| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Molecular and Cellular Biology,3 Baylor College of Medicine, Houston, Texas 77030
Department of Biology,4 The University of Texas at San Antonio, San Antonio, Texas 78249
Department of Pharmacology and Therapeutics,5 University of South Florida and Shriners Hospital for Children,Tampa, Florida 33612
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
isintegrin nd 
etalloproteinase with 
hrombo
pondin-like repeats-1) is expressed and hormonally regulated in the ovary by LH and the progesterone receptor. To determine whether other family members might be expressed and regulated in the rodent ovary, those closely related to ADAMTS1 (ADAMTS4 and ADAMTS5) were analyzed in the mouse ovary by reverse transcription-polymerase chain reaction as well as by Western blot, immunohistochemical, and immunocytochemical analyses using highly specific antibodies. Prior to ovulation, ADAMTS4 and ADAMTS5 were coexpressed in granulosa cells of most follicles, whereas ADAMTS5 was also present in granulosa cells of atretic follicles. Following ovulation, ADAMTS1 and ADAMTS4 (but not ADAMTS5) were expressed in multiple cell types, including those within the highly vascular ovulation cone that marks the site of follicle rupture, endothelial cells of newly forming corpora lutea, and cumulus cells within the ovulated cumulus cell-oocyte complex (COC). Versican, a substrate for ADAMTS1 and ADAMTS4, colocalized with these proteases and hylauronan on the cumulus cell surface. To further characterize induction of these proteases and associated molecules, COCs and granulosa cells were isolated from preovulatory follicles and treated with FSH. In expanded COCs and differentiated granulosa cells, FSH induced expression of ADAMTS4 and versican message and protein, whereas increased levels of ADAMTS1 protein was observed in the media of granulosa cells where it was stabilized by heparin in this in vitro system. These studies provide the first evidence that ADAMTS1, ADAMTS4, and ADAMTS5 are expressed in spatiotemporal patterns that suggest distinct as well as some overlapping functions that relate to the broad expression pattern of versican in granulosa cells of small follicles, expanded COCs, and endothelial cells of the mouse ovary.
ADAMTS4, ADAMTS1, ADAMTS5, cumulus cells, cumulus oocyte complex, follicular development, granulosa cells, ovary, ovulation, versican
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Ovulation is also dependent on the formation by the cumulus-oocyte complex (COC) of a specialized hyaluronan-rich matrix surrounding the oocyte that consists of components normally associated with sites of inflammation [6, 7]. Hyaluronan provides the polymeric structural backbone of this ECM, whereas specific hyaluronan binding factors provide the structural cross-links to stabilize the matrix. These factors include TSG-6 (tumor necrosis factor alpha [TNF
-stimulated gene-6; also known as TNF-induced protein 6, or TNFIP6]) [8–13], the serum-derived factor II [14], and versican (chondroitin sulfate proteoglycan 2; CSPG2) [15–17]. Mice null for TSG-6 (Tnfaip6) [13], factors regulating the production of TSG-6 (cyclooxygenase-2 [COX-2]; ptgs2; and the PGE receptor subtype EP2, or Ep2), [10, 11] and the serum component II [14] all present the same anovulatory phenotype, indicating that these factors are essential for proper matrix formation, stability, or both. Mice null for versican are embryonically lethal, precluding detection of an ovarian phenotype [18]. However, versican is markedly induced in granulosa cells of preovulatory follicles by LH and is present in high amounts in the ECM surrounding the COC [16]. Versican is a substrate for metallo-matrix proteins (MMPs) as well as a subset of proteases belonging to the ADAMTS ( isintegrin nd etalloproteinase with hrombopondin-like repeats) family that includes ADAMTS1, ADAMTS4, and ADAMTS5 [19].
ADAMTS1, the original member of this novel family of proteinases, was first cloned by differential display reverse transcription-polymerase chain reaction (RT-PCR) from a colon carcinoma cell line [20] and subsequently from ovaries of hormonally primed immature rats [21]. Since then, ADAMTS1 has been shown to be expressed in many different tissues [22, 23], and 18 additional members of this family have been identified [24]. ADAMTS4 and ADAMTS5 are structurally and functionally related to ADAMTS1 (Fig. 1; [25–30]). Each has a signal peptide, a prodomain that is cleaved by furin at the time of secretion, a metalloproteinase domain that selectively proteolyzes proteoglycans of the hyalectan family, a disintegrin domain with the potential to bind integrins, and a carboxy terminal region with a number of thrombospondin (TSP)-like repeats. Proteolytic cleavage at the carboxy terminus of each ADAMTS generates smaller peptides. The specific proteases that generate these variant forms are not yet completely known [25]. Based on their similar enzymatic activities these ADAMTSs comprise a subfamily now designated as aggrecanases. ADAMTS4 (originally called aggrecanase-1) has been associated with cartilage remodeling and inflammation [31]. ADAMTS5 (originally given the trivial name implantin), based on its expression in uterine tissue during implantation [32], has been shown to be expressed at different levels in many different tissues [33]; most recently in synovial cells, suggesting that it as well as ADAMTS4 contribute to cartilage remodeling [34, 35]. However, the expression patterns and functional roles of ADAMTS4 and ADAMTS5 during development and organogenesis, including that of the ovary, remain unclear.
|
Biochemical and in vitro assays have implicated a role for ADAMTS1 in inflammation [20, 36] and as an antiangiogenic factor [22], whereas the phenotype of mice null for ADAMTS1 indicate that it affects organogenesis, particularly the development of the reproductive tract, including the ovary [37, 38]. Of interest, ADAMTS1, ADAMTS4, and ADAMTS5 preferentially cleave one or more of the high-molecular-weight hyalectan proteoglycans, aggrecan, brevican, or versican [19, 39, 40], a process that for ADAMTS4 requires the TSP-1 motif and the presence of chondroitin sulfate linkages on the core protein [29, 30, 40]. Whereas aggrecan is preferentially expressed in cartilage (10% of its dry weight) and brevican is highest in brain, versican is more widely expressed and has been shown to be essential for heart formation [18]. In the ovary, versican is expressed in small follicles where it localizes to the surface of granulosa cells. LH dramatically increases the expression of the V0, V1, and V3 variants of versican in granulosa cells of preovulatory follicles [16] where it localizes to the ECM that surrounds the cumulus cells in a pattern similar to that observed for ADAMTS1 [17]. Prior to ovulation and within the ovulated COCs, the V1 variant of versican is cleaved as documented by the appearance of a 
70 kDa fragment [17] that has been generated in vitro by the actions of ADAMTS1 (and related family members) [19]. These recent observations indicate that ADAMTS1 (or related family members) act in vivo as well as in vitro to cleave this proteoglycan. Although the functional significance of versican cleavage is not known, in the COCs it may be critical for the functional integrity, structural integrity (or both) of the hyaluronan matrix, as well as movement of the cumulus cells within the expanding matrix. In support of this, ADAMTS1 expression is markedly reduced in follicles of progesterone receptor null (PRKO) mice that fail to ovulate in response to the LH surge [41]. In these same follicles, the amount of the 70 kDa fragment of versican V1 was decreased but not abolished [17]. That versican continues to be cleaved even when ADAMTS1 is low led us to hypothesize that other members of the ADAMTS family might be expressed and functional in the ovulating follicles.
