| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
research-article |
Division of Reproductive Sciences,3 Oregon National Primate Research Center, Oregon Health & Science University West Campus, Beaverton, Oregon 97006
Departments of Medical Informatics4 and Obstetrics and Gynecology,5 Oregon Health & Science University, Portland, Oregon 97239
ABSTRACT
Proteolytic activities are essential for follicular growth, ovulation, as well as for luteal formation and regression. Using suppression subtractive hybridization (SSH), a novel mouse ovary-selective gene (termed protease serine 35, Prss35) was identified. Analysis of the mouse genome database using the Prss35 sequence led to the identification of a homologous protease (protease serine 23, Prss23). PRSS35 possesses general features that are characteristic of serine (Ser) proteases, but is unique in that the canonical Ser that defines this enzyme family is replaced by a threonine (Thr). In contrast, PRSS23 possesses the standard catalytic Ser typical for this family of proteases. As determined by real-time polymerase chain reaction (PCR), the Prss35 mRNA levels increased around the time of ovulation and remained elevated in the developing corpus luteum. Steroid ablation/replacement studies demonstrated progesterone-dependent regulation of Prss35 gene expression prior to follicle rupture. Prss35 gene expression was localized to the theca cells of pre-antral follicles, the theca and granulosa cells of pre-ovulatory and ovulatory follicles, as well as to the developing corpus luteum. In contrast, Prss23 mRNA levels decreased transiently after ovulation induction and again in the postovulatory period. Prss23 gene expression was noted primarily in the granulosa cells of the secondary/early antral follicles. PRSS35 and PRSS23 orthologs in the rat, human, rhesus macaque, chimpanzee, cattle, dog, and chicken were identified and found to be highly homologous to one another (75–99% homology). Collectively, these results suggest that the PRSS35 and PRSS23 genes have been conserved as critical ovarian proteases throughout the course of vertebrate evolution.
corpus luteum, follicle, granulosa cell, ovary, theca cell
Proteases comprise a group of structurally and functionally diverse proteins that catalyze the hydrolysis of peptide bonds. The highly selective and limited cleavage of specific protein substrates by the five major groups of proteases (serine-, metallo-, aspartic-, cysteine-, and threonine-proteases) are critical for a number of essential biological processes [1]. For example, proteases remodel extracellular matrix (ECM) proteins, modulate growth factor/cytokine actions by regulating their bioavailability, and control the levels of certain cell surface/receptor proteins. Therefore, proteases are critical determinants of cellular proliferation, differentiation, migration, and adhesion. As such, enzymatic proteolysis is indispensable for maintaining homeostasis, tissue remodeling, and angiogenesis or angiolysis [2, 3]. The recent completion of several large-scale genome sequencing projects and the development of high-throughput differential screening methodologies have led to the identification of hundreds of mouse and human proteases [4, 5]. From these studies, it has been determined that the mouse and human genomes contain 628 and 553 protease-encoding genes, respectively [4]. Many of theses proteases are novel and require additional studies to determine their functional significance. Although limited information exists with regard to their biochemical and cellular activities, it is highly likely that some of these novel proteases play critical roles in ovarian physiology.
The female reproductive tract is unique in that rapid and extensive tissue remodeling is required throughout the course of each menstrual or estrus cycle. In the mammalian ovary, the tightly regulated proteolytic remodeling of the extracellular architecture is required for follicular growth, breakdown of the follicular wall during ovulation, as well as for luteal formation and regression [6, 7]. To date, several matrix-degrading proteases, such as matrix metalloproteinases (MMPs) and their tissue inhibitors (TIMPs), a disintegrin and metalloproteinase with thrombospondin-like motifs (ADAMTS) family members, and the plasminogen activator (PA)/plasmin system, have been postulated to play critical roles in the remodeling of the ECM throughout the ovarian cycle [6–15]. This complex expression pattern of proteases and associated inhibitors that are responsible for the cyclic changes in the ovary is, in turn, regulated by a variety of endocrine (i.e., FSH and LH) and intra-ovarian mediators, (i.e., prostaglandins [PGs], cytokines, steroid hormones, and growth factors) [6, 9, 13–15]. In addition, proteases are key regulators of follicular development, since they are responsible for determining growth factor (e.g., insulin-like growth factors) availability and consequently, their capacity to bind and activate the appropriate receptors [16].
Recently, the roles of several proteases in ovarian function have been investigated through the creation of specific null mutant mice. However, mice that lack the genes for several of the MMPs (MMP-1, MMP-2, MMP-7, MMP-9, MMP-11 and MMP-12), TIMPs (TIMP-1, TIMP-2, and TIMP-3) or the components of the PA/plasmin system (urokinase type PA, tissue type PA, and plasminogen) have been found to be fertile [2, 6–8, 17–20]. To date, only the deletion of the gene that encodes ADAMTS1 has led to an ovarian phenotype that includes a decrease in ovulatory efficiency (70% reduction compared to wild-type controls) [21, 22]. These findings indicate that either redundant proteolytic systems exist within the ovary or key proteolytic elements critical for ovarian function have yet to be identified. In support of the latter, we have recently identified a mouse gene that is selectively expressed in the ovary [23] and encodes a novel enzyme with homology to Ser protease family members. The studies described herein were conducted to determine the molecular and cellular characteristics of this unique putative Ser-like protease. In-depth bioinformatic analyses were also carried out to identify homologous proteases and characterize their evolutionary relationships.
