MOLECULAR REPRODUCTION AND DEVELOPMENT 70:262–270 (2005)

Identification and Cloning of Caprine Uterine

1,2 1 3 1 S¸ABAN TEKIN, MARIA B. PADUA, GARY R. NEWTON, AND PETER J. HANSEN * 1Department of Animal Sciences, University of Florida, Gainesville, Florida 2Department of Biology, University of Gaziosmanpas¸a, Tokat, Turkey 3Cooperative Agricultural Research Center, Prairie View A&M University, Prairie View, Texas

ABSTRACT The uterine have been des- istics suggesting it is not an inhibitory serpin. Mol. cribed in , , and as a highly diverged Reprod. Dev. 70: 262–270, 2005. group of the large superfamily of serpin proteins that ß 2005 Wiley-Liss, Inc. typically function as proteinase inhibitors. Here, the range of species that possess and express a uterine Key Words: caprine; ovine; serpin; ; endome- serpin is extended to the . Sequencing of trium; cDNA amplified from total RNA from a pregnant goat at day 25 of pregnancy resulted in a 1,292 bp full- length consensus cDNA sequence for caprine uterine INTRODUCTION serpin (CaUS). The predicted amino acid sequence of The uterine serpins found in sheep (Ing and Roberts, the caprine precursor showed 96%, 82%, 55%, and 1989), cattle (Mathialagan and Hansen, 1996), and pigs 56% identity to OvUS, BoUS, PoUS1, and PoUS2, (Malathy et al., 1990) are part of a large superfamily of respectively. The signal peptide extends from amino proteins of about 500 known members that, prototypi- acids 1 to 25, resulting in a secreted protein of 404 cally, fold into a conserved structure and inhibit serine amino acids and 46,227 Mr (excluding carbohydrate). proteinases through a unique suicide-like mechanism Both the goat and sheep uterine serpins have a nine (Irving et al., 2000; Silverman et al., 2001). Like several amino acid insert in the Helix I region that is not found other members of the serpin family, the uterine serpins in bovine or porcine uterine serpins. A total of 13 amino have functions distinct from proteinase inhibition. The acids in CaUS are different than those for the nearest best studied of these, ovine uterine serpin can inhibit homologue, ovine uterine serpin. One of these is in proliferation of ab T lymphocytes (Segerson et al., 1984; the site of cleavage of the signal sequence, where a Skopets and Hansen, 1993; Skopets et al., 1995; Peltier single nucleotide substitution (G ! C) changed the et al., 2000a) and has been proposed to protect the cysteine for the sheep, bovine, and porcine to a conceptus from destruction by the maternal immune serine. In addition, the amino acid at the putative P1– system (Hansen, 1998). The porcine uterine serpins P10 site (the scissile bond for antiproteinase activity) is form complexes with the iron-containing protein, uter- a valine for CaUS, BoUS, PoUS1, and PoUS2 versus oferrin (Baumbach et al., 1986), and may be involved in an alanine for OvUS. The hinge region of all five of metabolism of that protein. the uterine serpins (P17–P9) is distinct from the In the present report, we provide data that extend the consensus pattern for inhibitory sequences and it is range of species in which uterine serpins exist to the goat unlikely, therefore, that the uterine serpins possess (Caprus hircus). Moreover, the coding regions for ovine prototypical proteinase inhibitory activity. The goat and caprine uterine serpin (CaUS) exhibit very similar uterine serpin was immunolocalized to the glandular sequence identity and the caprine protein shows the epithelium of the from a pregnant nanny same tissue specificity in the uterus as other uterine at day 25 of pregnancy. There was also immunoreac- serpins. tive product in scattered luminal epithelial cells. No immunoreaction product was detected in endometrium MATERIALS AND METHODS from a nanny at day 5 of the estrous cycle. Western RNA Isolation blotting of uterine fluid collected from the pregnant Intercaruncular endometrium was collected from the uterine horn of a unilaterally-pregnant goat revealed uterus of a nanny at day 25 of pregnancy and snap frozen the presence of a protein band at Mr 56,000 that reacted with monoclonal antibody to OvUS. In conclu- sion, the range of species in which uterine serpins are Grant sponsor: USDA NRICGP; Grant number: 2001-35204-10797. present and expressed in the uterus includes the goat in *Correspondence to: Dr. Peter J. Hansen, Department