The studies described herein were undertaken to determine the temporal and spatial expression patterns of the ADAMTS1 related family members ADAMTS4 and ADAMTS5. Using semiquantitative RT-PCR analyses and highly specific antibodies, the expression patterns of ADAMTS1, ADAMTS4, and ADAMTS5 were examined in hormonally primed mice. Our results document that unlike ADAMTS1, ADAMTS4 and ADAMTS5 are clearly detected in granulosa cells of small follicles. ADAMTS1 and ADAMTS4 but not ADAMTS5 were expressed in significant amounts in cumulus cells of ovulated COCs collected from the oviducts, ovarian cells that comprise the ovulation cone at the site of follicle rupture, and endothelial cells of the newly forming corpus luteum. Granulosa cells in culture secrete ADAMTS1 and ADAMTS4 in association with the 70 kDa V1 form of versican, indicating that they are functionally active in this context.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Gestyl (equine chorionic gonadotropin; eCG) was purchased from Professional Compounding Center of American (Houston, TX). Pregnyl (hCG) was purchased from Organon Special Chemicals (West Orange, NJ), and FSH (oFSH-16) was a gift from the National Hormone and Pituitary Program (Rockville, MD). Forskolin was purchased from Calbiochem (San Diego, CA). Dulbecco modified Eagle medium (DMEM)-F12 medium, penicillin-streptomycin, TRIzol, and Klenow were from Invitrogen (Carlsbad, CA). Fetal bovine serum (FBS) was obtained from Hyclone (Logan, UT). Oligonucleotide poly(dT) was purchased from Amersham Pharmacia Biotech (Newark, NJ); and avian myeloblastoma virus (AMV) reverse transcriptase, Taq polymerase, and thermo-cycle buffer were from Promega (Madison, WI). Radiolabeled [32P]dCTP was purchased form ICN (Los Angeles, CA). Oligonucleotide primers for RT-PCR reactions were from Sigma-Genosys (Houston, TX).
Animals and Tissue Isolation
Immature female C57BL/6 mice were purchased from Harlan Sprague-Dawley, Inc. (Indianapolis, IN). Progesterone receptor knockout (PRKO) mice were a gift from Dr. John Lydon (Department of Molecular and Cellular Biology, Baylor College of Medicine). Mice were provided food and water ad libitum, housed under a 14L:10D schedule, and treated in accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals. Protocols were approved by the Institutional Animal Care and Use Committee, Baylor College of Medicine (Houston, TX).
Ovaries Twenty-one-day-old immature female mice were either left untreated (immature) or were injected i.p. with 4 IU eCG to stimulate the growth of preovulatory follicles [17, 42]. After 48 h, some of the eCG-treated mice were injected i.p. with 5 IU hCG to initiate ovulation and luteinization [16]. Mice were killed before and at selected time intervals following eCG and hCG injection (2, 4, 8, 12, 16, 24, and 48 h). Whole ovaries were collected (3–8 per time point) and frozen immediately for subsequent RNA extraction. Total RNA was prepared using ovaries collected from mice in three separate experiments that included each hormone treatment.
Cumulus cell-oocyte complexes
Immature (Day 21 of age) mice were injected i.p. with 4 IU of eCG, killed 48 h later, and the ovaries collected. Nonexpanded COCs were obtained by manual needle puncture of the largest (preovulatory) follicles within the ovaries. COCs were separated from granulosa cells, collected by pipette, pooled, and frozen at –80°C for subsequent RNA or protein extraction. Granulosa cells that were released by needle puncture of the follicles and ovaries were pooled, collected by centrifugation, and frozen at –80°C. COCs and granulosa cells were also isolated from ovaries of eCG-primed mice exposed to hCG for 8 h. At this time the COCs have begun to expand but have not yet ovulated. Ovulated (fully expanded) COCs were collected by needle puncture of the oviducts of mice 16 h or 24 h after hCG. Whole ovaries (WO) collected from the same eCG,hCG 24-h primed mice were also frozen and stored at –80°C. Each pool of COCs and granulosa cells as well as WO were obtained from 12 to 14 mice. In other experiments, nonexpanded COCs (100) were collected from the ovaries of eCG-treated mice, pooled, and plated in one well of a Falcon 24-well plate (Becton Dickinson, Franklin Lakes, NJ) in defined medium [11] containing 1% FBS without or with forskolin (10 µM) or FSH (100 ng/ml). After 16 h the COCs were extracted for RNA (see below). Granulosa cells from the same eCG-treated ovaries were pooled and plated in the same medium with two ovarian equivalents of granulosa cells (150 000/2 ml of media) added per well in a 12-multiwell Falcon culture dish. The cells were cultured in medium alone or were stimulated with forskolin (10 µM) or FSH (50 ng/ml) for selected time intervals [43]. Each pool represents COCs or granulosa cells from 12 mice.
For analyzing the synthesis and secretion of ADAMTS proteins and versican in culture, granulosa cells were isolated from eCG-treated mice and plated on serum-coated wells (12-well plate) in serum-free defined DMEM-F12 media for 24 and 48 h without or with FSH (50 ng/ml), heparin (H3149; Sigma, St. Louis, MO; 5 and 20µg/ml), versican (5µg of crude extract of human aorta), or hyaluronic acid (H1751; Sigma; 50µg/ ml). At 24 h and 48 h, media samples (1.5 ml) were removed, centrifuged (Eppendorf) for 2 min to remove any cells and debris, and then concentrated to 60 µl using Centricon 30 filters according to the manufacturer's instructions (Millipore, Bedford, MA). The cultured cells were immediately lysed in 100 µl of boiling SDS sample buffer [43, 44]. For the Western blots, 30 µl of media and 40 µl of the cell extract were resolved SDS-PAGE using gradient gels.
Isolation of RNA From Ovarian Cells and COCs
Total RNA was isolated from whole ovaries, granulosa cells, and COCs using a Qiagen RNeasy Mini-Kit (Qiagen Sciences, Valencia, CA) according to the manufacturer's instructions. The extracted RNA was resuspended in diethyl pyrocarbonate water, quantified by absorption at 260/ 280 nm, and stored at –80°C. A "no-RT" PCR reaction using ribosomal protein L19 primers was performed on each sample to ensure that each was DNA-free.
