| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Departments of Biochemistry,
b Animal Sciences,
c and Veterinary Pathobiology, University of Missouri-Columbia, Columbia, Missouri 65211
d Department of Veterinary Science, University of Kentucky, Lexington, Kentucky 40506
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The cloning of bovine and ovine PAG-1 revealed that these proteins contain key mutations within their catalytic center that likely render them enzymatically inactive [4, 10]. However, despite these changes to the active site, the close resemblance of PAG-1 to pepsinogens and other aspartic proteinases has permitted the construction of atomic models for bovine and ovine PAG-1 [10]. It is apparent that the three-dimensional structure of these molecules is similar to that of other aspartic proteinases and that these molecules are still capable of binding peptide ligands. Indeed, the aspartic proteinase inhibitor, pepstatin, can be used as an affinity ligand to purify some PAGs (unpublished results). In short, bovine and ovine PAG-1 appear to have a physiological role other than in proteolysis, perhaps functioning through their ability to bind peptide or protein ligands.
Recently, related molecules have been identified in cattle [11, 12], sheep [13], and pigs [14]. In all three species, the PAGs appear to be products of a relatively large gene family. So far, 12 distinct bovine [12], 9 distinct ovine [12, 13], and 2 porcine [14] PAG cDNAs have been characterized. Likewise, several PAGs have been purified and shown to possess distinct amino terminal sequences [6, 13, 15]. As in the case of ovine and bovine PAG-1, some of the recently characterized PAG cDNAs encode proteins possessing mutations that would probably inactivate them as proteinases [10, 12–14]. Until recently, PAGs were believed to be confined to the Artiodactyla order (even-toed ungulates). In these species, they provide as much as 3–5% of the total mRNA within the placenta and may represent the principal transcripts in the mature placenta [11, 14]. So far, no corresponding molecules have been characterized outside the Artiodactyla, specifically within rodents and primates, despite decades of research characterizing placental-specific molecules in these species [16–18]. In this report, we describe the molecular cloning of a cDNA for an equine (Perrisodactyla order) placental aspartic proteinase that appears to be related to the PAG family. A similar gene is expressed in the placenta of the closely related zebra.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A Uni-Zap cDNA library (Stratagene, La Jolla, CA) was prepared from polyadenylated RNA from the extraembryonic membranes of six Day 25 equine conceptuses. DNA restriction enzymes, alkaline phosphatase-conjugated anti-rabbit IgG, avian myeloblastosis virus reverse transcriptase, 5-bromo-4-chloro-3-indolyl-1-phosphate (BCIP), and nitro blue tetrazolium (NBT) were purchased from Promega (Madison, WI). [
-32P]Deoxy-ATP (3000 Ci/mmol) and [-35S]deoxy-ATP (1000–1500 Ci/mmol) were purchased from NEN Research Products (Boston, MA). Taq-polymerase was from Gibco BRL (Grand Island, NY). Sequenase v.2 was from US Biochemical (Cleveland, OH). MagnaGraph nylon membranes were from Micron Separations Inc. (Westboro, MA), and SDS-PAGE molecular weight markers were from Bio-Rad (Richmond, CA). All other reagents were purchased from Sigma Chemical Co. (St. Louis, MO).
Animals
All animals used in these experiments were maintained and handled according to protocols approved by the Animal Care and Use Committees at the University of Missouri and the University of Kentucky.
Complementary DNA Cloning and DNA Sequencing
An equine Day 25 extraembryonic membrane cDNA library was screened with a probe mixture consisting of 32P-labeled PAG cDNA from cattle, sheep, and pig. Probes were generated by polymerase chain reaction (PCR) in the presence of [32P]deoxy-ATP with 1/10 the normal concentration of deoxy-ATP. Hybridization was carried out at 42°C in 50% formamide, 5-strength SSC (single-strength SSC is 0.15 M sodium chloride and 0.015 M sodium citrate), 0.5% SDS, 5-strength Denhardt's, and 0.1 mg/ml herring sperm DNA [19]. Twenty-four phage positives were in vivo excised and sequenced by the dideoxy-nucleotide chain-termination method [19]. One of the phage clones (ep12.1) was fully sequenced in both directions.