of Animal addition to the previously described sheep, cow, and Sciences, University of Florida, PO Box 110910, Gainesville, FL 32611-0910. E-mail: [email protected]fl.edu . In all of these species, the uterine serpin is derived Received 6 July 2004; Accepted 27 August 2004 primarily from glandular epithelium, is secreted into Published online in Wiley InterScience (www.interscience.wiley.com). the uterine lumen, and contains sequence character- DOI 10.1002/mrd.20206 ß 2005 WILEY-LISS, INC. CAPRINE UTERINE SERPIN 263 in liquid nitrogen. Total RNA was isolated from endo- The concentration of the concentrated DNA was deter- metrial tissue samples with the TriPure reagent (Roche mined spectrophotometrically and purity was asses- Biochemicals, Indianapolis, IN). The quantity of RNA sed by electrophoresis as described in the previous was assessed spectrophotometrically and integrity of section. RNA was examined by electrophoresis. Sheep inter- Sequencing of the DNA samples was performed at the caruncular endometrium was collected at 140 days of University of Florida DNA Sequencing Core Laboratory gestation from the pregnant uterine horn of a Rambouil- using ABI Prism BigDye Terminator cycle sequencing let crossbred ewe made unilaterally-pregnant as des- protocols developed by Applied Biosystems (Perkin- cribed earlier (Bazer et al., 1979). The uterus was Elmer Corp., Foster City, CA). The fluorescently-labeled removed after captive bolt stunning and exsanguina- extension products were analyzed on an Applied Bio- tion, and tissue was snap frozen in liquid nitrogen. Total systems Model 373 Stretch DNA Sequencer or 377 DNA RNA was isolated from endometrial tissue samples Sequencer or on a 3100 Genetic Analyzer (Perkin-Elmer). using TRI reagent (Sigma-Aldrich, St. Louis, MO) ac- Double strand sequencing of CaUS cDNA amplified cording to the manufacturer’s protocol. RNA concentra- using primer set 1 was made by using primer set 1 and tion was determined spectrophotometrically. primer set 3 (forward primer, TTT TCA GCC CAA TCT CAC C; reverse primer, GGC ATC TTA ACC ATC GTA) Cloning of Caprine Uterine Serpin synthesized by Integrated DNA Technologies Inc. CaUS cDNA was amplified from total endometrial (Coralville, IA). Partial sequencing (single strand) of 50 RNA by reverse transcription-polymerase chain reac- and 30 ends of CaUS cDNA amplified using primer set 2 tion (RT-PCR) using primer sets 1 (forward primer, CAC was performed using primer set 2. Nucleotide sequences GTC CCA CAG GAG AAT G; reverse primer, CTC were aligned and assembled using programs in the AAC TTG GGG GTT GAG GAC T) and 2 (forward Sequencher 3.0 software package (Gene Codes Corp., primer, TCA GTA GAT AAC AGC GGG CTC C; reverse Ann Arbor, MI). primer, GAA TTG TAC TCT TTT TAT TCA TGG), The consensus sequence of CaUS cDNA (submitted which, based on the sequence for OvUS (Ing and to GenBank Third Party Annotation database and Roberts, 1989; GenBank accession number M21027), assigned accession number TPA: BK005554) was were estimated to produce products of 1,290 and aligned with other uterine serpin sequences (OvUS, 1,400 bp, respectively. The RT-PCR was performed GenBank accession number M21027; BoUS, GenBank using SuperScript One Step RT-PCR kit with Platinium accession number L22095; PoUS1, GenBank accession Taq (Invitrogen, Carlsbad, CA). To verify that PCR pro- number M30315; PoUS2, GenBank accession number ducts were amplified from RNA only, the SuperScript NM213845) using multiple sequence alignment analy- reverse transcriptase was omitted from control reac- sis with the ClustalW program (Higgins et al., 1994) at tions. cDNA synthesis and pre-denaturation reactions the European Bioinformatics Institute (http://www.ebi. were 1 cycle of 508C for 30 min and 948C for 2 min, ac.uk/clustalw/) and with the Genedoc program at http:// respectively. PCR amplification consisted of 35 cycles of www.psc.edu/biomed/genedoc/. The inferred amino acid 948C for 30 sec, 468C for 30 sec, and 728C for 2 min and sequence of CaUS was determined using the TRANS- 1 cycle of 728C for 5 min for final extension. A 2 reaction LATE program of EXPASY (http://us.expasy.org/) and buffer (containing 0.4 mM of each dNTP and 2.4 mM aligned with inferred amino acid sequences of other MgSO4) supplied in the kit was used in all PCR runs. uterine serpins using ClustalW