All of the protocols were approved by the Institutional Animal Care and Use Committee at the Oregon National Primate Research Center (ONPRC). Female C57BL/6 mice, which were 21 days of age upon arrival, were purchased from Jackson Laboratories (Bar Harbor, ME). At 25 days of age, one group of mice was sacrificed to provide the unstimulated control ovaries. A second group of mice was injected i.p. with 5 IU each of eCG. Forty-eight hours after administration of eCG, one group of mice was killed to provide ovaries at the pre-ovulatory phase of the stimulated estrus cycle. The remaining mice were injected i.p. with 5 IU hCG. Subgroups of the latter were killed at various intervals between 3 h and 48 h post-hCG treatment. The ovaries collected between 3 h and 12 h post-hCG treatment encompass the interval preceding follicular rupture, whereas the 24 h and 48 h post-hCG treatment groups encompass the postovulatory (i.e., luteal) phase of the stimulated estrus cycle.
Steroid ablation/replacement studies were performed to investigate the regulation of mouse Prss35 and Prss23 gene expression by steroid hormones. To inhibit steroid synthesis, eCG-primed C57BL/6 mice were injected i.p. with 12, 60, 120, or 240 mg/kg body weight (BW) of the 3ß-hydroxysteroid dehydrogenase (HSD3B) inhibitor trilostane (TRL; Sanofi-Aventis, Malvern, PA), 3 h after hCG injection. Another group of mice was cotreated with TRL (240 mg/kg BW) plus the synthetic progestin R5020 (Promegestone; 0.5 or 5 mg/kg BW; DuPont NEN, Boston, MA), 3 h after hCG injection. The designated amount of TRL and R5020 were dissolved in 50 µl of dimethylsulfoxide. Control animals received vehicle i.p. at 3 h post-hCG injection. Ovaries from each mouse were collected 8 h after hCG administration.
For RNA isolation and steroid extraction, ovaries were flash frozen in liquid N2 and stored at –80°C. For in situ hybridization, ovaries were embedded in O.C.T. compound (Sakura Finete, Torrance, CA) using a liquid propane bath held in liquid N2, and then stored at –80°C.
The general care and housing of rhesus macaques (Macaca mulatta) at the ONPRC have been described previously [24]. Ovaries were obtained from anesthetized adult female rhesus macaques during an aseptic ventral midline laparotomy [25], and then flash frozen in liquid N2 and stored at –80°C.
RNA Extraction and cDNA Synthesis
RNA was extracted from each ovary using TRIzol (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. For cDNA synthesis, RNA (1 µg) was DNase treated and then reverse-transcribed using random primers (50 ng; Invitrogen) and Moloney Murine Leukemia Virus reverse transcriptase (200 U; Invitrogen).
Total RNA (20 µg) was separated on 1% agarose-formaldehyde gels and transferred to a nylon membrane, as described previously [23]. Prior to transfer, RNA quality and concentration were assessed by ethidium bromide staining followed by visualization under UV light. The nylon membranes were pre-hybridized for 2–6 h at 42°C in 5x sodium chloride-sodium phosphate-EDTA, 50% formamide, 5x Denhardt solution and 0.25% SDS. The probe, which corresponds to the SSH-derived novel gene designated P3D9, was generated from a cloned cDNA fragment obtained from the previously described ovary-selective cDNA library [23]. The PCR-amplified P3D9 cDNA insert (249 bp) was radiolabeled with [32P]-dCTP using the Rediprime II DNA Labeling System (GE Healthcare Bio-Sciences Corp., Piscataway, NJ). The probe was denatured in a boiling water bath for 5 min prior to quenching with ice. Each membrane was hybridized with the relevant probe overnight at 42°C in the same solution used for prehybridization. The membrane was then washed in 2x sodium chloride/sodium citrate (SSC) and 0.1% SDS at room temperature three times for 5 min, followed by two washes in 0.125x SSC and 0.25% SDS for 15 min at 60°C. The blots were rinsed in 4x SSC and imaged using a phosphoimager (BioRad, Hercules, CA). Equivalent RNA loading was verified by probing the same (stripped) blots with a [32P]-dCTP labeled, sequence-verified ß-actin PCR product [23].