To clone equine PAG (ePAG) from zebra placenta, total cellular RNA (1 µg) isolated from term placenta was reverse-transcribed into cDNA for use as template in PCR [19]. PCR was performed with a general PAG 5' oligonucleotide (AGGAAAGAAGCATGAAGTGG), containing the start codon, and an antisense ePAG oligonucleotide encompassing the termination codon (GAGCCCGGGTCACATTTATACCGCAGTAGCCAG).
The conditions were 94°C (1 min), 42°C (1 min), 72°C (1 min) for 35 cycles. The resulting PCR product was cloned by using the TA cloning kit (Invitrogen, Carlsbad, CA) and sequenced.
RNA Isolation and Northern Analysis
RNA was collected from equine Day 32 (ovulation is Day 0) conceptus tissue and term placenta. For Day 32 tissue, the extraembryonic membranes were separated from the embryo, and the chorionic girdle was dissected away from the surrounding membranes prior to RNA isolation. The tissues were frozen on dry ice and stored at -80°C until RNA isolation. Term placenta was collected immediately after delivery of the placenta, frozen in liquid nitrogen, and stored at -80°C until RNA isolation. Other organ samples (skeletal muscle, liver, kidney, glandular stomach, esophageal stomach, heart, lung, spleen, and testes) were obtained from adult animals. The tissues were extracted by the acid guanidinium thiocyanate-phenol-chloroform method [20] with a LiCl extraction [21]. Total RNA was separated by electrophoresis on 1% formaldehyde-agarose and transferred onto nylon membranes [19]. Equine PAG mRNA was detected by hybridization with a 32P-labeled ep12.1 cDNA fragment (bases 255–556) as described for the cDNA library screening.
In Situ Hybridization
The ep12.1 clone was used as template to generate a 35S-labeled 302-base pair (bp) probe (bases 235–537 of the cDNA) by PCR, with a 3-fold higher concentration of the antisense oligonucleotide relative to the sense oligonucleotide as described previously [14]. Specific activity was approximately 3 x 107 cpm/100 ng of PCR product. A sense cDNA probe (used as control) was obtained by using only the sense primer. In situ hybridization was performed by standard methods [19] with some modifications [14]. Briefly, frozen sections (10 µm) of extraembryonic membranes were attached to gelatin-coated slides, fixed with 4% (w:v) paraformaldehyde in PBS, acetylated in 0.1 M triethanolamine/0.25% acetic anhydride, dehydrated in graded ethanol, air dried, and hybridized to 35S-labeled probe (3 x 106 cpm/slide) for 12–16 h at 42°C. Slides were coated with Kodak NTB-2 (Eastman Kodak, Rochester, NY) emulsion, dried at room temperature, exposed for 1–2 days at 4°C, developed, and counterstained with hematoxylin/eosin.