and Genedoc. The puta- PCR products were separated by electrophoresis using a tive signal sequence was determined by use of SignalP 0.8% (w/v) agarose (Fisher Scientific, Fair Lawn, NJ) gel 3.0 from Center for Biological Sequence Analysis, Tech- in Tris-acetate buffer (40 mM Tris-acetate, 2 mM EDTA) nical University of Denmark (Bendtsen et al., 2004), pH 8.5 containing 0.5 mg/ml ethidium bromide (Sigma- available at http://www.cbs.dtu.dk/services/SignalP/. Aldrich), visualized on an ultraviolet UV translumina- Identification of start codons was achieved using the tor and photographed using either a Polaroid camera TRANSLATE program at EXPASY. The N-glycosyla- and film (Fisher Scientific, Pittsburgh, PA) or a Sony tion sites were found using the NetNGlyc1.0 from Mavica CD400 digital camera (Sony, Japan). Center for Biological Sequence Analysis, Technical Uni- versity of Denmark at http://www.cbs.dtu.dk/services/ Sequencing of Caprine Uterine Serpin NetNGlyc/as. Identification of the putative reactive site Amplicons were gel-purified using the S.N.A.P. gel scissile bond of CaUS (i.e., the P1–P10 site) was based on purification kit from Invitrogen before sequencing. the location of the putative scissile bond in ovine uterine Briefly, PCR products were run on 0.8% (v/v) agarose serpin (Ing and Roberts, 1989). gels containing 0.5 mg/ml ethidium bromide, bands corresponding to CaUS or ovine uterine serpin were Immunohistochemistry cut out, and excised bands were loaded into spin filters Endometrium for immunohistochemistry was collect- provided in the kit and centrifuged at 14,000g at 48C for ed from nannies of mixed breeds on either day 5 of 10 min. The filtrate was washed with DEPC water the estrous cycle (day of estrus ¼ day 0, n ¼ 1) or day 25 (Biotecx, Houston, TX) and concentrated by centrifuga- of pregnancy (n ¼ 1) after captive bolt stunning and tion twice at 5,000g for 20 min using the Centricon YM30 exsanguination. Cross-sections of the uterine horns (Millipore Corporation, Bedford, MA) filter apparatus. were fixed immediately in 4% (w/v) paraformaldehyde 264 S. TEKIN ET AL. for 24–48 hr, processed and embedded in paraffin for Samples of goat uterine fluid (5 mg protein), sheep immunohistochemistry. uterine fluid (4.5–9 mg protein), and, as a negative For immunohistochemistry, 5 mm sections were cut control, (5 mg) were separated by SDS–PAGE using a manual microtome (American Optical Company, on 4%–15% Tris-HCl ready gels (Biorad, Hercules, CA). Buffalo, NY), fixed on poly-lysine coated slides, and Proteins were transferred electrophoretically at 80 V deparaffinized by washing in Safeclear tissue clearing (200 mA) for 1 hr at room temperature in a de- agent (Fisher Diagnostics, Middletown, VA) and ethanol gassed buffer of 25 mM Tris, 193 mM glycine, and (100%, 95%, and 80%, v/v) three times, two times, and 20% (v/v) methanol onto 0.2 mm nitrocellulose mem- once each for 5 min, respectively. Sections were micro- branes (Hybond ECL, Amersham Biosciences, Piscat- waved while immersed in 10 mM citrate, pH 6.0. The away, NJ). Open sites were blocked with blocking buffer procedure was performed three times for 2 min each and [1% (w/v) gelatin and 0.3% (v/v) Tween-20 in 10 mM specimens were allowed to cool between procedures. Tris þ 150 mM NaCl, pH 7.6] at room temperature Slides were then washed twice in deionized water (5 min overnight. Blots were probed with anti-OvUS each), once in phosphate-buffered saline [PBS; 0.1M (HL 218, Leslie et al., 1990) (hybridoma supernatant sodium phosphate, pH 7.4 containing 0.9% (w/v) sodium diluted 1:25,000 in blocking buffer) for 1 hr. After chloride] containing 2% (v/v) hydrogen peroxide for washing, blots were exposed to a 1:8,000 dilution of goat 5 min, and once in PBS for 5 min. All further steps were anti-mouse IgG conjugated to horseradish peroxidase in performed in a humidity chamber at room temperature blocking buffer provided in the ECL Western blotting and slides were washed for 3 min between each step with detection kit (Amersham Biosciences) for 1 hr. Presence PBS-GS [PBS containing 2% (v/v) goat serum]. Sections of second antibody was detected by reacting the blot with were then processed for immunohistochemistry using a ECL Western blotting detection reagent as per manu- universal monoclonal detection kit (Biomeda, Foster facturer’s instructions