Semiquantitative RT-PCR Analyses
The message levels for each gene were analyzed by semiquantitative RT-PCR performed as described previously [45] using specific primer pairs for ADAMTS4, ADAMTS5, ADAMTS1 [42, 46], veriscan [16], progesterone receptor [46], COX-2 [11], TSG-6 [10], syndecan-1, MMP17, and the internal control ribosomal protein L19 (RP-L19) as shown below. Briefly, total RNA (300 ng) was reverse transcribed by using oligo poly(dT) and AMV reverse transcriptase at 42°C for 75 min, followed by 95°C for 5 min. DNA products were amplified by adding [32P]dCTP, Taq polymerase, and Thermocycle buffer to the reaction mixtures at 94°C for 1 min, 60°C for 2 min, and 72°C for 2 min. The cycle number was chosen by determining the linear range of amplification for each gene. The amplified cDNA products were resolved on a 5% polyacrylamide gel, which was dried and exposed to x-ray film. The radioactive PCR products were quantified by using a Storm 860 PhosphorImager and Image Quant software v.x.x (Molecular Dynamics, Sunnyvale, CA). The expression of each gene in whole ovaries of hormone-treated mice was analyzed using RNA prepared from three separate experiments as described above. The ratio of the amplified gene of interest to L19 was determined and then used to calculate the relative fold induction (mean ± SEM) of each gene in response to hormone treatments. For COCs, the levels of message were determined using RNA prepared from three separate pools of COCs at each time point in vivo or in culture. These data are depicted as the ratio of the amplified gene to L19. The primers used for the RT-PCR reactions were as follows: Adamts1, forward 5'-CAGTACCAGACCTTGTGCAGACCTT, reverse 5'-CACACCTCACTGCTTACTGGTTTGA (20 cycles); Adamts4, forward 5'-GAGCAGTGTGCTGCCTACAA, reverse 5'-TGCTGCCGTACAAGGATATG (25 cycles); Adamts5, forward 5'-ACGGCATTATTGGCTCAAAG, reverse 5'-GGATCTGCACGATCAGGATT (22 cycles); Mmp17, forward 5'-AGGCGAAGCATTCTTTTTCA, reverse 5'-GCCGTGTGTGGTCATCATAG (32 cycles); Cspg-2 (versican), forward 5'-CTGCAGACGACATGGAGCTA reverse 5'-GTTGAGGCATGGGTTTGTTTT (25 cycles); Syndecan-1, forward 5'-GTCCTATGAACTCTACTGCTTCTGC, reverse 5'-CACAGATCTTTTGTCTAGCTGAGTG (20 cycles); Ptsg2 (COX-2,) forward, 5'-TGTACAAGCAGTGGCAAAGG, reverse 5'-GCTGTGGATCTTGCACATTG (24 cycles); Tnfaip6 (TSG-6), forward 5'-TTCCATGTCTGTGCTGCTGGATGG, reverse 5'-AGCCTGGATCATGTTCAAGGTCAAA (24 cycles); Nr3c3 (PR), forward 5'-CCCACAGGAGTTTGTCAAGCT, reverse 5'-TAACTTCAGACATCATTCCGG (26 cycles); and L19, forward 5'-CTGAAGGTCAAAGGGAATGTG, reverse 5'-GGACAGAGTCTTGATGATCTC (20 cycles). Primers for PCR amplification of the desired cDNA products/gene were designed using the Web-based program Primer3 (www.genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi) and the program Amplify 1.2. All cDNA products were sequenced to verify the authenticity of the gene products.
Western Blot
Protein samples from whole ovaries, COCs, and isolated granulosa cells were prepared either by homogenization in whole cell extract buffer (10 mM NaHPO4 pH 7.4, 1 mM Na2EDTA, 1 mM dithiothreitol, 400 mM KCl, 10% glycerol, 5 µg/ml aprotinin, 5 µg/ml leupeptin, 1 µM pepstatin, 1 mM phenylmethylsulfonyl fluoride, 5 µM NaF, and 0.5 µg/ml okadeic acid) [47, 48], urea extract buffer (6 M urea and 0.1% Triton containing protease inhibitors, EDTA, benzamidine, phenylmethylsulfonyl, and aprotinin) [17], or boiling SDS sample buffer [49] as indicated in the text and as described previously. Extracts were resolved by SDS-PAGE (12%) and transferred to Immobilon-P nylon membranes (Millipore Corp, Bedford, MA). Membranes were blocked in Tris-buffered saline and Tween-20 (TBST; 10 mM Tris pH 7.5, 150 mM NaCl, and 0.05% Tween-20) containing 3% nonfat Carnation instant milk (Nestlé Co., Solon, OH). As reported previously [17], primary antipeptide affinity-purified antibodies for ADAMTS1 p110 (pro-form) and ADAMTS1 p85 (metalloproteinase-secreted form) were diluted 1:3000 in TBST and incubated for 1 h at room temperature. Epitope-specific affinity-purified antibodies for ADAMTS4 (JSCYNH and JSVMA) [25], ADAMTS5 p55 (JSCKNG), and MMP17 (M3684; Sigma) were diluted 1:1000 (1.2 µg/ml), 1:500 (2 µg/ml), 1:500 (2 µg/ml), and 1:000 (0.8 µg/ml), respectively, in TBST and incubated overnight at 4°C. Following primary antibody, all blots were washed in TBST and incubated with secondary antibody (horseradish peroxidase-linked anti-rabbit immunoglobulin G [IgG]; 1:20 000; Amersham Biosciences) in 3% milk TBST for 1 h at room temperature. After washing in TBST, enhanced chemiluminescence detection was performed by using Pierce Super Signal according the manufacturer's specifications (Pierce) and appropriate exposure of the blots to Kodak x-ray film. Specific bands were quantified by densitometric analyses using a Molecular Dynamics (Sunnyvale, CA) Personal Densitometer.
Immunohistochemistry
Tissue and cellular localization of proteins was analyzed by immunohistochemistry on 7-µm sections of paraffin-embedded, 4% paraformaldehyde-fixed tissues as described previously [17]. Sections were rehydrated and boiled 10 min in 10 mM sodium citrate pH 6.0, then treated for 10 min with 1 µg/ml proteinase K (Sigma). Endogenous peroxidase activity was quenched by 10 min of treatment with 0.1% hydrogen peroxide followed by a PBS wash. Nonspecific antibody binding was blocked by incubation with 20% normal goat serum in PBS followed by incubation with the primary antibodies diluted in 10% normal goat serum overnight at 4°C. Primary antibodies for ADAMTS1, ADAMTS4 (JSCVMA and JSCYNH), ADAMTS5, and MMP17 (Sigma) as described above were diluted (1:500, 1:50 [20 µg/ml], 1:50 [20 µg/ml], and 1:30 [22 µg/ml], respectively). The next day, the slides were washed in PBS containing 0.025% Tween-20 and then incubated with secondary antibody (biotinylated goat anti-rabbit antiserum; Vector Laboratories, Burlingame, CA) for 1 h at room temperature. The sections were washed and incubated with streptavidin-conjugated horseradish peroxidase for 30 min, washed, and reacted with the diaminobenzidine substrate (3,3'-diaminobenzidine)-substrate solution (Vector Laboratories) for 2 min, dehydrated, and mounted in Permount.