Full-Length cDNA Sequence Determination
An ovarian 
phage cDNA library was constructed and amplified using the SMART cDNA Library Construction kit (Clontech Laboratories Inc., Mountain View, CA) according to the manufacturer's directions. Briefly, the RNA used to generate double-stranded cDNA for insertion into the phage vector (TriplEx2; Clontech) was isolated from ovaries obtained 6 h after hCG injection in eCG-primed mice. The resulting cDNA was ligated into the TriplEx2 expression vector and packaged using the Gigapack III Gold Packaging Extract (Stratagene, La Jolla, CA). Recombinant phage that comprised the mouse ovarian cDNA library was screened with a [32P]dCTP-labeled P3D9 probe, which was generated as described above. Positive clones were screened twice and plaque purified. The recombinant TriplEx2 phage was subsequently converted to the corresponding pTriplEx2 plasmid. For sequencing, plasmid DNA was isolated from individual clones using the QIAfilter Plasmid Midi Kit (Qiagen Inc., Valencia, CA). Sequencing was performed using the ABI 3130 Genetic Analyzer (Applied Biosystems, Foster City, CA). The full-length P3D9 cDNA sequence was deposited in GenBank [26] with accession number DQ223037. In the mouse genome database, the P3D9 gene sequence has been annotated as protease serine 35 (Prss35).
Using the full-length Prss35 cDNA sequence as the query, a corresponding paralogous gene, protease serine 23 (Prss23, GenBank accession no. AK034439) was identified in the National Center for Biotechnology Information (NCBI) mouse genome database. For Prss23 sequence validation, the putative open reading frame (ORF) was amplified by PCR using mouse ovarian cDNA. The primers used are listed in Table 1. PCR was performed with initial denaturation at 95°C for 5 min, followed by denaturation at 95°C for 45 sec, annealing for 45 sec (see Table 1 for annealing temperatures), and extension at 72°C for 90 sec (35 cycles). The PCR product was purified (QIAquick PCR Purification Kit; Qiagen) and sequenced by the ONPRC Molecular and Cellular Biology Core facility. The corresponding contig was generated using the Vector NTI Suite 7.1 bioinformatics software (Invitrogen).
|
A dot blot that contained ~2 µg of mRNA isolated from 18 different adult mouse tissues, including the ovary and four whole embryos at different stages of development, was purchased from Clontech. A radiolabeled probe for mouse Prss35 was generated using the 870-bp PCR product that encompasses nucleotides 509-1378 of the full-length cDNA sequence (see Table 1 for primer sequence). The mouse Prss23 probe was generated by PCR amplification of a 486-bp cDNA fragment using the primers outlined in Table 1. The PCR parameters for Prss35 and Prss23 included an initial denaturation step at 95°C for 5 min, followed by denaturation at 95°C for 45 sec, annealing for 45 sec (see Table 1 for temperatures), and extension at 72°C for 90 sec (38 cycles). The PCR products were purified, sequenced as described above, and labeled with [32P]dCTP using the Rediprime II DNA labeling system (GE Healthcare Bio-Sciences). All hybridization and washing procedures were performed according to the manufacturer's directions. The hybridization signals were assessed using a phosphoimager (BioRad).
The full-length sequences of the mouse Prss35 and Prss23 genes were used to design primer sets and TaqMan probes for real-time PCR assays (Primer Express software; Applied Biosystems). The oligonucleotide primers were synthesized by Invitrogen, while the TaqMan probes were synthesized by Applied Biosystems. The forward and reverse primers, as well as the TaqMan probe sequences for the mouse Prss35 and Prss23 are listed in Table 1. The expression levels of Prss35 and Prss23 mRNAs were analyzed using the TaqMan PCR Core Reagent Kit with the ABI PRISM 7900HT Sequence Detection System (Applied Biosystems). The PCR reactions were conducted in sealed 96-well optical plates under the following conditions: 2 min at 50°C, 10 min at 95°C, and 40 cycles of 15 sec at 95°C (DNA melting) and 1 min at 60°C (primer annealing/extension). In each sample, primers and a Taqman probe that allowed real-time PCR amplification of 18S rRNA were included as an internal control, as described previously [14]. The number of amplification cycles for the fluorescence to reach a determined threshold level (CT) was recorded for every unknown sample and an internal standard curve. The internal standard curve, which was used for relative mRNA quantification, was generated from five 10-fold dilutions of mouse ovarian cDNA pooled from each of the different stages of a stimulated estrus cycle. The CT values of the unknown samples were used to extrapolate the amount of RNA equivalents from the internal standard curve. The values for Prss35 and Prss23 were divided by the 18S rRNA values to normalize the data.
Ovarian homogenates were prepared in 500 µl of 0.05 M phosphate buffer (pH 7.4). Each homogenate was extracted with 3 ml ether, with the organic phase evaporated to dryness and stored in redistilled ethanol until analyzed by RIA at the ONPRC Endocrine Service Core. As TRL cross-reacts with the progesterone anti-sera used in the RIA, progesterone was separated from TRL and other components by column chromatography prior to performing the RIA as previously described [27, 28]. The aqueous fraction from each homogenate was assayed for protein content using the Bio-Rad DC Protein Assay kit with BSA as the standard (Bio-Rad). The ovarian progesterone levels are expressed as pg per µg of protein.