Expression of Recombinant ePAG
The coding region (without the signal peptide) of ep12.1 was amplified by PCR and engineered to contain in-frame, flanking ndeI restriction sites for cloning into the pET11a vector (Novagen, Madison, WI) for expression in bacterial strain BL21(De3)pLysS (sense: GTGTGCATATGTTAGTCACAATCCCTCTTGTG; antisense: GTGTGCATATGTGGGGTGGCCTCTGCATTTA) [22]. Recombinant ePAG expression was induced with 1 mM isopropyl-ß-D-thiogalactopyranoside. The protein was produced as insoluble inclusion bodies. The bacteria were lysed in a French pressure cell, and the inclusion bodies were isolated by centrifugation. The inclusions were washed two times in 20 mM Tris (pH 8.0), 1 mM EDTA, 20 µM PMSF, 1% Triton X-100 (v:v); this was followed by two washes in the same buffer containing 1 mg/ml deoxycholic acid. The washed inclusion bodies were solubilized in 7.5 M guanidine-hydrochloride (G-HCl), 20 mM Tris (pH 9.0), 100 mM 2-mercaptoethanol, 1 mM EDTA, 1 mM glycine. The solution was cleared by centrifugation, diluted to a final protein concentration of 50 µg/ml with 6 M G-HCl, 20 mM Tris (pH 9.0), 1 mM EDTA, 1 mM 2-mercaptoethanol, 1 mM glycine and allowed to air oxidize for 3 h via rapid stirring in a beaker at room temperature. The diluted protein was dialyzed against 12 volumes of 20 mM Tris (pH 7.0), 150 mM NaCl, 1 mM EDTA, 1 mM glycine, 20 µM PMSF, and 0.02% sodium azide (NaN3) for 12 h. The dialysis was repeated a second time for another 12 h. The refolded protein was concentrated by ultrafiltration with a 30 000 molecular weight cut-off filter (Amicon, Beverly, MA) to 2.5 ml and centrifuged twice (10 000 x g) to remove insoluble material. The protein solution was then passed over a Superose-12 gel filtration column (1 x 30 cm; Pharmacia, Piscataway, NJ), equilibrated in the same buffer, in order to separate ePAG monomers from larger aggregates (data not shown). Fractions containing a protein peak that corresponded to a relative molecular weight of 40 000 were pooled and dialyzed against 20 mM Tris (pH 7.0), 1 mM EDTA, 20 µM PMSF, 0.02% NaN3. The solution was concentrated to 1 ml and applied to a DEAE column (LKB Ultropac TSK DEAE-5PW; Pharmacia), equilibrated in the same buffer. The column was eluted with a salt gradient from 0 to 0.4 M NaCl in loading buffer (data not shown). Fractions were analyzed for proteolytic activity as described below. Fractions containing ePAG (40–60 mM NaCl fractions) were pooled, dialyzed against 10 mM Tris (pH 8.0), 1 mM EDTA, 20 µM PMSF, 0.02% NaN3, and concentrated by using a Centricon 30 (Amicon).
Proteinase Assays and Processing of the ePAG Zymogen
The hemoglobin (Hb) proteinase assay was performed as described previously [23]. Briefly, a 0.1-ml solution containing enzyme (1 µg) and [14C]methyl-Hb (0.04 µCi, in pH 7.5, 7.0, 6.5, 5.5, 4.5, 4.0, or 3.0 buffer) was incubated at 37°C for 2–4 h. Nonhydrolyzed Hb was then precipitated with 0.2 ml of 10% w:v trichloracetic acid (TCA) and centrifuged at 10 000 x g for 5 min, and 0.2 ml of the supernatant solution containing small 14C-labeled peptides that had been liberated by the enzyme were counted in a liquid scintillation counter. Assays were performed in the presence and absence of 1 µM pepstatin A. Controls included assays run in the absence of enzyme to determine background cpm levels and assays run with pepsin as a positive control.
To test the ability of ePAG to catalyze the removal of its own pro-peptide (or the pro-peptide of an adjacent ePAG), 400-ng aliquots of purified recombinant ePAG were incubated for 45 min in pH 7.0, 6.0, 5.5, 4.5, 3.5, or 2.5 buffer at 37°C (20 µl final volume). Incubations were carried out in the presence and absence of pepstatin A (1 µM). After the incubation, the samples were neutralized with 1 µl of 1 M Tris (pH 8.0) and 20 µl of double-strength SDS-loading dye. Proteins in the samples were separated by SDS-PAGE and visualized by silver staining [19].
To sequence the processed form of ePAG, 10 µg of refolded ePAG was incubated in 40 mM citrate buffer, pH 4.0 (100 µl final volume), for 1.5 h at 37°C. The protein was precipitated with 400 µl of 10% TCA, neutralized with 10 µl of 1 M Tris (pH 8.5), mixed with SDS-PAGE loading dye, and resolved by SDS-PAGE alongside 10 µg of unprocessed ePAG. The proteins were transferred to a polyvinylidene difluoride (PVDF) membrane (Immobilon-PSQ; Immobilon, Millipore, MA) and stained with Coomassie R250; the unprocessed and processed forms were then subjected to microsequencing.