Immunocytochemistry
Immunocytochemical approaches were used to localize ADAMTS1, ADAMTS4, versican, and hyaluronan in ovulated COCs collected from the oviducts 24 h after hCG (12 h after ovulation). Isolated COCs were immobilized on laminin-coated coverslips, fixed with 4% paraformaldehyde, washed with PBS, and incubated with various primary antibodies. Granulosa cells were isolated from eCG-primed mice and cultured on serum-coated coverslips in defined medium containing FSH (100 ng/ml) for 16 h. The granulosa cells were fixed in 4% paraformaldehyde and processed in a manner identical to that of the COCs. COCs were incubated overnight at 4°C with primary antibodies to ADAMTS1, ADAMTS4 (JSCVMA and JSCYNH), and versican (all diluted 1:00 in PBS containing 3% BSA) [50] or with biotinylated hyaluronic acid-binding protein (HABP; Seikagaku, Falmouth, MA; 0.05 µg/200 µl PBS containing 3% BSA). Antibody localization was visualized with fluorescein isothiocyanate-conjugated labeled anti-rabbit IgG and Streptavidin AlexaFluor 568 (Molecular Probes, Invitrogen, Eugene, OR). Nuclei were visualized by 4',6'-diamidino-12-phenylidole (DAPI) in Vectashield D mounting medium (Vector Laboratories).
Statistics
The RT-PCR data are represented as the mean ± SEM. Data were analyzed using GraphPad Prism Programs (analysis of variance [ANOVA] or t-test; GraphPad Prism, San Diego, CA) to determine significance. Values were considered significantly different if P < 0.05 or P < 0.01.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our first experiments were designed to determine when and where ADAMTS4 and ADAMTS5 were expressed during follicular development, ovulation, and luteinization in the mouse. As shown in Figure 2A, ADAMTS4 message was detected in total RNA prepared from whole ovaries of immature (Day 21) mice. Treatment of these mice with eCG for 48 h to stimulate the development of preovulatory follicles did not alter the level of ADAMTS4 message. However, at 2 and 4 h after treatment with hCG, ADAMTS4 mRNA levels increased markedly (P < 0.05 and P < 0.01 respectively, compared with those of eCG). ADAMTS4 message levels decreased significantly at 8 h (P < 0.01), exhibited a slight increase at 12–16 h, and decreased further at 24–48 h post-hCG when the ovulated follicles had fully luteinized. Levels of ADAMTS4 message in the luteinized ovaries were similar to those observed in ovaries of immature and eCG-primed mice. This pattern of expression differs from that of ADAMTS1 (Fig. 2A; shown schematically by the dashed line [42, 46]).
|
In other tissues, a secreted, membrane-bound form of ADAMTS4 p68 has been identified. Biochemical assays have shown that this mature form of the protein is active, but the specificity of the enzyme is broadened in the carboxy-terminal truncated p53 form [25]. These biochemical data suggest that the activity of ADAMTS4 is itself altered by one or more specific proteases. To determine whether ADAMTS4 protein was present, regulated, or cleaved in the mouse ovary, Western blots were carried out using whole cell extracts prepared from ovaries of hormonally primed immature mice and specific peptide affinity-purified antibodies that had been found to preferentially recognize the mature (p68; JSCYNH) and processed (p53; JSCVMA) forms, respectively. As shown in Figure 2B, the p68 form of ADAMTS4 is abundant and remains at a relatively constant level in ovaries of immature, eCG-treated, and eCG,hCG treated mice. In contrast, the amount of the ADAMTS4 p53 form was low in ovaries of immature, eCG-treated, and eCG,hCG 4-h treated mice but increased 6- to 8-fold at 8–48 h after hCG. When the amount of ADAMTS4 protein present in both bands was combined, the highest level in the ovary was observed in the periovulatory (12–16 h post-hCG) and postovulatory (24–48 h post-hCG) periods. That ADAMTS4 protein levels were higher than the message in postovulatory luteinized ovaries may indicate that the protein is more stable than the message at this time.
To determine whether ADAMTS4, like ADAMTS1, was hormonally regulated by progesterone receptor, RNA prepared from ovaries of PRKO mice was analyzed by RT-PCR. As shown in Figure 2C, ADAMTS4 was induced dramatically (P < 0.01) by hCG in heterozygote (PRHET) and PRKO ovaries at 4 h, as seen in Figure 2A. Although the levels of ADAMTS4 mRNA in the heterozygous mice remained unchanged at 8, 12, and 16 h, the amount of ADAMTS4 message in PRKO ovaries at 12 h after hCG was significantly reduced (P < 0.05). Thus, ADAMTS4 does not appear to be a major target of progesterone receptor.
Cell-Specific Localization of ADAMTS4 Message in the Mouse Ovary
To determine the cell specific expression and localization of ADAMTS4 protein in the mouse ovary, immunohistochemical analyses were performed using the epitope-specific antibodies as described above for the Western blots. As shown in Figure 3A (left panels), ADAMTS4 protein was localized selectively to granulosa cells of small and growing follicles in the immature mouse ovary. Following treatment with eCG, granulosa cells in preovulatory follicles as well as small follicles were immunopositive, but the intensity and distribution of the signal were altered. At higher magnification (40x) the signal in the preovulatory granulosa cells appeared to be distributed in patches on the surface of the cells rather than evenly coating the cells, which was seen at other time intervals analyzed. Following treatment with hCG to stimulate ovulation and luteinization, ADAMTS4 protein was detected evenly on the surface of granulosa cells. These data suggest that ADAMTS4 may be bound to different granulosa cell surface molecules in response to hormonal stimulation. In the hCG-treated ovaries ADAMTS4 was also observed in theca cells (Fig. 3A inset; eCG,hCG 4 h), a pattern that was confirmed by in situ hybridization (data not shown). ADAMTS4 protein remained detectable in corpora lutea and small follicles present in ovaries of eCG,hCG 48-h treated mice.
|
The most dramatic shift in the localization of ADAMTS4 during the periovulatory period was observed in ovaries collected at 16 h after hCG (Fig. 3B; eCG,hCG 16 h; red arrow). At this time, newly ovulated follicles were noted by the presence of ovulation cones (stigmae), which are characterized by a rich vascular network associated with the surface epithelium (Fig. 3C). At these sites, immunostaining of ADAMTS4 was enriched and localized to the surface of cells. At higher magnification (20x and 40x inset; white arrows), ADAMTS4 was also localized distinctly to endothelial cells of the newly forming corpora lutea. ADAMTS1 (Fig. 3B, right panels) was also localized to the cells within the ovulation cone as well as to endothelial cells of the newly forming corpora lutea (Fig. 3B). These results indicate that both ADAMTS proteases exhibit overlapping expression during this critical stage of remodeling that occurs at the surface of the ovary at the sight of rupture and during neovascularization of the corpus luteum.