In situ hybridization riboprobes for the mouse Prss35 and Prss23 genes were generated by PCR using the primers listed in Table 1. The PCR parameters included a 5-min initial denaturation step at 95°C, followed by 45 sec of denaturation at 95°C, 45 sec of annealing (see Table 1 for annealing temperatures), and 90 sec of extension at 72°C (35 cycles). The purified PCR products were ligated with SP6 or T7 promoters using the Lign'Scribe No-Cloning Promoter Addition Kit (Ambion, Austin, TX). The products were purified and sequenced, as described above. SP6 or T7 RNA polymerase was used to synthesize radiolabeled sense and antisense riboprobes using [35S]UTP and the MAXIscript In Vitro Transcription Kit (Ambion). Mouse ovarian cryosections (10 µm) were placed onto microscope slides (SuperFrost Plus; Fisher Scientific, Pittsburgh, PA) and stored at –80°C. Tissue sections were fixed in 4% paraformaldehyde for 15 min at 4°C. Hybridization was performed using the mRNAlocator kit (Ambion) according to the manufacturer's directions. Each slide was coated with emulsion NTB-2 (Eastern Kodak Co., Rochester, NY) and exposed for 8 to 10 days at 4°C. Sections hybridized with [35S]UTP-labeled sense probes served as negative controls. The slides were processed using D-19 developer and fixer (Eastman Kodak Co.) and counterstained with hematoxylin. Bright and dark field images were used to visualize the tissue histology and hybridized probe, respectively.
Identification and Analysis of PRSS35 and PRSS23 Orthologs
Using the mouse Prss35 and Prss23 sequences as the query, the corresponding human orthologs (PRSS35, GenBank accession no. NM_153362; PRSS23, NM_007173) were identified in the NCBI human genome database. A PCR approach was subsequently employed to amplify the corresponding rhesus macaque PRSS35 and PRSS23 ORFs using the primer sets developed from the human sequences (Table 1). Using rhesus macaque ovarian cDNA, PCR was performed with the following parameters: 5 min of initial denaturation at 95°C, 45 sec of denaturation at 95°C, 45 sec of annealing (see Table 1 for annealing temperatures), and 90 sec of extension at 72°C (35 cycles). The resultant PCR products for rhesus macaque PRSS35 and PRSS23 were purified and sequenced as described above. The rhesus macaque PRSS35 and PRSS23 cDNA sequences have been deposited in GenBank [26] with accession numbers DQ223038 and DQ223039, respectively.
The NCBI conserved protein domain database (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=cdd) and the MEROPS database (http://merops.sanger.ac.uk/index.html) were used to compare the putative catalytic domains of PRSS35 and PRSS23 with known Ser proteases. The signal peptides and proprotein cleavage sites were predicted using the Prop 1.0 and SignalP 3.0 programs (http://www.cbs.dtu.dk).
The sequence alignment and phylogenetic analysis were performed with the following orthologs extracted from the NCBI nonredundant databases, using the mouse PRSS35 and PRSS23 proteins as the query: for PRSS35, rat, NP_001008560; human, NP_699193; chimpanzee, XP_527441; bovine, NP_001030534; dog, XP_539023; and chicken, XP_419860; and for PRSS23, rat, NP_001007692; human, NP_009104; chimpanzee, XP_508909; bovine, XP_581199; dog, XM_542268; and chicken XP_417210. The full-length PRSS35 and PRSS23 amino acid sequences from different species were compared using the CLUSTALW sequence alignment program [29]. The phylogenetic tree was constructed by the neighbor-joining method using the PHYLIP software (http://evolution.gs.washington.edu/phylip.html) [30].
Multiple data-points per experiment were analyzed by one-way ANOVA followed by the Student-Newman-Keuls post-hoc test using the SigmaStat software (SPSS Inc., Chicago, IL). A P value < 0.05 was considered to be statistically significant.
Identification of a Putative Ovary-Selective Protease
We have previously reported the isolation of 83 novel genes, via the differential screening technique SSH, that are selectively or exclusively expressed in the mouse ovary [23]. In the current study, the expression of one such novel cDNA (SSH cDNA clone P3D9) was analyzed throughout the course of a mouse stimulated estrus cycle. Northern blot analysis showed that the expression of the P3D9 gene was detectable at all phases of the stimulated estrus cycle (Fig. 1). After hCG treatment of eCG-primed mice, P3D9 mRNA expression appeared to increase at the end of the peri-ovulatory interval (i.e., 9 h and 12 h post-hCG administration, follicular rupture at ~14 h post-hCG administration). This high level of gene expression was maintained through the postovulatory period (48 h post-hCG administration).
