin of seminal progastricsin, we have isolated and sequenced cDNA clones from two human prostatic gland libraries and performed 227

228

SZECSI ET AL.

immunohistochemical studies and in situ hybridization on tissue from the male reproductive tract. MATERIALS AND METHODS N-Terminal Protein Sequencing and Mass Spectrometry Human seminal progastricsin from pooled semen samples and gastric progastricsin were purified by anionexchange, gel filtration, and fast performance liquid chromatography [15]. Automatic Edman degradation was carried out with an Applied Biosystems 477A Sequencer equipped with an on-line phenylthioydantoin analyzer model 120A (Applied Biosystems, Foster, CA). A matrix-assisted laser desorption/ionization mass spectrometry time-of-flight (MALDI-MS TOF) procedure was performed with a-cyano4-hydroxycinnamic acid as the matrix. Immunohistochemistry and In Situ Hybridization Human prostate glands, seminal vesicles, epididymis, testis, and gastric mucosa were collected at surgery, necropsy, or endoscopy. The tissue specimens were immersion-fixed in Bouin's solution for 4-24 h and embedded in paraffin; 5-1m sections were mounted on silanized slides and dewaxed in xylene. For immunohistochemistry, sections were processed according to the ABC-peroxidase method (Vector, Burlingame, CA) with 3,3'-diaminobenzidine tetrahydrochloride/nickel sulfate as chromogen. The primary antibody, raised in rabbits against human gastric progastricsin, was diluted 1:2000. Sections were incubated overnight at 4°C. The following controls were used: 1) omission of primary antibody, 2) incubation with primary antibody preabsorbed with progastricsin (10 pg/ml and 100 ag/ ml) overnight at 4C, and 3) sections of human gastric mucosa as positive controls. For in situ hybridization, sections were permeabilized with 0.02% porcine pepsin in 0.2 M HCl/5 mM EDTA for 20 min at 37 0C and postfixed in 4% paraformaldehyde for 15 min. The sections were prehybridized for 2 h at 420 C in 4strength saline sodium citrate (single-strength SSC is 0.15 M sodium chloride, 0.015 M sodium citrate, pH 7.0), 10% dextran sulfate, single-strength Denhardt's solution (0.02% [v/ w] Ficoll 400, 0.02% [v/w] polyvinylpyrrolidone, 0.02% [v/ w] bovine albumin [fraction V]), and 15% formamide containing 400 gg/ml heat-denatured salmon sperm DNA. Hybridization was performed overnight in the prehybridization solution with 200 ng/ml of a 3'-digoxigenin-ddUTPlabeled (Boehringer-Mannheim, Mannheim, Germany) synthetic 38 oligomer probe corresponding to bp 120-156 (from the initiation codon) of human progastricsin, and sections were washed twice with 4-strength SSC for 5 min at room temperature, four times in 0.4-strength SSC for 15 min at 48°C, and once in single-strength SSC for 10 min at room