MMP17 and Syndecan-1 Are Expressed in theRodent Ovary
Previous studies have indicated that ADAMTS4 p68 is cleaved at the cell surface to release the more broadly active ADAMTS4 p53. This cleavage removes the C-terminal spacer domain, a process that can be mediated by the membrane protease (MT4-MMP) MMP17 [28]. To determine whether MMP17 was present in the rodent ovary, RT-PCR analyses were performed. As shown in Figure 4A, the message for MMP17 is present in ovaries of immature and eCG-primed mice, as well as in ovaries of mice treated with hCG. No significant changes were observed in response to hormone. Western blot analyses also indicated that there were no major changes in the amount of MMP17 (data not shown). However, immunohistochemical analyses show that MMP17, like ADAMTS4 and ADAMTS1, is enriched in the ovulation cones (Fig. 4B). Previous studies have also indicated that attachment of ADAMTS4 p53 to the cell surface involves its binding to the cell surface chondroitin sulfate proteoglycan syndecan-1 [28]. To determine whether syndecan-1 was expressed in the ovary, additional RT-PCR analyses were carried out. As shown, syndecan-1 message is expressed in the mouse ovary and increased significantly in ovaries of eCG,hCG-treated mice (Fig. 4C). When isolated ovarian compartments were analyzed, regulated expression of syndecan-1 was clearly evident (see Fig. 6A).
|
|
ADAMTS5 Is Expressed in Granulosa Cells at Specific Stages of Follicular Development
RT-PCR analyses show that ADAMTS5 mRNA is expressed in the immature mouse ovary (Fig. 5A). Unlike ADAMTS1 and ADAMTS4, ADAMTS5 message did not change significantly in response to eCG or hCG during the periovulatory period (hCG 2–16 h) and after luteinization (hCG 24–48 h) (Fig. 5A). Using a peptide-specific antibody for ADAMTS5 (JSCKNG), a single band at 55 kDa protein was observed in whole cell extracts prepared from ovaries of immature, eCG and eCG,hCG-primed mice (Fig. 5B). Levels of immunoreactive ADAMTS5 were not demonstrably different at any time interval. The mature, secreted p70 form of ADAMTS5 was not observed in these samples. Immunohistochemical analyses showed that ADAMTS5 p55 protein was present in granulosa cells of ovaries of immature mice and eCG-primed mice (Fig. 5C). In the adult cycling ovary ADAMTS5 signal localized preferentially to granulosa cells of antral follicles that appeared to exhibit signs of atresia. ADAMTS5 was not observed in granulosa cells of ovulating follicles (data not shown).
|
ADAMTS4 and ADAMTS1 Message and Protein Continue to be Expressed in Ovulated COCs
Previous studies have indicated that ADAMTS1 protein is expressed by COCs before ovulation as well as in ovulated COCs collected from oviducts 16 h following hCG, or approximately 4–6 h after ovulation [17]. To determine whether ADAMTS4 and ADAMTS5 as well as ADAMTS1 were present in COCs, granulosa cells and COCs were collected from eCG,hCG primed mice either before ovulation (8 h post-hCG treatment) or following ovulation (16 and 24 h post-hCG treatment). As controls, granulosa cells and COCs were collected from preovulatory follicles of eCG-treated mice (designated 0 h; Fig. 6). Total RNA was prepared and used in semiquantitative RT-PCR analyses, or whole cell (protein) extracts were made for Western blot analyses. Regulated expression of ADAMTSs was compared against that of versican, a known substrate for ADAMTS1 and ADAMTS4 [19, 39]; and progesterone receptor, a factor important in ADAMTS1 expression in vivo [42, 46]; as well as markers known to be selectively expressed in expanded COCs (namely, COX-2, and TSG-6) [8, 10, 11, 51, 52]. As shown in Figure 6A, ADAMTS4 message was present in COCs collected from ovaries of eCG-primed mice and was induced significantly (P < 0.01) 9- to 13-fold and 14-fold in COCs collected 8, 16, and 24 h after hCG, respectively. ADAMTS4 was also increased significantly (P < 0.05) in granulosa cells isolated from ovaries at 8 h and 16 h, and in whole ovaries isolated at 24 h compared with controls (0 h, eCG). In marked contrast, ADAMTS5 message was highest in the nonexpanded COCs at 0 h and declined significantly (P < 0.01) in response to hCG. ADAMTS5 was low in granulosa cells and not altered by hormones. ADAMTS1 message, like that of ADAMTS4, was induced dramatically (P < 0.01) in COCs and granulosa cells collected 8 and 16 h after hCG. Although message levels for ADAMTS1 dropped in COCs collected at 24 h, the amount remained significantly higher (P < 0.05) than that observed in nonexpanded COCs of eCG-primed mice.
Progesterone receptor mRNA was also induced (P < 0.01) in COCs and granulosa cells in a pattern similar to that of ADAMTS1. Expression of progesterone receptor, like that of ADAMTS4, remained significantly higher in whole ovaries collected at 24 h after hCG than in granulosa cells of eCG-primed mice. Versican exhibited an expression pattern in COCs and granulosa cells similar to that of ADAMTS1 with 7- and 5-fold increases observed in COCs and granulosa cells at 8 and 16 h, respectively (P < 0.01; Fig. 6). The membrane-associated proteoglycan syndecan-1 was also induced significantly (P < 0.01) in COCs collected 8, 16, and 24 h after hCG. Known markers of expanded and ovulated COCs; namely, COX-2 and TSG-6, were induced robustly (69-fold and 44-fold, respectively), and selectively compared against nonexpanded control samples (data not shown). Thus, expression of ADAMTS4 as well as ADAMTS1 (but not ADAMTS5) message is increased in the expanded COCs before (8 h post-hCG) and after (16 and 24 h post-hCG) ovulation.
Western blot analyses confirmed the induced expression of ADAMTS4 and ADAMTS1 protein in ovulated COCs (Fig. 6B). As shown, the p68 and p53 forms of ADAMTS4 were present in nonovulated COCs collected from eCG-primed mice (designated 0 h) and increased (2-fold and 5-fold, respectively) in ovulated COCs collected at 16 h. When these same samples were analyzed for ADAMTS1, the ovulated COCs contained 20-fold and 74-fold more of the intracellular (p110) and secreted (p85) forms of ADAMTS1, respectively, than the nonovulated COCs. Thus, levels of both ADAMTS4 and ADAMTS1 protein are elevated in the ovulated COCs compared with nonovulated and nonexpanded COCs of eCG-primed mice. ADAMTS5 protein was not detectable in these samples (data not shown).
To compare the localization of ADAMTS4 with that of ADAMTS1 within the ovulated COCs, immunocytochemical analyses were performed. Intense staining of both ADAMTS1 p85 and ADAMTS4 was observed to the cumulus cells within several clusters of ovulated COCs (Fig. 6C, upper panels). The most intense staining was observed at (or on) the surface of the cumulus cells as shown at higher magnification and by DAPI staining (see insets). However, whereas ADAMTS4 appeared evenly distributed at the cell surface, ADAMTS1 exhibited intense staining to specific foci. Furthermore, punctate staining of ADAMTS1, but not ADAMTS4, was observed within the focal plane of the matrix (Fig. 6C; lower panels), suggesting that ADAMTS1 is present not only at the cell surface but also in more limited amounts within the matrix. When the localization of versican was analyzed, this proteoglycan localized not only to the ECM but also to the surface of cumulus cells (Fig. 6D). Versican was clearly not present in the nuclei of these cells as shown by DAPI staining (insert). Because versican is a hyaluronan-binding protein, the localization of hyaluronan was also analyzed (Fig. 6, E and F). Versican and hyaluronan were tightly colocalized to the matrix as well as to the surface of the cumulus cells. Thus, the cumulus cells are not only the site of synthesis of these factors, but they become encased within these matrix molecules.
