Ha Emek Medical Center, Afula and Technion Israel Institute of Technology, Haifa, Israel

Transcription

1 FERTILITY AND STERILITY VOL. 79, SUPPL 1, MARCH 2003 Copyright 2003 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Regulation of matrix metalloproteinase-9 (MMP-9), tissue inhibitor of MMP, and progesterone secretion in luteinized granulosa cells from normally ovulating women with polycystic ovary disease Izhar Ben-Shlomo, M.D., a,b Shlomit Goldman, Ph.D., a and Eliezer Shalev, M.D. a,b Ha Emek Medical Center, Afula and Technion Israel Institute of Technology, Haifa, Israel Received February 15, 2002; revised and accepted July 2, Reprint requests: Eliezer Shalev, M.D., Department of Obstetrics and Gynecology, Ha Emek Medical Centre, Afula 18101, Israel (FAX: ; shalev_e@hotmail.com). a Laboratory for Research in Reproductive Sciences, Department of Obstetrics and Gynecology, Ha Emek Medical Center. b Rappaport Faculty of Medicine, Technion Israel Institute of Technology /03/$30.00 doi: /s (02) Objective: To investigate the regulation of MMP-9, TIMP-1, and progesterone via three signal transduction pathways in luteinized granulosa cells from normal ovulatory and PCOD women. Design: In vitro study. Setting: Laboratory for Research in Reproductive Sciences, Department of Obstetrics and Gynecology, Ha Emek Hospital, Afula, Israel. Patient(s): Ten normal ovulatory and 10 women with polycystic ovary disease (PCOD) treated in an assisted reproduction program. Intervention(s): Cultured cells were exposed to phorbol 12-myristate 13-acetate (TPA), acting via protein kinase C (PKC), to epidermal growth factor (EGF), acting via protein tyrosine kinase (PTK), and to forskolin, acting via protein kinase A (PKA). Main Outcome Measure(s): Secretion of MMP-9, TIMP-1, and progesterone. Result(s): Phorbol 12-myristate 13-acetate elicited an increase in MMP-9 and TIMP-1 secretion in both groups and apparently did not affect progesterone secretion. Epidermal growth factor did not change significantly neither MMP-9 nor TIMP-1 secretion but dose dependently decreased MMP-9 TIMP-1 ratio and increased progesterone secretion in the PCOD group. Forskolin inhibited MMP-9 activity and increased TIMP-1 and progesterone secretion in both groups. Progesterone production was inversely related to the ratio of MMP-9 TIMP-1 regardless of cell origin. Conclusion(s): In this preliminary study, similar and divergent patterns have emerged in the regulation of MMP-9 and TIMP-1 in human luteinized granulosa cells. Repressing MMP-9 TIMP-1 ratio may have an important modulatory effect on progesterone secretion. (Fertil Steril 2003;79(Suppl 1): by American Society for Reproductive Medicine.) Key Words: Matrix metalloproteinases, luteinized granulosa cells, polycystic ovary syndrome, tissue inhibitor of metalloproteinase, progesterone The polycystic ovary syndrome or disease (PCOD) is frequently associated with anovulation infertility and frequent abortions. The main features of this disease state are as follows: anovulation, hyperandrogenism, disturbed hypothalamic function in the form of reversed LH to FSH ratio, and decreased insulin sensitivity, for which the primary underlying cause is still enigmatic (1 3). Once infertility in these women is overcome by ovulation induction, although the majority carry their pregnancies to term, there is an increased rate of miscarriage. Several reasons were suggested for this, among which were dysfunction of the corpus luteum, obesity, and follicular phase exposure of oocytes to high levels of LH (4 6). Attempts to reverse this tendency, known as luteal support regimens, include stimulation of the corpus luteum either by injection of hcg or by provision of progesterone. A fundamental aspect of corpus luteum (CL) formation and regression is dynamic and extensive tissue remodeling remodeling via MMPs (7, 8). It is well known from studies in other mammals that regression of the corpus luteum is associated with a significant increase in the activity 694