temperature. The sections were then incubated with alkaline phosphatase-conjugated Fab fragments of antidigoxigenin sheep antibody diluted 1:500 (Boehringer-Mannheim) for 2 h at room temperature. Phosphatase staining was performed after a new series of washes with 5-bromo4-chloro-3-indolyl phosphate/nitro blue tetrazolium. Messenger RNA and Northern Blot RNA was extracted with guanidine thiocyanate from human seminal vesicles and prostate gland obtained from organ donors 15 min after circulation arrest at Aarhus University Hospital, Aarhus, Denmark. Poly(A)+ RNA was fractionated by oligo(dT)-cellulose chromatography. Poly(A) + RNA was denatured and subjected to electrophoresis on 1.5% agarose gel containing glyoxal and transferred to a nylon membrane (Hybond-N; Amersham Corp., Arlington Heights, IL). The membrane was hybridized overnight at 65°C to a nick-translated 1.35-kb cDNA of human gastric progastricsin previously obtained from a human stomach cDNA library [20]. Library Construction, Screening, and DNA Sequencing First-strand synthesis was carried out with use of oligo(dT),,,,18 and M-MLV reverse transcriptase (Gibco BRL, Gaithersburg, MD). Double-stranded cDNA was prepared according to the RNAse H method. After methylation of internal EcoRI sites and addition of EcoRI linkers, the cDNA was ligated into the EcoRI site of XgtlO0, in vitro-packaged, and transfected in E. coli strain C600 Hfl. The library containing 7.2 X 105 independent recombinant phages was amplified once. Nylon filter (Hybond-N, Amersham) imprints of - 40 000 phage plaques were screened with the nick-translated 3 2P-labeled 1.35-kb full-length cDNA of progastricsin. Filters were prehybridized in 6-strength SSC containing 0.1% SDS, 5-strength Denhardt's solution, and 100 g/ml heat-denatured salmon sperm DNA for 3 h at 65°C and subsequently hybridized for 18 h at 650C. Hybridizing plaques were identified by autoradiography after washing at high stringency at 65°C in double-strength SSC for 30 min, double-strength SSC containing 0.1% SDS for 45 min, and 0.1-strength SSC for 15 min. The 5' clone was obtained by polymerase chain reaction (PCR) (annealing temperatures: 5 times at 37°C, 5 times at 420C, 5 times at 470C, and 25 times at 55 0C) of reverse-transcriptized poly(A)+ mRNA, with use of 100 pmol of primers corresponding to the 3' end of the longest clone obtained from the XgtlO0 library screening and to nucleotides 1-20 of gastric progastricsin cDNA. The resulting DNA was cut at the primer Sst I site and cloned into pUC19. An additional human prostate kgtll library was obtained (Cat. #HL1051b; Clontech, Palo Alto, CA) and screened as described above. Phages were

SEMINAL PROGASTRICSIN

229

purified by successive plating and hybridization. The EcoRI inserts were cloned into M13mp18 or pUC18. All clones were subcloned and sequenced with use of the chain termination procedure in both directions. General molecular cloning procedures were performed as described [21]. Amino acids and nucleotides are represented according to IUPAC-IUB. RESULTS A Northern blot of mRNA isolated from the seminal vesicles and the prostatic gland showed that a transcript of approximately 1.8 kb strongly hybridized to gastric progastricsin cDNA (Fig. 1). On the basis of tissue weight, a stronger hybridizing signal suggested that the seminal vesicles contained more progastricsin mRNA than did the prostate. However, the vesicles also contained more hybridizable degradation products. For this reason, poly(A) + mRNA from the prostatic gland was chosen for the construction of a kgtlO library. Immunohistochemistry of gastric mucosa showed the chief cells and mucous neck cells staining positive for progastricsin in a manner similar to that described by others [22, 231 (data not shown). Progastricsin-positive staining was observed in the epithelia of both seminal vesicles and prostatic glands (Fig. 2, A and B). No reactivity was found in the testis or any part of the epididymis. The amount of positive staining varied greatly both within individual subjects and within each section. Both in the seminal vesicle and prostate gland, as well as in the gastric mucosa control samples, positive staining was completely abolished by preincubation of the antibody with progastricsin. The epithelial cells of some glands of the seminal vesicles manifested intense cytoplasmic staining without apical enhancement, frequently with marked intensity around the nuclear edge. Other glands were completely negative. Within progastricsin-staining glands, staining was variable; occasional single cells might stain very intensely, whereas the neighboring cells were completely unstained (Fig. 2A). Progastricsin-positive cells in the prostatic gland were confined to a few acini, with uneven distribution among the individual cells (Fig. 2B). In contrast to the seminal vesicles, progastricsin staining in the prostatic gland often manifested supra-nuclear apical enhancement. Some cells stained very intensely; sometimes positive cells were interspersed with negative cells. The progastricsin staining was not related to the prostate-specific antigen that stained all glands. Benign hypertrophic nodules and neoplastic tissue manifested no progastricsin staining. In some cases of prostatic carcinoma, an accumulation of progastricsin-positive cells were present at tumor cell borders (Fig. 2C). The in situ hybridization showed that the progastricsin