ADAMTS4, PR, Versican, and Syndecan-1 Are Induced in COCs in Culture
To determine whether COC-expressed genes were regulated during hormone-induced expansion in vitro, nonexpanded COCs collected from ovaries of eCG-primed rats were cultured in defined medium containing 1% fetal bovine serum and either FSH (100 ng/ml) (Fig. 7) or forskolin (10 µM; data not shown). The expression patterns of ADAMTS4 and ADAMTS5 in the cultured COCs were similar to those observed in vivo (Fig. 6A). Whereas ADAMTS4 message increased 3.5-fold at 16 h after FSH treatment, ADAMTS5 expression was reduced by 80% (Fig. 7A). Expression of versican and syndecan-1 were increased in the COCs (6.5- and 2.8-fold, respectively) and granulosa cells (3- and 1.5-fold, respectively) cultured in the presence of FSH. Conversely, ADAMTS1 message was not induced by FSH in these same samples. Rather, levels of ADAMTS1 message levels were low in the expanded COCs. Furthermore, FSH did not modulate the expression of ADAMTS1 message in granulosa cells cultured with FSH for 16 h. Progesterone receptor, a putative regulator of ADAMTS1, increased 6-fold in COCs and 2-fold in granulosa cells (Fig. 7A). The known markers of COC expansion, COX-2 and TSG-6, were induced robustly (21-fold and 18-fold, respectively; data not shown) at 16 h in the expanded COCs in a manner similar to that observed in the in vivo-expanded COCs [10, 11].
|
Western blot analyses confirmed the expression of ADAMTS4 and ADAMTS1 protein in COCs and granulosa cells cultured in medium containing 1% serum and FSH (Fig. 7B). As shown, the p68 form of ADAMTS4 was present in similar amounts in nonexpanded COCs cultured overnight in the presence of FSH for 16 h. In contrast, the p53 form of ADAMTS4 was present in low abundance, nonexpanded COCs, but increased 10-fold in response to FSH. In granulosa cells, ADAMTS4 p68 was also present at relatively constant levels, whereas the amount of ADAMTS4 p53 increased 40-fold in response to FSH (Fig. 7B). Levels of ADAMTS1 p110 and p85 were low in the control (nonexpanded) and hormone-treated (expanded) COCs in culture compared with the marked (10-fold) increase in ADAMTS1 observed in granulosa cells in response to FSH. Furthermore, the FSH-induced increase in the amount of ADAMTS1 in granulosa cells at 16 h occurred despite the lack of any detectable change in message at this same time interval (Fig. 7A). These observations suggest that ADAMTS1 protein may be more stable in the differentiated cells. ADAMTS5 protein was not detectable in these samples (data not shown). Versican could not be determined in these cultures due to nonspecific interactions with serum proteins adherent to the cells.
Immunocytochemical analyses revealed that ADAMTS4 and ADAMTS1 were localized to discrete regions within the granulosa cells (Fig. 7, C and D). Specifically, ADAMTS4 p53 and p68 exhibited punctate staining suggestive of secretion vesicles. ADAMTS4 p53 (as determined by JCVMA antibody) also appeared to be associated with cytoskeletal structures and the cell surface. ADAMTS1 p110 was specifically localized to a distinct region of the cells, whereas ADAMTS1 p85 exhibited a more punctate staining pattern. Neither appeared to be associated with the cell surface, most likely because most is bound to factors present in the medium (Figs. 7B and 8).
|
Medium from Cultured Granulosa Cells Contains Versican, ADAMTS1, and ADAMTS4
The thrombospondin-1 (TSP-1)-like repeats of ADAMTS1 and ADAMTS4 are similar to those in other ECM molecules such as thrombospondins -1 and -2, and pleiotrophin [22, 53]. These domains appear to be critical for the binding of these factors to matrix molecules containing chondroitin sulfate, heparan sulfate moieties (or both), such as those present in the cell surface receptor, syndecan-1 [54] as well as versican [50, 55, 56]. That ADAMTS1 and ADAMTS4 protein is highly concentrated in the matrix of ovulated COCs (Fig. 6B) and in media samples of granulosa cells cultured in the presence of serum (Fig. 7B) suggested that these proteases might be stabilized by glycosaminoglycan-containing molecules. Furthermore, heparin has been shown to be a potent inhibitor of ADAMTS1 activity [57], most likely because of its potential to bind the TSP domain of this protease. To address the possibilities that heparin might alter the function or stability of ADAMTS1 and ADAMTS4 in cultured granulosa cells and thereby alter versican V1 cleavage, granulosa cells were isolated from ovaries of eCG-primed mice and cultured in serum-free DMEM-F12 medium with or without FSH/testosterone (T) for 48 h.
When granulosa cells were cultured in serum-free medium alone for 48 h, the amount of the ADAMTS1 (p85) secreted into the medium was low, whereas ADAMTS4 (p68) was clearly detected and not altered by FSH (Fig. 8A). In contrast, the amount of ADAMTS1 (p85) increased dramatically in the media of cells cultured with 5 and 20 µg/ml of heparin alone (400-fold) or with FSH for 48 h (480-fold), whereas the increase of ADAMTS4 (p53) in the media of cells cultured with heparin alone (2-fold) or in the presence of FSH and 5 and 20 µg heparin (3-fold and 20-fold, respectively) was more modest. In the same media samples of these cultures, little or no versican V1 70 kDa was observed in the absence of FSH (Fig. 8A). However, in the presence of FSH, a 70 kDa immunoreactive band was clearly evident. Furthermore, the 70 kDa versican V1 band observed in cells treated with FSH exhibited an identical migration pattern to that of versican extracted from human aorta, and the migration of 70 kDa versican from human aorta and was not altered by heparin in the absence or presence of cells (data not shown). Collectively, these data show that FSH increases the synthesis/cleavage of versican V1 to the 70 kDa fragment, and that ADAMTS4 but not ADAMTS1 may be the more active protease in this serum-free culture system. Conversely, when heparin, a potent inhibitor of ADAMTS activity was present, the cleaved product of versican V1 was not observed, but elevated levels of ADAMTS1 (p85) and ADAMTS4 (p53) in the media were observed, which is consistent with the ability of heparin and heparin-containing matrix molecules to bind as well as to inhibit ADAMTS1 and ADAMTS4 activity.