|
Since SSH yields only partial cDNA fragments (249 bp for clone P3D9), an alternative methodology was employed to isolate the full-length cDNA sequence. For P3D9, a cDNA library constructed from mouse ovarian cDNA was screened to identify the full-length sequence. Of five clones isolated, one clone contained a 3439-bp insert that corresponded to the apparent size of the P3D9 transcript, as determined by Northern blot analysis (Fig. 1). The isolated cDNA contains a single large ORF that is to encode a protein of 409 amino acids with a molecular mass of 45.7 kDa (Fig. 2). The 3'-untranslated region of the P3D9 gene is quite large (2057 bp), and contains a consensus polyadenylation signal sequence, as well as a long terminal repeat (position 2029–2627 bp) derived from a mouse retrotransposon [31]. The protein encoded by the P3D9 gene possesses a putative signal sequence at the N-terminus (Met1-Gly20) and proprotein cleavage sites (Arg122-Gln123 and Arg208-Glu209), which indicate that this protein is directed to the secretory pathway. The protein also contains a consensus His active site domain (position: Leu165-Cys170), which is indicative of trypsin proteases (PROSITE; PS00134: [Leu, Ile, Val, or Met]-[Ser or Thr]-Ala-[Ser, Thr, Ala, or Gly]-His-Cys). Analysis of the NCBI DNA database using the Basic Local Alignment Search Tool (BLAST) [32] revealed that the full-length P3D9 cDNA sequence matches a full-length RIKEN cDNA clone (GenBank accession no. AK031411). The P3D9 and RIKEN sequences also match the mouse Prss35 (GenBank accession no. BC075675), a putative protease that was identified from a high-throughput, full-length cDNA sequencing project [33]. To maintain consistency of the nomenclature, the gene that corresponds to the full-length P3D9 sequence will be referred to as Prss35. It was noted from the BLAST results that the putative protein encoded by the P3D9/Prss35 gene possesses significant homology (E probability value of 2e–93) to another uncharacterized protease, termed Prss23, which was also originally derived from a high-throughput RIKEN full-length cDNA sequencing project (Supplemental Fig. 1, available online at www.biolreprod.org). The database–derived mouse Prss23 sequence was compared and validated by PCR amplification of the putative ORF using mouse ovarian cDNA and overlapping primer sets. The mouse PRSS23 is similar to PRSS35 in that it possesses a putative signal sequence at the N-terminus and two proprotein cleavage sites (Supplemental Fig. 1, available online at www.biolreprod.org). The PROSITE domain noted in the mouse PRSS35 protein was also identified in the mouse PRSS23 protein (Fig. 3A).
|
|
Analysis of Mouse PRSS35 and PRSS23 Proteins
The deduced mouse PRSS35 and PRSS23 proteins share 58.9% homology and 46.9% identity at the amino acid level. Both PRSS35 and PRSS23 proteins possess His and Asp residues (positions: PRSS35 169 and 272; PRSS23 174 and 245, respectively) that are highly conserved among Ser proteases (Fig. 3A). Surprisingly, within the PRSS35 protein, the canonical Ser residue that defines Ser proteases is substituted by a Thr at position 342. Although highly homologous to PRSS35, PRSS23 contains the canonical Ser residue at position 315, and thus possesses a consensus catalytic triad. Several amino acids adjacent to the catalytic Ser are known to be highly conserved in Ser proteases [1, 34, 35]. One such amino acid includes a Gly residue that serves to stabilize the catalytic Ser residue and its amide group by hydrogen bonding the carbonyl oxygen of the substrate [36]. Alignment of the mouse PRSS35 and PRSS23 sequences with representative Ser family members revealed that the conserved Gly residue is present in mouse PRSS35 and PRSS23 at positions 340 and 313, respectively (Fig. 3B). Furthermore, another Gly residue, which is positioned at the C-terminus of the catalytic Ser residue, was found to be conserved in mouse PRSS35 and PRSS23. Mouse PRSS35 and PRSS23 also possess an Asp residue that is in close proximity to the putative catalytic Thr and Ser residues, respectively (Fig. 3B). For trypsin and trypsin-related proteases, this Asp residue lies at the bottom of a pocket that is critical in determining substrate specificity [1, 34].
Tissue Distribution of Mouse Prss35 and Prss23 Gene Expression
Initial Northern blot analysis demonstrated that mouse Prss35 mRNA expression was high in the ovary but was absent in a variety of other mouse tissues (i.e., brain, heart, kidney, spleen, thymus, liver, stomach, small intestine, skin, lung, testis, and skeletal muscle; data not shown). To investigate fully the expression of the Prss35 and Prss23 genes in different mouse tissues, an mRNA dot-blot was employed (Fig. 4). The only significant source of Prss35 gene expression was found to be the ovary. In contrast, the expression of Prss23 mRNA was detected in a wide range of mouse tissues analyzed. High levels of Prss23 mRNA expression were observed in the pancreas, submaxillary gland, and reproductive tissues, including the ovaries.
|
Phase-Selective Expression Profile of Mouse Prss35 and Prss23 mRNA
To investigate the expression of mouse Prss35 and Prss23 mRNA throughout the course of a stimulated estrus cycle, real-time PCR was performed (Fig. 5). The mouse Prss35 mRNA levels in ovaries increased towards the end of the peri-ovulatory interval (4.2-fold increase, control vs. 8 h post-hCG, P < 0.05) and remained elevated after ovulation (3.8-fold increase, control vs. 24 h post-hCG, P < 0.05). In contrast, the mouse Prss23 mRNA levels were highest in ovaries from the control and eCG-primed mice. The lowest level of Prss23 gene expression was observed 4 h after hCG administration (2.9-fold decrease relative to control, P < 0.05), and again at 24 h and 48 h post-hCG treatment (2.4-fold and 2.9-fold decreases relative to control, respectively; P < 0.05).