and exposure of X-ray film (Fuji City, CA) according to the manufacturer’s protocol and Medical X-ray Film, Fuji Foto Film Co, Tokyo, Japan). with all steps at room temperature. Briefly, sections were sequentially incubated with tissue conditioner for RESULTS 5 min, PBS-GS for 3 min, and primary antibody for 1 hr. Sequence of Caprine Uterine Serpin cDNA Primary antibody was a monoclonal antibody to ovine Sequencing of cDNA amplified from total mRNA from uterine serpin (HL-708; Leslie et al., 1990) used as a pregnant goat at day 25 of pregnancy using primer set hybridoma-conditioned medium at a 1:1,000 to 1:1,600 1 resulted in a 1,265 bp product (Fig. 1). This amplicon dilution in washing buffer for 1 hr. A similar dilution was missing the beginning of the signal sequence and of hybridoma cell culture medium (BD Biosciences, stop codon. An additional PCR product was produced Palo Alto, CA) was used as negative control. Slides were using primer set 2. Sequencing of 50 and 30 ends of this then sequentially incubated with secondary antibody cDNA resulted in information on 545 and 756 bp (biotinylated goat anti-mouse immunoglobulin) for 1 hr, streptavidin-alkaline phosphatase reagent for 30 min, and 3-amino, 9-ethylcarbazole for 10 min (all reagents from the kit). Slides were washed under tap water and counterstained with diluted hematoxcylin (four drops hematoxcylin in 1 ml deionized water) provided in the kit. Coverslips were mounted and tissue exam- ined for immunolocalization using bright field micro- scopy with a Zeiss Axioplan microscope (Carl Zeiss, Inc., Go¨ttingen, Germany). Photographs were prepared using a Sony Mavica digital camera mounted on the microscope.

Western Blotting Uterine fluid was obtained from the ligated uterine horn of Alpine (n ¼ 2) and Rambouillet or Ram- bouillet crossbred ewes (n ¼ 5) made unilaterally preg- nant as described earlier (Bazer et al., 1979). Uterine fluid was collected at day 100 of gestation for goats and day 140 of gestation for sheep after captive bolt stunning and exsanguination. Samples of uterine fluid were pre- pared for sodium-dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE) by dilution in an equal part of loading buffer [0.125M Tris, HCl, pH 6.8 con- Fig. 1. Cloning of caprine uterine serpin (CaUS). Uterine serpin cDNA was amplified by reverse transcription-polymerase chain reac- taining 20% (w/v) sucrose, 10% (w/v) SDS, a trace tion (RT-PCR) using primer set 1 and total RNA from endometrium amount of bromophenol blue, and 5% (v/v) 2-mercap- from an ewe at day 140 of gestation (OvUS) and a goat at day 25 of toethanol] and boiling for 3 min. gestation (CaUS). CAPRINE UTERINE SERPIN 265 fragments, respectively. A 1,292 bp CaUS, full-length Localization of Caprine Uterine Serpin consensus cDNA sequence was generated by alignment to Endometrial Epithelium of the two small fragments with the 1,265 bp long cDNA Immunoreactive proteins, detected with an antibody sequence (Fig. 2). There were two positions where se- specific to OvUS, were localized to the endometrial quence information was available from two PCR pro- glandular epithelium of a pregnant nanny collected on ducts only and nucleotides were not identical. In these day 25 of pregnancy. Immunoreactivity was also de- cases, the identity of the largest sequence was chosen for tected in scattered luminal epithelial cells (Fig. 5C). No the consensus sequence because double-strand sequenc- immunoreaction product was detected in endometrium ing was performed on this product. The degree of iden- from a nanny at day 5 of the estrous cycle (Fig. 5A) or in tity between the consensus sequence of the CaUS cDNA sections at any day in which primary antibody was ex- and previously published sequences as determined by cluded (Fig. 5B,D). ClustalW alignment were 98%, 90%, 76%, and 77% for OvUS, BoUS, PoUS1, and PoUS2, respectively (Fig. 3). Presence of Caprine Uterine Serpin As determined using ClustalW, the predicted amino in Uterine Fluid acid sequence of the CaUS precursor showed 96%, 82%, 55%, and 56% identity to OvUS, BoUS, PoUS1, and Western blotting of uterine fluid collected from the PoUS2, respectively (Fig. 4). Like for the ovine, bovine, pregnant uterine horn of a unilaterally-pregnant goat and porcine genes, there are two start codons for the revealed the presence of a protein band at