Antiserum
To generate ePAG antiserum, inclusion bodies isolated from E. coli expressing recombinant ePAG were solubilized in 7.5 M G-HCl, 20 mM Tris (pH 9.0), 1 mM EDTA, and 100 mM 2-mercaptoethanol. Solubilized ePAG (0.25 mg) was diluted to 0.5 ml with 8 M urea/20 mM Tris (pH 8.0), mixed with 0.5 ml of Freund's complete adjuvant, and injected s.c. at multiple sites along the back of a New Zealand White rabbit. The rabbit was boosted with antigen at 4- to 5-wk intervals with 0.1 mg of the solubilized inclusion proteins in Freund's incomplete adjuvant. Blood was collected from the central ear vein 12–14 days after each booster injection and allowed to clot at 4°C overnight, and the serum was stored at -20°C.
Tissue Culture and Western Blot Analysis
Individual Day 25 conceptuses and term placenta (approximately 4 g of term placenta) were minced into small (1 mm3) pieces, washed three times in Dulbecco's Modified Eagle's medium containing penicillin (100 U/ml), streptomycin (100 µg/ml), and Fungizone (0.5 mg/ml), and cultured in 150-mm3 flasks for 24 h at 37°C under 5% CO2:95% air. The culture media (~10 ml) were collected, centrifuged to remove cellular debris, and frozen at -20°C until use. Proteins released into the medium (10 µg) were analyzed by electrophoresis on 10% SDS-polyacrylamide gels and stained with Coomassie brilliant blue [19].
For Western blots, secretory proteins (2 µg) from Day 25 conceptus cultures or term placental cultures were separated by SDS-PAGE and electrophoretically transferred onto nitrocellulose membranes. The membranes were blocked in 5% nonfat dry milk in PBS-Tween (0.05%), reacted with ePAG antiserum (1:1000), washed, and incubated with alkaline phosphatase-conjugated anti-rabbit IgG (Promega) diluted 1:2500. The blots were then washed and stained in a mixture of NBT and BCIP according to the manufacturer's instructions. Nonspecific binding was determined by incubation of membranes with normal rabbit serum (1:250) instead of antiserum.
To analyze secretory proteins from chorionic girdle cells (a gift from O.J. Ginther, University of Wisconsin, Madison), the chorionic girdle from a Day 34 conceptus was diced and placed in the upper chamber of a Matrigel invasion chamber (Collaborative Biochemical, Bedford, MA). The cells were cultured in RPMI containing 10% fetal bovine serum (FBS), 1% penicillin-streptomycin, and Hepes. Invasive girdle cells were able to migrate through the Matrigel and enter the lower chamber; this allowed collection of their secretory proteins [24]. Media were harvested from both the upper and lower chambers after 48, 72, and 96 h of culture. Samples (20 µl) of culture media were resolved on SDS-PAGE for Western blotting as described above.
Cloning the ePAG Promoter
To clone the proximal region of the ePAG promoter, PCR was performed upon equine genomic DNA with three sense oligonucleotides that annealed to conserved regions within the promoters of three previously cloned PAG genes (boPAG-1, boPAG-2, and porcine PAG-2; unpublished results) along with an antisense oligonucleotide that annealed to exon 1 of ePAG (ep12.1e1r: AAGCACTCTGAGAGGGTC). One of the general PAG promoter oligonucleotides (PAGprom2: TGyCwAGCATkGCCCCn where y: c/t, w: a/t, k: g/t, and n: any nucleotide) generated a product of the expected size (435 bp) with the antisense oligonucleotide. PCR conditions were 96°C (10 sec), 55°C (15 sec), and 72°C (30 sec) for 36 cycles. The resulting product was blunt end-ligated into EcoR5-cut pBluescript vector (Stratagene, La Jolla, CA) and sequenced in both directions by the dideoxy-chain termination method [19]. Another PCR reaction was performed to confirm that the original PCR fragment represented the ePAG promoter. In this, an oligonucleotide specific for the putative ePAG promoter was generated (ePAGprom1: GCATGGCCCCTGGCACTT). The ePAGprom1 oligonucleotide was used with an antisense oligonucleotide from exon 2 of the ePAG gene (ep12.1e2r: CAGGTAGTTCCTCATGGG) with equine genomic DNA as template. PCR conditions were 96°C (10 sec), 55°C (15 sec), and 72°C (60 sec) for 36 cycles. The resulting PCR product (~1.6 kilobase pair) was partially sequenced.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
An equine Day 25 placental cDNA library was screened with a probe mixture consisting of 32P-labeled PAG cDNA from cattle, sheep, and pig. Of the ten thousand plaques that were plated, 0.5% hybridized to the probe. Twenty-four of these were randomly selected, plaque purified, and partially sequenced. All of the clones encoded the same protein, referred to here as "equine PAG" (ePAG) to indicate its relationship with PAGs from ruminants and swine. One of the clones (ep12.1) was fully sequenced in both directions. The cDNA for ep12.1 was 1371 bp and contained an open reading frame of 388 codons encoding a polypeptide with a theoretical molecular weight of 42 897 (Fig. 1). From the inferred amino acid sequence, ePAG is a secretory protein as determined by its predicted 15-amino acid signal sequence [25]. A single potential site for N-glycosylation (NNS) is present close to the carboxyl terminus.