When the cellular levels of ADAMTS1 and ADAMTS4 were analyzed, ADAMTS1 p110 decreased markedly (70%–80%) in the presence of heparin, and little or no ADAMTS1 p85 was observed (Fig. 8B). In contrast, levels of ADAMTS4 (p68 and p53) remained relatively constant (Fig. 8B). RT-PCR analyses showed that expression of ADAMTS4 and ADAMTS1 message in granulosa cells cultured with FSH in the presence of heparin (5 and 20 µg/ ml media) was identical to that observed with FSH alone (data not shown). Thus, the presence of heparin did not alter granulosa cell function but did alter the amount (stabilization) of ADAMTS1 and ADAMTS4 in the media.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Specifically, ADAMTS4 and ADAMTS5 were expressed in small preantral and antral follicles of immature mice and eCG-treated mice, a pattern that closely mirrors that of versican [16]. The coexpression of versican, ADAMTS4, and ADAMTS5 in granulosa cells of small, growing follicles indicates that these proteases may have overlapping functions such as regulating the functional activities of versican and other matrix molecules at this stage of development. Versican is a complex and multifunctional molecule that has been shown recently to alter cell proliferation, migration, and differentiation, depending on what isoform or region of the molecule is present and active [55]. Versican V1 is a potent stimulator of neuronal differentiation, whereas the G3 (carboxy-terminal) domain of V1 is less effective [55]. In contrast, versican G3 promotes tumor growth and angiogenesis [56]. Because the granulosa cell layer of small follicles is avascular, and because the granulosa cells at this stage are not yet highly differentiated, the role of versican or cleaved forms of versican in these small follicles remain to be determined. At other stages of follicular growth ADAMTS4 and ADAMTS5 appear to have disparate functions. Specifically, immunohistochemical analyses show that ADAMTS5 is localized to atretic as well as healthy follicles, and although ADAMTS5 is expressed in nonovulated COCs isolated from preovulatory follicles, its expression decreased in response to hCG. Conversely, ADAMTS4 appeared to be present in all follicles and is increased in COCs in response to hCG. These results extend and confirm the patterns observed in bovine follicles [23], indicating that expression of ADAMTS5, but not ADAMTS4, decreases during preovulatory follicles development. Based on these expression profiles, it appears that ADAMTS4 and ADAMTS5 have different roles at early stages of follicular growth as well as during ovulation.
In the ovulating follicle, ADAMTS4 message is increased rapidly and is associated with an LH-induced increase in the processing of ADAMTS4 p68 to its more active p53 form, suggesting that ADAMTS4 may exert specific functions in ovulation. Prior to ovulation, ADAMTS4 was selectively induced in theca cells within 2–4 h after administration of hCG. Specific induction of ADAMTS4 (but not ADAMTS1 [42, 46]) in these cells and at this stage of follicular growth suggests that ADAMTS4 may participate in the proteolytic cascade, by which components of the basal lamina are degraded before ovulation and luteinization. The cell-specific and regulated expression of ADAMTS4 as well as ADAMTS1 in the mouse differs from that observed in bovine [23] and equine [58] follicles, suggesting that regulated expression of these proteases may also exhibit some species specificity. Because versican is localized to the theca cell layer as well as the antral cavity of mouse [16] and bovine [15] preovulatory follicles, the increased expression of message and processing of ADAMTS4 protein in granulosa cells and theca cells of periovulatory follicles in the mouse could contribute to the cleavage of versican or other substrates (or both) at both sites before ovulation. The molecular basis for the highly reproducible decline in ADAMTS4 mRNA at 8 h after hCG is not known but suggests that specific signaling cascades that affect the transcription of this gene are being altered at this time. Because ADAMTS4 is not a major target of progesterone receptor in the murine [42, 46] or porcine [59] ovary, other factors must be involved.
The pronounced expression of ADAMTS4, MMP17, and ADAMTS1 in cells that comprise ovulation cone, a prominent structure that forms at the surface of the ovary after the COC has been extruded, was striking. At this site, ADAMTS4 may be more actively cleaved, allowing it to bind syndecan-1 or other matrix molecules. ADAMTS4, along with ADAMTS1, may function to modify versican (or other substrates) to facilitate ovulation. Following ovulation and the release of the COCs, these proteases may act to permit or restrict cell migration to facilitate repair of the surface epithelium. As noted, the ovulation cone is rich in blood vessels, and therefore the regulation of angiogenesis at this site may be mediated in part by the interplay of versican with ADAMTS4 and ADAMTS1, the latter of which has been previously implicated as potent antiangiogenic factors [22].
Equally impressive is the marked expression of ADAMTS4 along with ADAMTS1 and versican in COCs collected from ovaries 8 h after exposure to hCG as well as in ovulated COCs collected from the oviduct at 16 h post-hCG. These data indicate that these factors affect the dynamic events that contribute to the formation of the COC matrix within the follicle and that these factors continue to be expressed and functional in ovulated COCs. During the formation and remodeling of the COC matrix, versican, ADAMTS1, and ADAMTS4 may act in concert to permit cumulus cell migration away from the oocyte, a highly dynamic process as shown by time-lapse photography [11]. Once expansion is complete, the high concentration of these factors in the matrix may play a role in the release of the COC from the surface of the ovary as well as in the collective movement and function of the COCs as they are transported down the oviduct. The immunocytochemical data provide evidence that versican, hyaluronan, and the ADAMTS proteases colocalize at (or on) the cumulus cell surface, suggesting that ADAMTS1 and ADAMTS4 may bind specific cell surface matrix molecules. Some potential binding sites are CD44, a hyaluronan binding factor that can also bind versican [55]; and syndecan-1 [54], a proteoglycan that binds ADAMTS4 [28]. Because the matrix is not undergoing rampant degradation even 8 h following ovulation (or 24 h after hCG), whatever proteolytic activity is going on within the COC matrix is tightly controlled, allowing the matrix to provide a protective shield around the oocyte. As observed in culture, the presence of cell surface proteoglycans containing heparin or chondroitin sulfate moieties may not only provide attachment sites to stabilize these proteases, but may also render them less active or inactive. That versican is tightly colocalized with hyaluronan indicates that it is localized with other hyaluronan-binding factors (i.e., II; data not shown) and Ptx3 [60, 61], and that interactions with these factors may exert additional stringency on versican cleavage by ADAMTS proteases.
When COC expansion was analyzed in culture, the expression of ADAMTS4, versican, and syndecan-1 as well as progesterone receptor, COX-2, and TSG-6 transcripts was induced in response to FSH or forskolin. These results indicate that the induction of these genes in cultured COCs mimics that observed COCs expanded in vivo following exposure to LH. That progesterone receptor was induced in the COCs in vivo as well as in COCs cultured in the presence of FSH or forskolin indicates that progesterone receptor may affect cumulus cell function in the mouse as it does in porcine COCs [59, 62]. However, expression of ADAMTS1 message was low in control and FSH-expanded COCs in culture. These data indicate that the COC culture system mimics some but perhaps not all events that occur in vivo, and that factors in addition to cAMP (FSH) and progesterone receptor affect expression of the ADAMTS1 gene in these murine cumulus cells. What these factors are remain to be determined.
Even when granulosa cells were cultured with FSH/T, the amount of ADAMTS1 message at 16 h did not differ markedly from that at 0 h, as we reported previously [46]. In marked contrast, the amount of ADAMTS1 protein secreted by granulosa cells into the media (containing 1% FBS) was increased strikingly. These results indicate that ADAMTS1 protein is synthesized, secreted, and stabilized by one or more factors either produced by granulosa cells or present in serum. This was further verified by the marked increase in ADAMTS1 protein observed in serum-free medium containing heparin, a heparan sulfate molecule. Heparin- and heparan sulfate-containing molecules are extremely important because they can bind and therefore alter the activity of many ECM-associated factors including ADAMTS1 [54, 57]. Heparin, for example, has been shown to exert anti-inflammatory and other regulatory effects by binding various growth factors [63]. Factors produced by granulosa cells that might act like heparin include versican [16], which has chondroitin sulfate side chains as well as the heparan sulfate proteoglycans, of which there are many, including the cell surface molecule syndecan-1 [54]. Therefore, the binding of ADAMTS1 and ADAMTS4 to these molecules likely accounts for the disparity in low expression of message compared with high levels of protein observed most dramatically in the cell culture medium. The binding of these proteases to matrix molecules either produced by cells or present in serum makes it difficult to analyze the functional activity of these ADAMTSs. That putative matrix molecules stabilize ADAMTS1 and ADAMTS4 also indicates that they are rapidly degraded unless they are bound to a matrix. Understanding the factors involved in this degradation process will be important for determining the specific functions of these ADAMTSs in vivo.