2 of MMPs (7 10). The role of extracellular MMPs in ovarian tissue remodeling during the life span of the follicle has been documented in numerous studies. In these, the mammalian ovary was revealed to express MMPs 1, 2, 3, 9, and possibly more (11, 12), as well as the naturally occurring tissue inhibitors of MMPs (TIMPs). Cultured luteinized granulosa cells from human ovaries were shown to express MMPs 9 and 2 and TIMP-1 (9, 13, 14). Furthermore, hcg was shown to suppress the overall activity of granulosa cell derived MMPs by both suppression of their production by the cultured cells and enhancement of TIMP production (9). In a previous study we found increased expression of matrix metalloproteinases 2 and 9 in cultured human luteinized granulosa cells from women with PCOD, compared with cells from normal ovulatory women (15). In view of all the above, we set out to explore different signal transduction pathways for MMP-9, TIMP-1, and progesterone in the granulosa cell, in normal ovulatory and PCOD women, both under gonadotropin stimulation. MATERIALS AND METHODS Patients and Induction of Ovulation The study population included 10 normal ovulatory women (controls; aged 25 to 36 years; treated with assisted reproduction for male factor infertility) and 10 women affected with PCOD (study group; aged 26 to 36 years) diagnosed by endocrine profile as described by us earlier (15). Criteria were hyperandrogenism, LH FSH ratio of 2, and typical ultrasonographic appearance of the ovary. Ultrasonic evidence was considered to be the presence of large ovaries with thickened cortex and multiple subcapsular follicles of 4 to 8 mm diameter before stimulation. These ultrasound criteria have been widely used in European studies to define polycystic ovarian morphology (16). Both groups of patients underwent induction of ovulation for IVF and ICSI-ET, as previously described (15). The hospital ethics committee, acting as institutional review board, approved the use of granulosa cells for the study. Each participating patient consented to this as well. The protocols used for controlled ovarian hyperstimulation were two as described elsewhere (17, 18): long or short GnRH-agonist before administration of gonadotropins. For the long protocol, DTRP-6-GnRH (Decapeptyl, Ferring, Malmo 246, Sweden; 3.75 mg controlled release) was injected IM during the midluteal phase, and ovarian downregulation was verified 2 weeks later, before hmg IU/d (Humegon, Organon, Os, The Netherlands) was begun. For the short protocol, subcutaneous DTRP-6-GnRH (0.1 mg/d) was begun on the 1st day of menstruation, and hmg ( IU/d) was added on the 3rd day. The two protocols were equally distributed between groups. Human chorionic gonadotropin (Pregnyl, Organon, Os, The Netherlands) 5,000 10,000 IU IM was administered when at least three follicles reached a mean diameter of 18 mm. Oocyte retrieval was performed transvaginally under ultrasonographic guidance (5.0 MHz probe, SSD-1400, Aloka, Tokyo, Japan) hours later. Culture Human luteinized granulosa cells were harvested only from large follicles ( 16 mm diameter). Cell preparation was commonly employed and was identical to that described in detail in our previous study (15). Cells were cultured in triplicates at a density of 300,000 cells per well in 0.5 ml human tubal fluid medium (HTF medium, Irvine Scientific, Santa Ana, CA), replete with 10% artificial serum substitute (SSR, Irvine Scientific). After 24 hours of culture to facilitate cell attachment, the medium was removed, and serumfree medium supplemented with antibiotics was added. The cells were cultured with various concentrations of M 12-O-tetradecanoylphorbol-13-acetate (TPA), M forskolin, or ng/ml epidermal growth factor (EGF). Cells were cultured for an additional 48 hours, and media were removed for analysis. Control consisted of serum-free medium alone. Viability of cells as assessed by trypan blue, was 86 91% in both clinical groups. Unless stated otherwise, all materials were purchased from Sigma Chemical (St. Louis, MO). Protein Assay The total protein content of luteinized granulosa cells was determined using protein assay kit with bovine serum albumin as the standard (Bio-Rad Laboratories, Inc., Washington, DC). Sample volume of 5 to 20 L was used in the assay. Substrate Gel Electrophoresis (Zymography) To detect proteolytic activity within the conditioned media (CM), substrate gel electrophoresis (zymography) on gels containing gelatin as the substrate was used (19). Conditioned media (10 L) diluted to contain 50 g protein, mass marker (10 L), and standard commercial gelatinases AandB(7 L; Oncogene Science, Cambridge, MA) were diluted with 4 sample buffer (5% sodium dodecyl sulfate [SDS], 20% Glycerol in 0.4 M Tris, ph 6.8 containing 0.02% bromophenol blue without 2-mercaptoethanol). Samples were electrophoresed through a 10% polyacrylamide gel containing 0.5% gelatin (5 mg/ml). After electrophoresis, gelatin gels were washed twice in 2.5% Triton X-100 for 15 minutes. Gelatin gels were incubated for 24 hours at 37 C in 0.2 M NaCl, 5 mm CaCl 2, 0.02% Brij 35, and 50 mm Tris, ph 7.5. The buffer was decanted and the gel stained with Coomassie Blue G in 30% methanol and 10% acetic acid for 10 minutes at room temperature on a rotary shaker. Stain was washed out with water until clear bands were seen. Finally, the gel was incubated for 30 minutes in 45% methanol and 5% glycerol before drying overnight between stretched sheets of cellophane. Areas where proteolytic activity degraded the gelatin are seen as absence of staining. These FERTILITY & STERILITY 695