|
A very similar transcript with 99% nucleotide sequence identity to the ePAG cDNA was cloned from the placenta of the closely related species, the zebra. Again, 10 cDNAs were isolated and partially sequenced; all of the sequenced regions were identical. As in the horse, there was no evidence for the extensive range of transcripts found in bovine and ovine placenta [12, 13]. The polypeptide encoded by the zebra PAG mRNA differed at only five positions out of 388 from the inferred sequence of ePAG (Fig. 1), including a Leu for Ser substitution at codon 343, which negated the potential site for N-glycosylation at N341.
Equine and zebra PAG have greatest identity with a cat PAG-like protein recently cloned in our laboratory (71% amino acid identity) and with rabbit pepsinogen F (69% amino acid identity), a gastric proteinase expressed for a short time in the neonatal rabbit [26] (Table 1). The amino acid sequence identity between ePAG and other members of the pepsinogen family is around 55%, whereas its identity with members of the bovine, ovine, and porcine PAG families ranges from 52% to 57% (Table 1). Additionally, ePAG possesses both of the aspartic acids implicated in the catalytic activity of aspartic proteinases (Fig. 1). The sequences around these two residues, IFDTGS in the amino terminal lobe and IVDTGT in the carboxyl terminal lobe, conform to the consensus (hydrophobic-hydrophobic-DTGT/S) found in catalytically active members of this class of proteinase [1, 2].
|
Expression of ePAG mRNA in Conceptus Tissue
Expression of the ePAG transcript within tissues of the equine conceptus (extraembryonic membranes plus embryo) was examined. Northern blot analysis of two Day 32 equine conceptuses revealed that the 1.6-kilobase (kb) ePAG mRNA was highly expressed in placental tissue, including the cells that form the chorionic girdle, but was completely absent in embryos (Fig. 2A). The ePAG transcript was also detectable in placenta collected at term (Fig. 2B). Northern blotting was performed upon RNA isolated from various adult tissues as well. No ePAG message was detectable in RNA from equine skeletal muscle, adrenal gland, liver, kidney, glandular stomach, esophageal stomach, heart, lung, spleen, or testis (data not shown). From these data, it would appear that the ePAG transcript is expressed predominantly, if not exclusively, in the placenta.
|
In situ hybridization with 35S-labeled probes revealed that the expression of ePAG was restricted to certain cell populations within Day 25 extraembryonic membranes and was completely absent in others (Fig. 3). The morphology of the labeled cells, although poorly conserved in these frozen sections, indicated that the ePAG mRNA was present predominantly in the allantochorion and the avascular chorion. No specific hybridization signals were noted with the sense probe. No message was detected either in Day 15 uterine endometrium or in Day 25 embryos (data not shown).