Mr 57,000 CaUS precursor (Fig. 4), resulting in proteins of 429 that reacted with monoclonal antibody to OvUS (Fig. 6). amino acids and 424 amino acids with predicted molec- The Mr of CaUS was identical to the major band of OvUS ular weights of 49,094 and 48,426 (excluding carbohy- identified in Western blots of uterine fluid from preg- drate). The signal peptide extends from amino acids 1 to nant ewes. In addition to the major band at Mr 57,000, 25, resulting in a secreted protein of 404 amino acids and other bands of lower molecular weight were also iden- 46,227 Mr (excluding carbohydrate). There were two N- tified; these were less in number and abundance in goat glycosylation sites for CaUS, at positions 222 and 268; than sheep. the site at 222 is conserved with other uterine serpins except for the bovine gene and the site at 268 is con- DISCUSSION served with all other known uterine serpins. Depending on phylogenetic methods used, the uterine Both the goat and sheep uterine serpins have a 9 serpins are classified as either a separate clade of the amino acid insert in the Helix I region that is not found serpin superfamily (Irving et al., 2000; Peltier et al., in bovine or porcine uterine serpins. A total of 13 amino 2000b) or as a highly-diverged group of the a1-anti- acids in CaUS are different than those for the nearest trypsin clade (Irving et al., 2000). The uterine serpins do homologue, OvUS. One of these is for the site of cleavage not appear to be inhibitory serpins: the hinge region of of the signal sequence, where a single nucleotide substi- the proteins (P17–P9 relative to the scissile bond) does tution (G ! C) changed the cysteine in the sheep, bovine, not show the consensus sequence conserved in inhibi- and porcine genes to a serine. All species have a con- tory serpins (Hopkins et al., 1993) and OvUS did not served KVP at P4–P2 but there are other differences exhibit any serpin-like antiproteinase activity (Ing and between species at the reactive center. The amino acid at Roberts, 1989). The reactive center does not appear to be the putative P1–P10 site (the scissile bond for anti- well conserved within uterine serpins. In addition to proteinase activity) is a valine for CaUS, BoUS, PoUS1, the large 39 amino acid insert between P1 and P10 in the and PoUS2 versus an alanine for OvUS. In the bovine, bovine, there are differences in amino acids at several there is a 39-amino acid insertion containing two positions near the scissile bond including the P1, P10, repeats of KAKEVPAVVKVPM and a similar KEVP- and P20 residues. The goat is more similar to the cow and VVVKVP repeat located between the P1 and P10. This pig at the P1 residue, with all species sharing a valine insert is not present in the goat. The P20 site is not well while the sheep has an alanine. conserved between species, being alanine for goat, One area of conservation between the species is a threonine for sheep and cow, and lysine for the two pig KEVPVVVK motif beginning at P30. Because a motif genes. The hinge region of all five of the uterine serpins similar to VVVK is present in pepstatin, Mathialagan (P17–P9) is distinct from the consensus pattern for and Hansen (1996) proposed that this motif may be inhibitory sequences (Hopkins et al., 1993). Rather than responsible for uterine serpins possessing inhi- the consensus glycine at P15, uterine serpins have bitory activity. Indeed, OvUS does exhibit weak pepsin threonine (CaUS and OvUS), valine (BoUS), or lysine inhibitory activity (Mathialagan and Hansen, 1996). (PoUS1 and PoUS2). Instead of threonine or serine at However, a synthetic peptide corresponding to the P7– P14, uterine serpins have aspartic acid (CaUS, OvUS, P150 region of OvUS did not inhibit pepsin activity and and BoUS) or asparagine (PoUS1 and PoUS) and the the exact role of the conserved KEVPVVVK motif re- P12–P9 region of uterine serpins are not filled with mains unclear. the alanines, glycines, and that predominate in One striking difference between the sequence of CaUS the consensus sequence for inhibitory serpins. All of the as compared to the uterine serpins for other species is uterine serpins have a conserved KEVPVVVK motif that the cysteine at the site of cleavage of the signal upstream of the P1–P10 reactive center. sequence in sheep, cow, and pig is a serine in goat. 266 S. TEKIN ET AL.