In summary, these studies provide the first evidence that ADAMTS4 and ADAMTS5 protein as well as message are expressed in different ovarian cells and in follicles at different stages of development, suggesting that these protease play multiple roles in multiple cell types and matrices in the mouse ovary. Most specifically, the results document that ADAMTS4 as well as ADAMTS1 are highly expressed in COCs before and after ovulation, suggesting that they may exert specific effects in the formation, stabilization, function (or a combination of these) of the matrix that surrounds and protects the oocyte. The elevated expression of ADAMTS4 along with ADAMTS1 at the site of rupture at the ovarian surface suggests that these proteases may also act as potential regulators of angiogenesis. Whether or not versican is a critical target for ADAMTS4 in each process and whether versican V1 cleavage by ADAMTS4, ADAMTS1, or both is essential in these processes remains to be determined. Of interest, ADAMTS4 and ADAMTS1 exhibit specific but also overlapping expression patterns in these regions of the ovary. Based on the abnormal function of ovaries in mice null for ADAMTS1, it appears that ADAMTS4 cannot substitute for all actions of ADAMTS1. The nonredundant functions of these two proteases may be dependent in part on where these proteases dock in the matrix or on cell surfaces. Conversely, ADAMTS1 is unlikely to substitute for ADAMTS4 in small follicles because the level of expression of ADAMTS1 in these small follicles is low or negligible. However, in small follicles, ADAMTS5, as well as ADAMTS9 [64] [23] and ADAMTS19 [65], might be redundant in granulosa cells, whereas ADAMTS9 and ADAMTS4 or ADAMT1 (or both ADAMTS4 or ADAMT1) may coshare functions in theca cells depending on the species [23, 58]. Finally, although versican has been identified as one substrate for ADAMTS1, ADAMTS4, and ADAMTS5, it is clear that the making and remodeling of ECMs that permit follicular growth and ovulation is complex, highly regulated, and mediated by a diverse set of factors.
|
FOOTNOTES |
|---|
2 Correspondence: JoAnne S. Richards, Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, TX 77030. FAX: 713 790 1275; joanner{at}bcm.tmc.edu
Received: 17 November 2004.
First decision: 13 December 2004.
Accepted: 17 January 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. A.H. Knauff, L. Franke, M. A. van Es, L. H. van den Berg, Y. T. van der Schouw, J. S.E. Laven, C. B. Lambalk, A. Hoek, A. J. Goverde, S. Christin-Maitre, et al. Genome-wide association study in premature ovarian failure patients suggests ADAMTS19 as a possible candidate gene Hum. Reprod., June 9, 2009; (2009) dep197v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-C. Su, E. A. Mendoza, H.-i. Kwak, and K. J. Bayless Molecular profile of endothelial invasion of three-dimensional collagen matrices: insights into angiogenic sprout induction in wound healing Am J Physiol Cell Physiol, November 1, 2008; 295(5): C1215 - C1229. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Deng, S. Lang, C. Wylie, and S. R. Hammes The Xenopus laevis Isoform of G Protein-Coupled Receptor 3 (GPR3) Is a Constitutively Active Cell Surface Receptor that Participates in Maintaining Meiotic Arrest in X. laevis Oocytes Mol. Endocrinol., August 1, 2008; 22(8): 1853 - 1865. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Duggavathi, D. H. Volle, C. Mataki, M. C. Antal, N. Messaddeq, J. Auwerx, B. D. Murphy, and K. Schoonjans Liver receptor homolog 1 is essential for ovulation Genes & Dev., July 15, 2008; 22(14): 1871 - 1876. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Sriraman, U. Eichenlaub-Ritter, J. W. Bartsch, A. Rittger, S. M. Mulders, and J. S. Richards Regulated Expression of ADAM8 (a Disintegrin and Metalloprotease Domain 8) in the Mouse Ovary: Evidence for a Regulatory Role of Luteinizing Hormone, Progesterone Receptor, and Epidermal Growth Factor-Like Growth Factors Biol Reprod, June 1, 2008; 78(6): 1038 - 1048. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Gao, C. De Geyter, K. Kossowska, and H. Zhang FSH stimulates the expression of the ADAMTS-16 protease in mature human ovarian follicles Mol. Hum. Reprod., July 1, 2007; 13(7): 465 - 471. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L. Russell and R. L. Robker Molecular mechanisms of ovulation: co-ordination through the cumulus complex Hum. Reprod. Update, May 1, 2007; 13(3): 289 - 312. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Zhu, P.C.K. Leung, and C.D. MacCalman Expression of ADAMTS-5/implantin in human decidual stromal cells: regulatory effects of cytokines Hum. Reprod., January 1, 2007; 22(1): 63 - 74. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Wen, H. Zhu, S. Murakami, P. C. K. Leung, and C. D. MacCalman Regulation of A Disintegrin And Metalloproteinase with ThromboSpondin Repeats-1 Expression in Human Endometrial Stromal Cells by Gonadal Steroids Involves Progestins, Androgens, and Estrogens J. Clin. Endocrinol. Metab., December 1, 2006; 91(12): 4825 - 4835. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. H. Ng, H. Zhu, C. J. Pallen, P. C.K. Leung, and C. D. MacCalman Differential effects of interleukin-1beta and transforming growth factor-beta1 on the expression of the inflammation-associated protein, ADAMTS-1, in human decidual stromal cells in vitro Hum. Reprod., August 1, 2006; 21(8): 1990 - 1999. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Hernandez-Gonzalez, I. Gonzalez-Robayna, M. Shimada, C. M. Wayne, S. A. Ochsner, L. White, and J. S. Richards Gene Expression Profiles of Cumulus Cell Oocyte Complexes during Ovulation Reveal Cumulus Cells Express Neuronal and Immune-Related Genes: Does this Expand Their Role in the Ovulation Process? Mol. Endocrinol., June 1, 2006; 20(6): 1300 - 1321. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Shimada, I. Hernandez-Gonzalez, I. Gonzalez-Robayna, and J. S. Richards Paracrine and Autocrine Regulation of Epidermal Growth Factor-Like Factors in Cumulus Oocyte Complexes and Granulosa Cells: Key Roles for Prostaglandin Synthase 2 and Progesterone Receptor Mol. Endocrinol., June 1, 2006; 20(6): 1352 - 1365. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. A. Tullet, V. Pocock, J. H. Steel, R. White, S. Milligan, and M. G. Parker Multiple Signaling Defects in the Absence of RIP140 Impair Both Cumulus Expansion and Follicle Rupture Endocrinology, September 1, 2005; 146(9): 4127 - 4137. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