|
Processing of ePAG and Proteolytic ActivityToward Hemoglobin
Purified, recombinant ePAG was tested in a standard proteinase assay with [14C]Hb as substrate [23]. Some activity was demonstrated toward [14C]Hb, although relatively long incubation times were required before appreciable 14C-labeled peptide production was evident (2–4 h). The pH optimum was approximately 4.5, but low activity was still detectable at pH 7.5 provided that the pro-peptide had been removed prior to the assay (Fig. 4A). Addition of 1 µM pepstatin A, 5 min prior to the start of the assay, eliminated all activity. The assay performed with an equal amount of pepsin at pH 4.5 generated 10-fold higher soluble radioactivity compared to ePAG (data not shown).
|
To further examine the activity of ePAG, the purified recombinant protein was subjected to different pH conditions (Fig. 4B). The presence of a faster-migrating ePAG protein after SDS-PAGE was assumed to represent the removal of the pro-peptide by a self-digestion mechanism similar to that observed with pepsinogen A [27]. The appearance of the faster-migrating, and presumed processed form, of ePAG became evident at pH 5.5. At pH 4.5, it was the predominant form. However, the mobility of the processed form at pH 4.5 was slightly different from that at pH 5.5, perhaps due to different cleavage sites within the pro-peptide at the different pH conditions. At pH 3.5 and 2.5, less of the mature form was produced, possibly because catalytic activity was reduced as a result of the low pH (Fig. 4B). Again, the presence of pepstatin A in the reaction mixture completely inhibited the appearance of the mature form of ePAG (data not shown). The doublet appearance of the purified recombinant protein was observed whenever ePAG was refolded from inclusion bodies and could represent minor proteolytic processing of the recombinant protein during the renaturation procedure.
To identify the site at which cleavage occurred, the processed form of ePAG was transferred to a PVDF membrane and subjected to microsequencing. Two signals were observed. The stronger began at Thr 34, while a second one indicated an additional cleavage site between Leu 22 and Lys 23 (Fig. 1).
Secretion of ePAG by Cultured Conceptus Tissue
Western blotting of conditioned media (2 µg of total protein) from Day 25 conceptuses (Fig. 5A) revealed two immunopositive bands of approximate Mr 41 000 and 37 000 (Fig. 5B). It is assumed that the two represented the zymogen as well as the mature, and presumably active, form from which the pro-peptide had been removed. The mobility of the more slowly migrating band was similar to that of ePAG expressed in bacteria, indicating that ePAG produced by trophoblast tissue is not extensively glycosylated. Equine PAG was also detectable by Western blotting of culture media from Day 15 conceptuses (data not shown) but did not appear to be released from cultured explants from term placenta (Fig. 5B), despite the presence of the ePAG mRNA in this tissue (Fig. 2B).
|
Equine PAG may be expressed by a specialized component of the early placenta, the chorionic girdle, which is a strip of invasive trophoblast encircling the spherical conceptus [28]. These cells are capable of migrating through the uterine epithelium and into the uterine stroma to form the endometrial cups, highly differentiated cells that serve an endocrine role during the first one third of pregnancy [29]. Analysis of medium from the culture of Day 34 chorionic girdle cells that had been placed in the upper well of a two-chambered culture vessel revealed that immunoreactive ePAG accumulated in the lower chamber, indicating that the protein may have been released by invasive cells able to migrate through the Matrigel barrier (Fig. 6). The immunoreactive proteins noted below the recombinant ePAG in the heavily loaded lanes (Figs. 5B and 6) are bacterial proteins present within the solubilized inclusion bodies.