3 were quantitated using the BioImaging gel documentation system (Dinco & Renium, Jerusalem, Israel) endowed with TINA software (Raytest, Staubenhardt, Germany). In addition, a validation/inhibition study with 10 mm EDTA was performed (data not shown). This procedure chelates the Zn 2 at the active site of MMPs and inhibits their activity (20). Western Blot Analysis To detect TIMP-1, CM and mass marker (10 L) were diluted with 4 sample buffer (5% SDS, 20% glycerol in 0.4 M Tris, ph 6.8 containing 0.02% bromophenol blue without 2-mercaptoethanol) and subjected to 10% polyacrylamid gel electrophoresis. After electrophoresis, the proteins (50 g/ lan) were blotted from the SDS-PAGE onto 0.45 m nitrocellulose membranes (Scheicher & Schuel, Dassel, Germany). Nonspecific binding sites were blocked by incubating the nitrocellulose membranes overnight with 20% nonfat milk and Tris-buffered saline containing 0.01% Tween-20. The membranes were then washed twice with Tris-buffered saline containing 0.5% Tween-20, and incubated for 1 hour with mouse antihuman TIMP-1 antibody (1.0 g/ml; Oncogene Science, Cambridge, MA) in 10% nonfat milk and Tris-buffered saline, containing 0.01% Tween-20. The membranes were subsequently washed with Tris-buffered saline, containing 0.5% Tween-20 and incubated for 1 h with horseradish peroxidase-conjugated anti-mouse rabbit secondary antibody (Jackson ImmunoResearch, West Grove, PA) in 10% nonfat milk and Tris-buffered saline, containing 0.01% Tween-20, then detected by enhanced chemiluminescence (Amersham International) and quantified by densitometry as above. Progesterone Assay Progesterone was assayed via enzyme immunoassay, a technique based on competition between an unlabeled antigen and an enzyme-labeled antigen for a fixed number of antibody-binding sites. The amount enzyme-labeled antigen bound to the antibody is inversely proportional to the concentration of progesterone present in the sample. Progesterone in the CM was assayed using the EIA kit from Diagnostic Systems Laboratory Inc. (Webster, TX). Statistical Methods Each experiment was repeated at least three times with at least three different sets of granulosa cells. Statistical analysis of the data was performed using Student s t test where two treatments were compared, and ANOVA test when more than two treatments were evaluated (e.g., dose-dependent responses). P.05 was considered significant. Correlation between progesterone values and MMP-9 TIMP-1 ratio was tested with correlation test. RESULTS Effects of TPA on MMP-9, TIMP-1, and Progesterone Secretion by Luteinized Human Granulosa Cells Normally Ovulatory vs. PCOD Patients Cells were incubated for 48 h with 10 9 M, 10 8 M, and 10 7 M of TPA. In cells originating from normal patients, TPA, at concentrations of 10 7 M, increased MMP-9 secretion by 210% 25.8 (P.01). In cells from PCOD patients, TPA, at a concentration of 10 7 M, increased MMP-9 secretion by % ( SE; P.01; Fig. 1). In cells from normal ovulatory women, 10 7 M TPA increased TIMP-1 secretion by % (P.05), whereas in cells from PCOD patients, 10 8 M and 10 7 M TPA increased TIMP-1 secretion by % and %, respectively (P.01; Fig. 2). The ratio of MMP-9/ TIMP-1 was increased to by 10 7 M TPA in cells from normal ovulatory women, whereas it remained constant in cells from PCOD patients (Fig. 3). 12-O-tetradecanoylphorbol-13-acetate (Fig. 4) did not significantly affect progesterone secretion. Effects of EGF on MMP- 9, TIMP-1, and Progesterone Secretion by Luteinized Human Granulosa Cells Normally Ovulatory vs. PCOD Patients Incubation of cells from both PCOD patients and normal ovulatory women in the absence or presence of EGF ( ng/ml) did not significantly affect either MMP-9 or TIMP-1 secretion (Figs. 1 and 2). However, in cells from PCOD patients, the MMP-9 TIMP-1 ratio was dose dependently reduced (P.05; Fig. 3). Progesterone secretion was increased in cells from PCOD patients: 53.4 ng per 10 g protein in the absence of EGF to 73 ng per 10 g protein in the presence of 100 ng/ml EGF (P.05) (Fig. 4). Effects of Forskolin on MMP-9, TIMP-1, and Progesterone Secretion by Luteinized Human Granulosa Cells Normally Ovulatory vs. PCOD Patients Cells were incubated for 48 hours in the absence and presence of forskolin 1 M 100 M. In cells from normal ovulatory patients, forskolin, at concentrations of 10 and 100 M, inhibited the secretion of MMP-9 by % and %, respectively (P.05 and P.01, respectively). In cells from PCOD patients, forskolin, at concentrations of 1, 10, and 100 M, inhibited MMP-9 secretion only by %, %, and %, respectively (P.05; Fig. 1). Concentrations of 10 and 100 M forskolin increased secretion of TIMP-1 from normal cells by % and %, respectively (P.05 and P.01; Fig. 2). In PCOD cells, TIMP-1 secretion was increased only after incubation with 100 M forskolin and only by % (P.01; Fig. 2). The ratio of MMP-9 to TIMP-1 was 696 Ben-Shlomo et al. Regulation of MMP-9 and TIMP-1 in granulosa cells Vol. 79, Suppl 1, March 2003