FIG. 1. Northern blot of human progastricsin mRNA. Ten micrograms of poly(A)+ RNA from human seminal vesicles (S)and prostate gland (P)were denatured, electrophoresed inagarose, blot-transferred to nylon, and hybridized to 32P-labeled human gastric progastricsin cDNA XG16-3. Arrows indicate positions of 1.90-, 1.71-, 1.32-, and 0.93-kb DNA markers.

transcript was localized to all the epitheliums of the seminal vesicles and to scattered epithelial cells of the prostate (Fig. 3, A and B). Protein sequencing of the 43 N-terminal amino acid residues of the seminal progastricsin and the 34 residues of the mature enzyme yielded a sequence identical to that deduced from gastric cDNA and the gtll library (Fig. 4). At position 29, protein sequencing gave a valine, whereas the nucleotide sequence of the PCR-derived clone coded for an

230

SZECSI ET AL.

alanine. The values for molecular weight obtained by mass spectrometry, both for seminal (Mr 41 173) and gastric progastricsin (Mr 41 388), were higher than that deduced from the nucleotide sequence ( 40 562). Despite extensive screening using different conditions, the prostate Xgtl0 library yielded only a 704-bp cDNA clone derived from the 3'-end of the progastricsin cDNA (bp 6071311). The missing 5' end of the progastricsin cDNA was obtained by PCR (bp-50-625). The gtl0 and the PCR clones together represent nearly the entire length of human progastricsin cDNA. The sequences of the clones manifested few nucleotide differences from the published genomic and gastric cDNA sequence of progastricsin, including the 3'-nontranslated region (Fig. 4). Substitutions were present at positions 263 (T/C), 301 (T/A), 846 (A/C), and 1304 (C/A). Only two of the substitutions gave changes in amino acids-29valine to 29alanine and 42serine to 42threonine. Both these changes are located in the PCR-amplified 5' end. A nearly full-length progastricsin clone (bp 1981188) was obtained from the kgtll library. This clone had only three conservative nucleotide substitutions (positions 486 (G/C), 842 (G/C), and 846 (A/C), the latter being identical with the changes in the XgtlO0 clone).

DISCUSSION Origin of Seminal Progastricsin

FIG. 2. Immunohistochemical localization of seminal progastricsin. All sections were incubated with antiprogastricsin 1:2000 overnight (detection with ABC peroxidase). A) Seminal vesicle; B)prostate gland; C) cancer of prostate gland. Note scattered positive cells inepithelia of seminal vesicles and prostate gland, and staining around neoplastic tissue.

Our results suggest the origin of seminal progastricsin to be the epithelia of both the seminal vesicles and the prostatic glands. This is consistent with our previous findings of progastricsin activity in extracts from both tissues. Our immunohistochemical staining results are inconsistent with those of Reid and coworkers [10, 11], who found progastricsin-positive cells in the prostate only and not in seminal vesicles. However, their findings of variable staining foci are similar to ours. Our results are consistent with those of Reese and colleagues [12], who reported progastricsin staining in both the seminal vesicles and in the prostate. There are several possible explanations of the discrepancy regarding staining in seminal vesicle samples. First, Reid and coworkers [10, 11] used 10% unbuffered formalin as fixative, which does not preserve antigen determinants as well as does Bouin's fixative, which we and Reese and coworkers used [12]. Second, Reid and coworkers [10,11] used an antibody raised against gastricsin, not against the zymogen. Most polyclonal antibodies react with both the enzyme and the zymogen, but occasionally some sera fail to show cross-reactivity. Third, we have observed great variability in staining intensity between different tissue blocks, some sections manifesting only weak staining in a few glands, whereas others manifest more intense and more widespread staining. By chance, Reid and coworkers may have investigated only tissue blocks that react poorly with the antibody.