|
Cloning of the ePAG Promoter
The structures for three PAG genes have so far been elucidated ([30]; unpublished results). Like most of the mammalian aspartic proteinases, the PAG genes exhibit a conserved nine-exon/eight-intron structure [30]. Despite the similarities in their gene structures to other aspartic proteinases, the promoter regions of the PAGs are unique and show no obvious sequence identity with the promoters of other aspartic proteinase genes ([30]; unpublished results). Since the transcribed region of the ePAG gene shares approximately equal sequence identity with the PAGs and pepsinogens, the ePAG promoter was cloned in an attempt to determine whether it could provide a link between ePAG and other members of the aspartic proteinase family. To accomplish the cloning, oligonucleotides directed toward conserved regions within the promoters of known PAG genes were used in conjunction with an antisense, exon 1 ePAG-specific oligonucleotide in PCR with equine genomic DNA as template. The resulting PCR product (435 bp) was cloned and sequenced. The sequence provided a complete match with exon 1 of the ePAG cDNA, and the upstream region aligned with the promoter of the boPAG-1 gene (Fig. 7). There was no similarity to the upstream promoter regions of either the pepsinogen genes or any other aspartic proteinase gene that has been cloned. As further confirmation that this clone represented the ePAG promoter and had not been produced artifactually, another PCR reaction was performed with an oligonucleotide specific for the presumed ePAG promoter and an antisense oligonucleotide within exon 2 of the ePAG gene. Sequence from this PCR product perfectly matched the sequence shown in Figure 7 and provided a 100% match with exon 2 of the ePAG cDNA.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Equine PAG represents the first such placentally expressed molecule to be identified outside the Artiodactyla order. It may well be misnamed, as there is no evidence that ePAG is a glycoprotein. Its electrophoretic mobility after SDS-PAGE analysis, for example, is identical to that of the nonglycosylated recombinant form produced in E. coli. The closely related PAG from zebra lacks the single potential site for N-glycosylation present on the horse polypeptide. As a consequence, the molecular weight of ePAG is much smaller than those of most Artiodactyla PAGs, which usually display electrophoretic mobilities consistent with molecular masses greater than 55 kDa. Equine PAG is also active as a proteinase, whereas some bovine and ovine PAGs appear to have accumulated amino acid changes that would likely render them inactive [4, 10]. On the other hand, as with other PAGs, the expression of ePAG seems to be confined to the trophoblast. Furthermore, the promoter region of its gene is PAG-like rather than pepsinogen-like, a feature discussed in greater detail below.
Equine PAG had earlier been anticipated to be a proteinase based on amino acid sequence analysis and computer modeling [10]. This prediction was confirmed in the proteinase assays performed here (Fig. 4, A and B), in which purified recombinant ePAG had the ability to cleave [14C]Hb. In contrast to the case in which an equal amount of pepsin was used, long incubation times (2–4 h) were necessary to obtain sufficient soluble radioactive product for analysis. There seem to be three possible explanations. First, the recombinant ePAG may not have refolded efficiently. Second, hemoglobin may not be an appropriate substrate for ePAG, because it lacks the optimal amino acid sequences required for proteolytic attack. If, like renin, ePAG has strict substrate requirements, then it is unlikely that high activity would be evident in such a general proteinase assay. Third, the primary function of ePAG may not be that of a proteinase. The residual activity observed could be a remnant of an evolutionarily ancient role, now of secondary importance, as has been observed with certain ribonucleases [31].
Due to their expression by binucleate trophoblast cells, bovine and ovine PAG-1 can be detected in the maternal circulation [9]. Binucleate cells compose up to 20% of the trophoblast within the mature ruminant placenta [7] and contain the PAGs as components of their secretory granules [4]. Subsequent to binucleate cell fusion with uterine epithelial cells, the PAGs are released into the maternal tissues. During implantation, the equine placenta possesses cells functionally homologous to ruminant binucleate cells. These compose the chorionic girdle, a strip of specialized cells encircling the spherical conceptus. The girdle cells are capable of migrating through the uterine epithelium to enter the uterine stroma where they enlarge and aggregate to form the "endometrial cups''—an endocrine tissue, supported by maternal blood capillaries, that is responsible for the production of a peptide hormone, eCG [28, 32, 33]. Since ePAG may be expressed by girdle cells (Fig. 6), it could be present along with eCG in the maternal circulation as the endometrial cups become established.