|
Progesterone-Dependent Regulation of Mouse Prss35 Gene Expression in the Peri-Ovulatory Interval
The administration of TRL at 3 h post-hCG reduced the levels of ovarian progesterone in a dose-dependent manner (Fig. 6A, upper panel). The ovarian progesterone levels decreased significantly at 60, 120, and 240 mg/kg BW of TRL relative to the control group, which received only vehicle (P < 0.05). As determined by real-time PCR, the ovarian Prss35 mRNA levels decreased significantly at 120 and 240 mg/kg BW of TRL (58% and 43% decreases relative to vehicle group, respectively; P < 0.05; Fig. 6A, lower panel). Supplementation of TRL-treated animals with the progestin R5020 restored the Prss35 mRNA expression levels to the control values (Fig. 6B). In contrast to the Prss35 mRNA levels, there was no significant change in the Prss23 mRNA levels after TRL treatment (data not shown).
|
Cellular Localization of Mouse Prss35 and Prss23 mRNAs
To ascertain the cellular sites of Prss35 and Prss23 gene expression, in situ hybridization was performed using specific [35S]UTP-labeled riboprobes (Fig. 7). The expression of mouse Prss35 mRNA was restricted to the theca cells at the early stages of follicular development. However, after eCG stimulation of follicular growth, mouse Prss35 mRNA was expressed in the granulosa cells of pre-ovulatory follicles. After hCG administration, the highest level of Prss35 riboprobe hybridization was observed in granulosa cells within ovulatory follicles around the time of rupture. High levels of mouse Prss35 mRNA were also localized to the cells that comprise the corpora lutea. In contrast to the dynamic pattern of Prss35 gene expression, mouse Prss23 mRNA was primarily detected in the granulosa cells of developing follicles. Significant Prss23 expression was mostly associated with secondary/early antral follicles throughout the course of the stimulated estrus cycle. While the granulosa cells in the antral follicles expressed Prss23 mRNA, these expression levels were considerably lower than those in the secondary/early antral follicles. No difference in Prss23 mRNA expression was observed between the atretic and healthy secondary/early antral follicles. No signal was detected when control sense probes for the mouse Prss35 and Prss23 genes were hybridized to the ovarian sections.
|
Identification and Phylogenetic Comparison of PRSS35 and PRSS23 Orthologs
PCR amplification of rhesus macaque ovarian cDNA using primers derived from the human PRSS35 and PRSS23 genes led to the identification of the rhesus macaque PRSS35 and PRSS23 genes. The corresponding rhesus macaque PRSS35 and PRSS23 proteins share 95.9% and 99.0% homology with the human PRSS35 and PRSS23 proteins, respectively. Using mouse PRSS35 and PRSS23 as the query, additional orthologous proteins were identified in the rat, chimpanzee, bovine, dog, and chicken genome databases. Of the annotated PRSS35 and PRSS23 orthologs in the databases, the dog PRSS35 and chimpanzee PRSS23 protein sequences were found to be truncated at the N- and C-terminus, respectively. The dog PRSS35 protein lacks 55 residues at the N-terminus, which are highly conserved in other PRSS35 orthologs. The chimpanzee PRSS23 (GenBank accession no. XM_508909) lacks seven nucleotides downstream of position 870, resulting in a premature stop codon. The dog Prss35 and chimpanzee PRSS23 genomic DNA sequences were re-analyzed and compared to the sequences from the other species. According to this analysis, these truncated proteins have come about from improper annotation of the respective genomic DNA sequences. The amino acid sequences generated from the corrected gene sequences (Supplemental Fig. 2, available online at www.biolreprod.org) were used to perform the sequence alignment and phylogenetic analysis. The alignment of the PRSS35 and PRSS23 proteins from eight different species revealed a high degree of homology between the individual orthologs (Fig. 8). PRSS35 protein homology ranges from 74.9% (rat vs. chicken) to 99.3% (human vs. chimpanzee). The mouse PRSS35 protein shares 93.9% homology with the rat ortholog, while the PRSS35 orthologs in human, rhesus macaque, and chimpanzee share 95.9–99.3% homology. Likewise, the PRSS23 proteins show high homology, ranging from 86.7% (rat vs. chicken) to 99.7% (human vs. chimpanzee). The mouse PRSS23 protein is 97.4% homologous to the rat ortholog. The PRSS23 orthologs in human, rhesus macaque, and chimpanzee are >99% homologous to one another. In addition to possessing canonical catalytic His and Asp residues, all of the PRSS35 orthologs contain the unique catalytic domain that contains a Thr in place of the standard Ser residue. Likewise, the putative catalytic triad, which encompasses a catalytic Ser residue, is completely conserved in all PRSS23 orthologs.