Fig. 2. Alignment of CaUS to produce a consensus sequence for number AY679516) ends of CaUS cDNA amplified using primer set CaUS cDNA. Sequences represent double strand sequencing of CaUS 2. Shaded bases are those where there were differences between cDNA amplified using primer set 1 (caus1; GenBank accession number sequences. The nucleotide under the shaded area is the predicted AY679515) and partial single-strand sequences of 30 (caus2; GenBank nucleotide. accession number AY679517) and 50 (caus3; GenBank accession CAPRINE UTERINE SERPIN 267

Fig. 3. Alignment of consensus CaUS cDNA sequence with other uterine serpin sequences. Shaded bases are those that are identical to the goat sequence. The CaUS cDNA (GenBank accession number TPA: BK005554) displayed 98%, 90%, 76%, and 77% identity with OvUS (GenBank accession number M21027), BoUS (L22095), PoUS 1 (M30315), and PoUS2 (NM213845), respectively. 268 S. TEKIN ET AL.

Fig. 4. Alignment of the inferred amino acid sequence of CaUS with three b sheets of a1-antitrypsin that form the consensus structure of the other uterine serpins. Shaded amino acids are those that are identical serpin superfamily are also indicated over their corresponding part of to the goat sequence. The arrowheads represent glycosylation sites and the sequence alignment. The hinge region for prototypical serpins the P1–P10 annotation indicates the position of the putative scissile (P17–P9) is enclosed within a box. There was 96%, 82%, 55%, and 56% bond. Regions corresponding to the nine a helices and strands of the identity of CaUS with OvUS, BoUS, PoUS 1, and PoUS2, respectively.

Nonetheless, it is likely that the signal sequence is are -induced genes (Moffatt et al., 1987; Ing cleaved at a similar position for all four species since the et al., 1989; Leslie and Hansen, 1991) and, at least for requirements for a small and neutral amino acid at sheep, are not produced by the uterus during the estrous position 1 (von Heijne, 1985) is maintained in the goat. cycle until day 16 (Ing et al., 1989). The fact that Overall, the nucleotide and inferred amino acid se- localization of CaUS at day 25 of pregnancy is mostly quences of CaUS and OvUS are very close. This would be limited to the glandular epithelium is reminiscent of the expected based on the recent divergence (about 5– situation in other species. In sheep, uterine serpin is 7 million years ago) of these two species (Carroll, 1988). limited to glandular epithelium after 30 days of proges- Other genes exhibit a similar degree of identity between terone treatment (Ing et al., 1989; Leslie and Hansen, goat and sheep. For example, the degree of identity in 1991) and at day 60 of pregnancy (Stephenson et al., nucleotide sequence between sheep and goat are 94%– 1989) although immunoreactive OvUS is also detected 96% for interferon-t (Ealy et al., 2004) and 93% for in luminal epithelium by day 120 of gestation (Moffatt receptor-a (Mansour et al., 2001), while the et al., 1987; Stephenson et al., 1989). Uterine serpin was degree of identity in amino acid sequence is 99% for also limited to endometrial glands for pregnant cows at interleukin-12 (de Rose et al., 2000), and 98% for growth 180 days of gestation (Leslie et al., 1990) and for preg- hormone releasing hormone receptor (Horikawa et al., nant pigs at day 60 of gestation (Murray et al., 1989). 2001). The major form of CaUS in uterine fluid had a molec- Results of immunohistochemistry and Western blot- ular weight as estimated by SDS–PAGE that was simi- ting confirmed that, like the ovine, bovine, and porcine lar to the molecular weight of OvUS. Assuming, like uterine serpins (Moffatt et al., 1987; Ing and Roberts, OvUS (Hansen et al., 1987), the goat protein is about 1989; Murray et al., 1989; Leslie et al., 1990; Malathy 5.6% carbohydrate and with a predicted molecular et al., 1990; Mathialagan and Hansen, 1996), CaUS is weight of the protein core based on the inferred amino produced in the endometrium and secreted into the acid sequence of the cloned cDNA of 46,000, the approxi- uterine lumen. The uterine serpins from sheep and cow mate molecular weight of the glycosylated protein is CAPRINE UTERINE SERPIN 269

Fig. 5. Immunolocalization of CaUS in goat endometrium. Shown are endometrium from a nanny at day 5 of the estrous cycle (AB) and at day 25 of gestation (CD). Sections were reacted with either a monoclonal antibody to OvUS (AC) or without antibody (BD).