231

SEMINAL PROGASTRICSIN

1304 C/A

846 A/C

PCR

-sO

263 T/C

301 T/A

I

I

198 42

AA

D

E-

I1311

842 846 G/C AIC

G/C I

p12 pl16 29

II

625

486

xgtll

i

607

t123

57

\I' '

I

S1188

314 S

129 S (D)

I

FIG. 4. Human seminal progastricsin cDNA clones and protein structure cDNA clones of human seminal progastricsin (kgt 10 and PCR-derived clone are from same individual; kgt 11clone isfrom different library). Nucleotides that differ from published genomic and gastric progastricsin sequence are marked. Nucleic acid numbering is from initiation codon. Coarsely hatched region represents leader sequence; finely hatched region represents propart. Arrow marks amino acid residues determined by protein sequencing. Potential phosphorylation sites are denoted by symbol ( (p12S, p16T, 129S, and 314S) and potential N-linked glycosylation site (57N) by symbol T.Ala29 and Thr42 (triangles) were deduced from nucleotide sequence; protein sequence manifested Val29 and Ser42. Active site aspartate residues are denoted by letter Din circle.

FIG. 3. In situ hybridization for seminal progastricsin transcript. Sections in A (seminal vesicles) and B (prostate) were hybridized with 200 ng/ml of 3' digoxigenin-ddUTP-labeled synthetic 38-mer oligo nucleotide complementary to progastricsin transcript. Note positive reaction in all epithelia of seminal vesicles and in some glands of prostate.

derived from a gene different from that of gastric progastricsin. The other substitutions we observed, including changes in amino acids, were present in the PCR-generated part of the sequence. These may have been due to the infidelity of the Taq polymerase, since the substitutions could not be confirmed from another cDNA clone from a different library. The change at position 263, which would have resulted in a Val-to-Ala substitution, could not be confirmed by protein sequencing. Gastric and Seminal ProgastricsinGene Products

Moreover, our finding of a progastricsin transcript by Northern blotting and in situ hybridization provides support for the interpretation that progastricsin is produced both in the seminal vesicles and in the prostate.

Structure of Seminal Progastricsin Like several others [16-19], we were unable to confirm the previously published protein sequence of gastric progastricsin [8]. The conservative nucleotide substitution at position 846, which did not result in amino acid substitution, appears to be a true difference from the previously reported sequences of gastric cDNA [16-20], since it appeared in clones from two different libraries. However, this change is probably the result of polymorphism and should not be taken as evidence that the prostate progastricsin is

The cDNA structure of prostate progastricsin cannot account for the charge differences between gastric and seminal progastricsin mentioned above. The observed charge difference in progastricsin disappeared upon activation, indicating that a charge difference should be present in the propart of the zymogen. Amino acid substitutions in this region are unlikely since the present results could not reveal any difference by N-terminal protein sequence covering all 43 residues in the propart and the 34 first amino acid residues of the enzyme moiety. The mass spectrometry spectra manifested a poor signal:noise ratio with rather broad peaks. The observed mass difference of 215 is not significant, but progastricsin from both sources appears to have a slightly higher mass than the theoretical mass calculated from the nucleotide sequence, indicating posttransitional modifications. However, the quality of the spectra did not allow any further characterization of the molecular differences. It seems likely that the charge difference of progastricsin

232

SZECSI ET AL.

is due to posttranslational modifications. Posttranslational modifications of aspartic proteases are well known. Phosphorylation has been reported from human [24], porcine [25], and bovine pepsin A [26]; glycosylation has been reported for simian pepsin A [27], porcine and bovine cathepsin D [28], proteinase A from baker's yeast [29], and human cathepsin E [30]. The five electrophoretic components of human pepsinogen A arise from posttranslational modifications of three gene products [31]. Human progastricsin contains 4 potential phosphorylation sites and a single Nlinked glycosylation site. Tang et al. have reported [32] phosphate-containing peptides in digest of human gastricsin even though the site of phosphorylation was not determined. N-linked glycosylation is unlikely to be the difference between seminal and gastric progastricsin since the ability of a monoclonal antibody to differentiate between the two zymogens was unaffected by endoglycosidase [15]. The observation of several genomic fragments of the human progastricsin gene reported by Hayano and coworkers [16] can be explained by the restriction fragment polymorphism described by Pals and colleagues [18]. Alternatively, these results may have been due to the fact that Hayano and coworkers used a rat cDNA probe that might have hybridized to other related aspartic protease genes. The minor sequence differences between the seminal and gastric samples are also compatible with the presence of two very similar genes. For example, the two genes of rat progastricsin differ only at a single base pair in the first exon [33], and the different human pepsinogen genes differ only by 7 base pairs out of 1164 translated base pairs [31]. The present results are thus best explained by polymorphism, even though the possibility of several genes cannot be completely ruled out. However, we consider this possibility to be remote. Both the fact that no genetic variation of human progastricsin has been observed at the protein level and the polymorphism described by Taggart and coworkers [19] suggest the existence of a single gene for human progastricsin. Likewise, our current results suggest gastric and seminal progastricsin to be products of the same gene. In all likelihood, the molecular difference observed is due to posttranslational modifications in the propart region of the zymogen, which seems to be tissue-dependent. It remains to be determined whether this difference affects the function of progastricsin of either derivation.