|
|
Phylogenetic analysis was subsequently utilized to investigate the relatedness of the PRSS35 or PRSS23 orthologs (Fig. 9). The mouse and rat PRSS35 orthologs clustered together, as did the primate (human, rhesus macaque, and chimpanzee) PRSS35 sequences. However, the chicken, dog, and bovine PRSS35 sequences were separate from the rodent and primate branches. The phylogenetic relationships of PRSS23 were very similar to those of PRSS35, with rodents and primates showing the greatest degree of relatedness.
|
PRSS35 and PRSS23 Genomic Organizations
To determine the chromosomal locations and potential exon/intron organizations of the PRSS35 and PRSS23 genes, the mouse and human cDNA sequences were aligned using the available genomic DNA databases (Fig. 10). The mouse Prss35 gene was located on chromosome 9 (9E3.1) and found to possess two exons separated by a single large intron (10 927 bp). The human PRSS35 gene localized to a syntenic region on the long arm of chromosome 6 (6q14.2) and was also found to consist of two exons separated by a 10 679-bp intron. The mouse and human PRSS23 genes localized to syntenic chromosomal regions (chromosome 7, 7D3 and chromosome 11, 11q14.1, respectively) and also possessed a two exon/1 intron organization, with the spacings between the first and second exon being 6575 bp and 6990 bp, respectively.
|
We have identified a novel mouse gene that corresponds to Prss35 through the differential screening technique SSH [23]. Subsequent BLAST [32] analysis using our original full-length Prss35 sequence revealed that a related, novel protease (Prss23) exists in the mouse genome. As suggested in previous studies [23] and confirmed in this study, the mouse Prss35 gene is selectively expressed in the ovary. Although Prss23 mRNA was detected in various tissues, the reproductive organs, including the ovary and testis, exhibited high levels of Prss23 gene expression. These data suggest that Prss35 is critical for ovarian function, whereas Prss23 plays roles in both male and female reproduction.
In the mouse ovary, increased levels of Prss35 mRNA expression were noted around the time of follicle rupture and throughout the development of the corpus luteum. The induction of Prss35 mRNA expression 8 h post-hCG was blocked by the administration of the HSD3B inhibitor TRL. When ovarian progesterone levels were reduced by 93% (240 mg/kg BW of TRL), there was a corresponding 43% reduction in Prss35 gene expression. The effects of TRL on Prss35 mRNA levels were reversed when the animals received a simultaneous injection of the progestin R5020. These results demonstrate that the induction of Prss35 gene expression prior to ovulation is, in part, regulated by progesterone. The residual Prss35 gene expression in the absence of steroid synthesis probably reflects theca cell-associated expression. Progesterone-independent Prss35 expression in the ovulatory follicle may also be due to its regulation via gonadotropins or other intra-ovarian mediators (e.g., PGs). Recently, it has been demonstrated that ADAMTS1 is a progesterone-regulated protease and a critical mediator of ovulation and cumulus matrix expansion [9, 10, 13]. In the present study, the ADAMTS1 expression was also inhibited by TRL treatment and subsequently restored by progestin (R5020) replacement (data not shown), which validates the steroid ablation/replacement protocol.
Prss35 mRNA expression localized to distinct regions of the ovary dependent upon the stage of the cycle analyzed. Initially, Prss35 gene expression was restricted to the theca cells of pre-antral follicles. However, following the development of pre-ovulatory and ovulatory follicles, both the theca and granulosa cell populations expressed Prss35 mRNA. Following ovulation, the cellular constituents of the corpus luteum expressed high levels of Prss35 mRNA. This temporal and spatial regulation of Prss35 gene expression indicates that it may be involved in the tissue remodeling that occurs during each reproductive cycle in the mouse ovary.
A unique spatiotemporal pattern of ovarian expression was also noted for the related Prss23 gene. The level of mouse Prss23 mRNA expression was highest during the early stages of follicular growth (i.e., in ovaries from immature and eCG-primed mice) but was low 4 h after ovulatory stimulus and in the postovulatory stage of the stimulated estrus cycle. In situ hybridization revealed that the mouse Prss23 gene was primarily expressed in the granulosa cells of the secondary/early antral follicles. In fact, Prss23 mRNA was not detectable in the primary follicles or in the cells that form the corpus luteum. Prss23 mRNA was detectable in the antral follicles, although these expression levels were low relative to those in the secondary/early antral follicles. This restricted pattern of Prss23 gene expression in the mouse ovary suggests that it may play an important role in the transition of pre-antral follicles to antral follicles, perhaps by allowing ECM remodeling and/or regulating growth factor availability [16].