49,000. This value corresponds closely to the predicted Peltier et al., 2000c). It is likely that these lower- molecular weight of glycosylated OvUS (major product molecular weight forms of the protein represent proteo- of Mr ¼ 49,000; Ing and Roberts, 1989). The presence of lytic breakdown products of the larger, secreted form. lower-molecular weight forms of CaUS in uterine fluid The lower-molecular weight forms of OvUS increase in is a phenomenon also seen in sheep (Moffatt et al., 1987; abundance with storage (Moffatt et al., 1987; Hansen and Liu, 1994) and the appearance of the major lower- molecular weight forms can be mimicked by treating purified OvUS with trypsin (Peltier et al., 2000c). In conclusion, the range of species in which uterine serpins are present and expressed in the uterus includes the goat in addition to the previously described sheep, cow, and pig. In all of these species, the uterine serpin is derived primarily from glandular epithelium, is sec- reted into the uterine lumen, and contains sequence characteristics suggesting it is not an inhibitory serpin.

ACKNOWLEDGMENTS The authors thank Ufuk Koca, Department of Horti- cultural Sciences, University of Florida for assistance with alignments; Conrad A. Matthee, Department of Zoology, Stellenbosch University for commenting on the data and for useful advice regarding interpretation; Rebecca Krisher, Department of Animal Sciences, Purdue University for assistance with procuring tis- Fig. 6. Western blotting to detect uterine serpin in uterine fluid from sues; and the University of Florida DNA Sequencing a unilaterally-pregnant sheep at day 140 of gestation and goat at day 100 of gestation. Represented in each lane are sheep uterine fluid (SUF: Core Laboratory for conducting sequencing. This is 9 mg), goat uterine fluid (GUF: 5 mg) and ovalbumin (Ova: 5 mg), which Journal Series No. 10327 of the Florida Agricultural was used as a negative control. Experiment Station. 270 S. TEKIN ET AL.

REFERENCES progesterone-induced proteins of the sheep uterus. Domest Anim Endocrinol 7:517–526. Baumbach GA, Ketcham CM, Richardson DE, Bazer FW, Roberts RM. Malathy PV, Imakawa K, Simmen RC, Roberts RM. 1990. Molecular 1986. Isolation and characterization of a high molecular weight cloning of the uteroferrin-associated protein: A major progesterone- stable pink form of uteroferrin from uterine secretions and allantoic induced serpin secreted by the porcine uterus, and the expression of fluid of pigs. J Biol Chem 261:12869–12878. its mRNA during pregnancy. Mol Endocrinol 4:428–440. Bazer FW, Roberts RM, Basha SMM, Zavy MT, Carton D, Barron DH. Mansour MM, Machen MR, Tarleton BJ, Wiley AA, Wower J, Bartol 1979. Method for obtaining ovine uterine secretions from unilater- FF, Goyal HO. 2001. Expression and molecular characterization of ally pregnant ewes. J Anim Sci 49:1522–1527. estrogen receptor alpha messenger RNA in male reproductive organs Bendtsen JD, Nielsen H, von Heijne G, Brunak S. 2004. Improved of adult goats. Biol Reprod 64:1432–1438. prediction of signal peptides: Signal P 3.0. J Mol Biol 340:783–795. Mathialagan N, Hansen TR. 1996. Pepsin-inhibitory activity of the Carroll RL. 1988. Vertebrate palaeontology and evolution. New York: uterine serpins. Proc Natl Acad Sci USA 93:13653–13658. Freeman. Moffatt J, Bazer FW, Hansen PJ, Chun PW, Roberts RM. 1987. De Rose R, Scheerlinck JP, Casey G, Wood PR, Tennent JM, Chaplin Purification, secretion, and immunocytochemical localization of the PJ. 2000. Ovine interleukin-12: Analysis of biologic function and uterine milk proteins, major progesterone-induced proteins in the species comparison. J Interferon Cytokine Res 20:557–564. uterine secretions of the sheep. Biol Reprod 36:419–430. Ealy AD, Wagner SK, Sheils AE, Whitley NC, Kiesling DO, Johnson Murray MK, Malathy PV, Bazer FW, Roberts RM. 1989. Structural SE, Barbato GF. 2004. Identification of interferon-t isoforms expres- relationship, biosynthesis, and immunocytochemical localization of sed by the peri-implantation goat (Capra hircus) conceptus. Domest uteroferrin-associated basic . J Biol Chem 264:4143– Anim Endocrinol 27:39–49. 