ACKNOWLEDGMENTS We thank Dr. A. Bjartell and Elise Nilsson, University of Lund, for their help with the in situ hybridization and immunohistochemistry, and Dr. P. Roepstorff, Odense University, for performing the mass spectrometry.

REFERENCES 1. Lundquist F, Seedorff H. Pepsinogen in human seminal fluid. Nature 1952; 170:11151116. 2. Seiffers MJ, Segal HL, Miller LL.Preliminary appraisal of the role of seminal pepsinogen I in human sterility. Fertil Steril 1965; 16:202-207. 3. Samloff I, Liebman W. Purification and immunochemical characterization of group II pepsinogens in seminal fluid. Clin Exp Immunol 1972; 11:405-414. 4. Hirsch-Marie H, Delafontaine D. Seminal pepsinogen: quantitation and correlation with the spermiogram. IntJ Androl 1978;1:397-404. 5. Moriyama A, Kageyama T, Takahashi K, Sasaki M. Purification of Japanese monkey prostate acid protease zymogen and its identification as a pepsinogen C-like zymogen. J Biochem 1985; 98:1255-1261. 6. Szecsi PB, Dalgaard D, Stakemann G, Wagner G, Foltmann B. The concentration of pepsinogen C in human semen and the physiological activation of zymogen in vagina. Biol Reprod 1989; 40:653-659. 7. Chulavatnatol M, Ruenwongsa P. Activation of proenzyme of acidic protease from human seminal plasma. Biochim Biophys Acta 1976; 452:525-532. 8. Foltmann B,Jensen A. Human progastricsin. Analysis of intermediates during activation into gastricsin and determination of amino acid sequence of the propart. EurJ Biochem 1982; 128:63-70. 9. Szecsi PB, Lilja H. Gastricsin-mediated proteolytic degradation of human seminal fluid proteins at pH levels found in the human vagina. J Androl 1993; 14:351-358. 10. Reid WA, Vongsorak L, Svasti J, Valler MJ, Kay J. Identification of the acid proteinase in human seminal fluid as a gastricsin originatingin the prostate. Cell Tissue Res 1984; 236:597. 11. Reid WA, Liddle CN, SvastiJ, KayJ. Gastricsin and malignant prostate. J Clin Pathol 1985; 38:639-643. 12. Reese JH, McNeal JE, Redwine EA, Samloff IM, Stamey TA. Differential distribution of pepsinogen II between the zones of the human prostate and the seminal vesicle. J Urol 1986; 136:1148-1152. 13. Foltmann B, Szecsi PB, Tarasova N. Detection of proteases by clotting of casein after gel electrophoresis. Anal Biochem 1985; 146:353-360. 14. Foltmann B, Tarasova NI, Szecsi PB. Methods for detection of proteinases: electrophoretic and immunological comparison of aspartic proteinases of different origins. In: Kostka V (ed.), Aspartic Proteinases and Their Inhibitors. Berlin: Walter de Gruyter & Co.; 1985: 491-507. 15. Szecsi PB, Kock C, Foltmann B. Seminal pepsinogen C is not identical with though very similar to gastric pepsinogen C. FEBS Lett 1988; 238:101-104. 16. Hayano T, Sogawa K, Ichihara Y, Fujii-Kuriyama Y, Takahashi K. Primary structure of human pepsinogen C gene. J Biol Chem 1988; 263:1382-1385. 17. Hayano T, Sogawa K, Ishihara T, Fujii-Kuriayma Y, Takahashi K. Primary structure of human pepsinogen C gene. J Biol Chem 1988; 263:14592. 