Ser proteases are characterized by a common molecular mechanism that involves three essential catalytic residues: His, Asp, and Ser [1, 34–37]. In the catalytic process (depicted in Fig. 11), the His residue extracts the hydroxyl hydrogen from the Ser residue, which is facilitated by the polarizing effect of an adjacent Asp residue. The nucleophilic oxygen of the Ser hydroxyl group then attacks the carbonyl carbon atom of the substrate's susceptible peptide bond. Covalent bonding between the Ser residue and the substrate yields the tetrahedral intermediate, which is stabilized by an oxyanion hole formed by two amide hydrogens from Ser and neighboring Gly residues. The mouse PRSS35 protein possesses the conserved His and Asp residues (positions 169 and 272, respectively), as well as a unique domain with a Thr residue (position 342) instead of the canonical Ser active site. In contrast, the amino acid sequence of mouse PRSS23 contains the catalytic His, Asp, and Ser residues (positions 174, 245, and 315, respectively) that are conserved among Ser proteases. Thr and Ser are polar amino acids and both contain a hydroxyl group as the side chain. The difference between these amino acids is that Thr contains a methyl group instead of the hydrogen attached to the ß carbon atom of the side chain. The highly conserved Gly residue that is adjacent to the catalytic Ser residue and serves to stabilize the oxyanion hole is also present in the PRSS35 protein. In addition, PRSS35 also possesses an Asp residue in close proximity to the putative catalytic amino acid. This orientation is highly conserved within the trypsin and trypsin-related groups of Ser proteases [1, 34]. Based on these findings, we propose that PRSS35 possesses a unique catalytic triad that utilizes a Thr in place of the Ser residue that generally defines this family of proteases. In this scenario, the hydroxyl group of the canonical Ser is replaced by the hydroxyl group of Thr (Fig. 11). This Ser to Thr substitution may play an important role in determining a unique pattern of substrate specificity for PRSS35. Alternatively, this substitution may yield an inactive protease that binds to specific substrates, thus protecting them from proteolysis. Recombinant PRSS35 and PRSS23 proteins will be needed to determine their proteolytic activities.
|
Analyses of mammalian PRSS35 and PRSS23 protein sequences reveal that they share a high level of homology. In particular, the putative catalytic residues and their associated peripheral sequences, which were originally identified in the mouse PRSS35 and PRSS23 proteins, are completely conserved in the corresponding orthologs. As determined by phylogenetic analysis of the full-length amino acid sequences, the PRSS35 and PRSS23 proteins are highly conserved among closely related species (i.e., in primates and rodents). Furthermore, in addition to being located on syntenic chromosomes, the genomic organization and exon/intron sizes of the PRSS35 and PRSS23 genes are almost identical in the mouse and human genomes. The sequenced genomes of bacterial (e.g., Escherichia coli), plant (e.g., Oryza sativa, Arabidopsis thaliana), unicellular eukaryote (e.g., Saccharomyces cerevisiae), and invertebrate (e.g., Drosophila melanogaster, Caenorhabditis elegans) species lack both Prss35 and Prss23 orthologs. These results suggest that both proteases arose and have been maintained through the course of vertebrate evolution, perhaps because they play critical roles in the physiology of vertebrate reproduction.
ACKNOWLEDGMENTS
We thank Wilma Perez, Griselda Irusta, Kunie Mah, Barbra Mason, and Dr. David Hess for technical support. We are also grateful to the members of the groups of Dr. Richard Stouffer, Dr. Robert Brenner, and Dr. Michael Conn for helpful suggestions and comments, and to Dr. Richard Stouffer for his critical review of this paper. We also thank the ONPRC/U54 Molecular and Cellular Biology, Imaging and Morphology, and Endocrine Services Cores for expert technical assistance.
FOOTNOTES
1Supported by NICHD HD42000, NICHD U54-18185, and NCRR RR00163. The following sequences have been submitted to Genbank: mouse putative serine protease 35 mRNA, complete cDNA: DQ223037; rhesus macaque putative serine protease 35 mRNA: DQ223038; and rhesus macaque putative serine protease 23 mRNA: DQ223039. 
Correspondence: 2 Jon D. Hennebold, Division of Reproductive Sciences, Oregon National Primate Research Center, Oregon Health & Science University West Campus, 505 NW 185th Ave., Beaverton, OR 97006. FAX: 503 690 5563; e-mail: henneboj{at}ohsu.edu
Received: 9 March 2006.
First decision: 31 March 2006.
Accepted: 21 July 2006.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  P. Wahlberg, A. Nylander, N. Ahlskog, K. Liu, and T. Ny Expression and Localization of the Serine Proteases High-Temperature Requirement Factor A1, Serine Protease 23, and Serine Protease 35 in the Mouse Ovary Endocrinology, October 1, 2008; 149(10): 5070 - 5077. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. L Espey, R. A Garcia, H. Kondo, B. Ishizuka, S. Yoshioka, S. Fujii, S. Hampton, and J. S Richards Expression of paralogs of cytochrome P45021a1 pseudogene (Cyp21a1-ps) and endogenous retrovirus SC1 (SC1) in the rat ovary during the ovulatory process J. Endocrinol., July 1, 2008; 198(1): 231 - 241. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. D. Gallardo, G. B. John, L. Shirley, C. M. Contreras, E. A. Akbay, J. M. Haynie, S. E. Ward, M. J. Shidler, and D. H. Castrillon Genomewide Discovery and Classification of Candidate Ovarian Fertility Genes in the Mouse Genetics, September 1, 2007; 177(1): 179 - 194. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||