4150. Hansen PJ. 1998. Regulation of uterine immune function by progester- Peltier MR, Liu WJ, Hansen PJ. 2000a. Regulation of lymphocyte one—lessons from the sheep. J Reprod Immunol 40:63–79. proliferation by uterine serpin: Interleukin-2 mRNA production, Hansen PJ, Liu W-J. 1994. Biochemical/physiological properties of endo- CD25 expression, and responsiveness to interleukin-2. Proc Soc Exp metrial serpin-likeproteins. Adv Contracept Delivery Sys 10:339–353. Biol Med 223:75–81. Hansen PJ, Ing NH, Moffatt RJ, Baumbach GA, Saunders PT, Bazer Peltier MR, Raley LC, Liberles DA, Benner SA, Hansen PJ. 2000b. FW, Roberts RM. 1987. Biochemical characterization and biosynth- Evolutionary history of the uterine serpins. J Exp Zool 288:165– esis of the uterine milk proteins of the pregnant sheep uterus. Biol 174. Reprod 36:405–418. Peltier MR, Grant TR, Hansen PJ. 2000c. Distinct physical and Higgins D, Thompson J, Gibson T, Thompson JD, Higgins DG, Gibson structural properties of the ovine uterine serpin. Biochim Biophys TJ. 1994. CLUSTAL W: Improving the sensitivity of progressive Acta 1479:37–51. multiple sequence alignment through sequence weighting, position- Segerson EC, Moffatt RJ, Bazer FW, Roberts RM. 1984. Suppression of specific gap penalties, and weight matrix choice. Nucleic Acids Res phytohemagglutinin-stimulated lymphocyte blastogenesis by ovine 22:4673–4680. uterine milk protein. Biol Reprod 30:1175–1186. Hopkins PC, Carrell RW, Stone SR. 1993. Effects of mutations in the Silverman GA, Bird PI, Carrell RW, Church FC, Coughlin PB, Gettins hinge region of serpins. Biochemistry 32:7650–7657. PG, Irving JA, Lomas DA, Luke CJ, Moyer RW, Pemberton PA, Horikawa R, Gaylinn BD, Lyons CE, Jr., Thorner MO. 2001. Molecular Remold-O’Donnell E, Salvesen GS, Travis J, Whisstock JC. 2001. cloning of ovine and bovine growth hormone-releasing hormone re- The serpins are an expanding superfamily of structurally similar but ceptors: The ovine receptor is C-terminally truncated. Endocrinology functionally diverse proteins. Evolution, mechanism of inhibition, 142:2660–2668. novel functions, and a revised nomenclature. J Biol Chem 276: Ing NH, Roberts RM. 1989. The major progesterone-induced proteins 33293–33296. secreted into the sheep uterus are members of the serpin superfamily Skopets B, Hansen PJ. 1993. Identification of the predominant proteins of inhibitors. J Biol Chem 264:3372–3379. in uterine fluids of unilaterally pregnant ewes that inhibit lympho- Ing NH, Francis H, McDonnell JJ, Amann JF, Roberts RM. 1989. cyte proliferation. Biol Reprod 49:997–1007. Progesterone induction of the uterine milk proteins: Major secretory Skopets B, Liu WJ, Hansen PJ. 1995. Effects of endometrial serpin-like proteins of sheep endometrium. Biol Reprod 41:643–654. proteins on immune responses in sheep. Am J Reprod Immunol Irving JA, Pike RN, Lesk AM, Whisstock JC. 2000. Phylogeny of the 33:86–93. serpin superfamily: Implications of patterns of amino acid conserva- Stephenson DC, Leslie MV, Low BG, Newton GR, Hansen PJ, Bazer tion for structure and function. Genome Res 10:1845–1864. FW. 1989. Secretion of the major progesterone-induced proteins of Leslie MV, Hansen PJ. 1991. Progesterone-regulated secretion of the the sheep uterus by caruncular and intercaruncular endometrium of serpin-like proteins of the ovine and bovine uterus. Steroids 56:589– the pregnant ewe from days 20 to 140 of gestation. Domest Anim 597. Endocrinol 6:349–362. Leslie MV, Hansen PJ, Newton GR. 1990. Uterine secretions of the cow von Heijne G. 1985. Signal sequences. The limits of variation. J Mol Biol contain proteins that are immunochemically related to the major 84:99–105.