18. Pals G, Azuma T, Mohandas TK, Bell GI, BaconJ, Samloff IM, Walz DA, Barr PJ, Taggart RT. Human pepsinogen C (progastricsin) polymorphism: evidence for a single locus located at 6p21.1-pter. Genomics 1989; 4:137-145. 19. Taggart RT, Cass LG,Mohandas TK, Derby P, Barr PJ, Pals G, Bell GI. Human pepsinogen C (progastricsin). Isolation of cDNA clones, localization to chromosome 6, and sequence homology with pepsinogen A.J Biol Chem 1989; 264:375-379. 20. Tang J, Wong R. Evolution in the structure and function of aspartic proteases. J Cell Biochem 1987; 33:53-63. 21. SambrookJ, Fritsch EF, Maniatis T. Molecular Cloning. Cold Spring Harbor, NY: Laboratory Press; 1989. 22. Meuwissen SGM, Mullink H, Bosma A, Pals G, Dfize J, Flipse M, Westerveld BD, Tas M, Brakk J, Kreuning J, Eriksson AW, Meyer CJLM.Immunocytochemical localization of pepsinogen I and II in the human stomach. In: Kreuning J, Samloff IM, Rotter JI, Eriksson AW (eds.), Pepsinogens in Man. Clinical and Genetic Advances. New York: Alan R. Liss; 1985: 185-197. 23. Varis KS. Peptic cells. In: Kreuning J, Samloff IM, Rotter JI, Eriksson AW (eds.), Pepsinogens in Man. Clinical and Genetic Advances. New York: Alan R. Liss; 1985: 177-184. 24. Jensen AL, Foltmann B. Determination of phosfoserine in human pepsinogens. Scand J Clin Invest 1995, (in press). 25. Ryle A. The porcine pepsins and pepsinogens. Methods Enzymol 1970; 19:316-336. 26. Meitner P, Kassell B. Bovine pepsinogens and pepsins: a series of zymogens and enzymes that differ in organic phosphate content. Biochem J 1971; 121:249-256. 27. Kageyama T, Takahashi K. The carbohydrate moiety of Japanese monkey pepsinogens. Its composition and site of attachment to protein. Biochem Biophys Res Commun 1977; 74:789-795. 28. Takahashi T, TangJ. Amino acid sequence of porcine spleen cathepsin D light chain. J Biol Chem 1983; 258:6435-6443. 29. Mechler B, Miller M, Mfiller H, Meussdoerffer F, Wolf DH. In vivo biosynthesis of the

SEMINAL PROGASTRICSIN vacuolar proteinases A and B in the yeast Saccharomyces cerevisiae. J Biol Chem 1982; 257:11203-11206. 30. Athauda SBP, Matsuzaki O, Kageyama T, Takahashi K. Structural evidence for two isozymic forms and the carbohydrate attachment site of human gastric cathepsin E.Biochim Biophys Res Commun 1990; 168:878-885. 31. Evers MPJ, Zelle B, Bebelman JP, van Beusechem V, Kraakman L, Hoffer MJV, Pronk JC, Mager WH, Planta RJ, Eriksson AW, Frants RR. Nucleotide sequence comparison of

233

five human pepsinogen A (PGA) genes; concerted evolution of the PGA multigene family. Genomics 1990; 4:232-239. 32. TangJ, MillsJ, Chiang L,de Chiang L.Comparative studies on the structure and specificity of human gastricsin, pepsin and zymogen. Ann NY Acad Sci 1967; 140:688-696. 33. Ishihara T, Ichihara Y, Hayano T, Katsura I, Sogawa K, Fujii-Kuriyama Y, Takahashi K. Primary structure and transcriptional regulation of rat pepsinogen C gene. J Biol Chem 1989; 264:10193-10199.