Although the PAGs constitute a large gene family within the Artiodactyla, only a single transcript for an ePAG has so far been identified in the horse and its close relative, the zebra (Perrisodactyla order). Southern blotting of equine genomic DNA with a short ePAG genomic probe (~350 bp encompassing exon 7, intron 7, exon 8) produced only two hybridizing bands (data not shown). Therefore, ePAG does not seem to belong to a large group of closely related genes [12]. Importantly, genes very similar to the one for ePAG are present within species only distantly related to the horse (order Perissodactyla). For example, a PAG-like molecule cloned as a cDNA from the placenta of a cat (order Carnivora) at Day 30 of pregnancy shares high sequence identity with ePAG (71% inferred amino acid identity). Similarly, the rabbit (order Lagomorpha) expresses an aspartic proteinase, pepsinogen F (pepF), which shares 69% amino acid identity with ePAG (Table 1). Rabbit pepF was originally identified as a gastric proteinase expressed for a short time (~1 wk) in the neonate [26]. Contrary to its name, rabbit pepF actually has greater sequence identity with members of the PAG family than with classical stomach pepsinogens. It is not known whether pepF is present in the placenta of the rabbit. Importantly, expressed sequence tags (accession numbers: W30491, W16236) from Day 19.5 mouse fetuses (order Rodentia) share greater sequence identity to ePAG and rabbit pepF than to other aspartic proteinases and may represent the murine homologue of ePAG and rabbit pepF. In short, it appears that ePAG/PepF-like genes have been conserved within four distinct orders of mammals widely separated in evolutionary distance. It remains to be determined whether they are also present in primates and other mammalian orders.
The relationship, both functional and genetic, between ePAG and the PAG family is not entirely clear. Most of the mammalian aspartic proteinases share a common genomic structure consisting of 9 exons of conserved length and 8 introns of varying lengths. Each gene is approximately 12 000 bp in length [34–36], but the promoter regions of the different subfamilies (e.g., pepsin A, gastricsin, renin, etc.) are distinct and share identity only with their homologues in other species. Likewise, the promoters of the PAG genes are distinct, having no obvious sequence identity with the promoters of other aspartic proteinase genes ([30]; unpublished results). These unique promoter regions are the likely explanation for the distinct tissue-specific distributions of the various aspartic proteinases. It seems possible that shared identity, or lack thereof, between promoter regions may be a useful indicator of genetic relationships among members of the mammalian aspartic proteinase family. With this in mind, the promoter of the ePAG gene was cloned and sequenced. There was 69% nucleotide sequence identity with the bovine PAG-1 promoter over the 360 bp examined but no similarity to the promoters of other aspartic proteinase genes (Fig. 7). This result strengthens the hypothesis that ePAG and possibly its relatives in the rabbit, cat, and mouse are most appropriately grouped with the PAG gene family rather than with the pepsinogens. Possibly, the ePAG/pepF group is a remnant of the primordial pepsinogen gene that ultimately underwent many rounds of duplication within the Artiodactyla to create the large PAG gene family found in modern day cattle, sheep, pigs, and their relatives.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: R. Michael Roberts, Department of Animal Sciences, University of Missouri, 158 Animal Science Research Center, Columbia, MO 65211–5300. FAX: 573 882 6827; e-mail: robertsrm@missouri.edu
3 Current address: Department of Animal Physiology, University of Agriculture and Technology, 10–718 Olsztyn, Poland.
Accepted: December 7, 1998.
Received: August 20, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  X. Chen, C. S. Rosenfeld, R. M. Roberts, and J. A. Green An Aspartic Proteinase Expressed in the Yolk Sac and Neonatal Stomach of the Mouse Biol Reprod, October 1, 2001; 65(4): 1092 - 1101. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Green, S. Xie, X. Quan, B. Bao, X. Gan, N. Mathialagan, J.-F. Beckers, and R. M. Roberts Pregnancy-Associated Bovine and Ovine Glycoproteins Exhibit Spatially and Temporally Distinct Expression Patterns During Pregnancy Biol Reprod, June 1, 2000; 62(6): 1624 - 1631. [Abstract] [Full Text]
|
|
|
|
 |
|
 A. L. Hughes, J. A. Green, J. M. Garbayo, and R. M. Roberts Adaptive diversification within a large family of recently duplicated, placentally expressed genes PNAS, March 28, 2000; 97(7): 3319 - 3323. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