4 FIGURE 1 The effects of TPA (12-O-tetradecanoylphorbol-13-acetate), M; EGF, ng; and forskolin, M on MMP-9, secretion by luteinized human granulosa cells in normal ovulatory and PCOD patients. Control culture in serum-free medium. Conditioned media were analyzed for MMP-9 by zymography. Data are presented as mean SEM from four independent experiments. * Significant difference compared with control. inversely related to forskolin concentration both in normal and PCOD. In the former it decreased to , whereas in the latter, only to (Fig. 3). Incubation of cells from normal ovulatory women with 10 and 100 M forskolin increased progesterone secretion from 62 ng per 10 g protein to and ng per 10 g protein, respectively (P.01). In cells from PCOD patients only incubation with 100 M forskolin increased progesterone secretion from ng per 10 g protein to ng per 10 g protein (P.01; Fig. 4). Production of Progesterone by Cultured Luteinized Granulosa Cells Is Inversely Correlated to MMP-9 TIMP-1 Ratio As the cumulative impression indicated a possible overall correlation between progesterone production and MMP-9 to TIMP-1 ratio, we combined all coupled observations in media conditioned by the above cultures. We then separated those from PCOD patients from those of normal ovulatory patients (Fig. 5). Correlation existed when progesterone values were plotted against MMP- 9 TIMP-1 ratio. For cultures from normal ovulatory patients, the correlation coefficient was r (95% confidence interval, ; P.0015) and for PCOD, r ( ; P.0063). DISCUSSION In our previous study, we found that luteinized granulosa cells from patients with PCOD have a higher level of total gelatinolytic activity, mainly because of higher production of MMP-9, but also somewhat because of lower levels of TIMP-1 (15). Although the role of MMPs is destructive in nature and characterizes luteal regression, it is not clear whether this higher MMP-9 TIMP-1 ratio underlies the functional impairment, known as one of the characteristics of corpus luteum in PCOD patients. Several studies have shown that spontaneous regression of the corpus luteum involves activation of MMPs and TIMP-1 (8, 21). It is interesting to note that hcg, the primary signal of an implanting pregnancy, is a rescue factor, which almost completely arrests MMP production in the corpus luteum (13, 22). Human chorionic gonadotropin acts primarily via the camp protein kinase A (PKA) signal transduction pathway. Forskolin is a prototypical stimulator of the camp pathway. In human vein endothelial cells it enhances the induction of MMP-9, whereas it suppresses this induction in human mesangial cells and keratinocytes (23, 24). In the ovary, forskolin mimics the steroidogenic effect of gonadotropins on granulosa cells in culture (25). In our study, in addition to boosting progesterone production in luteinized granulosa cells, forskolin decreased the ratio of MMP-9 to TIMP-1. Furthermore, it did so much more significantly in cells from normal FERTILITY & STERILITY 697

5 FIGURE 2 The effects of TPA (12-O-tetradecanoylphorbol-13-acetate), M; EGF, ng; and forskolin, M on TIMP-1, secretion by luteinized human granulosa cells in normal ovulatory and PCOD patients. Control culture in serum-free medium. Conditioned media were analyzed for TIMP-1 by Western blot. Data are presented as mean SEM from four independent experiments. * Significant difference compared with control. ovulatory women as compared with cells from PCOD patients. This may be additional evidence of preexisting defects in the cells from PCOD patients. Unlike forskolin, which induces differentiation and function such as progesterone production, TPA is a prototypical inducer of proliferation acting via PKC signal transduction pathway. In nonovarian tissues, mostly cancerous, TPA was shown to induce cell invasiveness and hyperexpression of MMPs with a much less prominent increase in TIMP-1 production, resulting in a total higher MMP TIMP-1 ratio FIGURE 3 The effects of TPA (12-O-tetradecanoylphorbol-13-acetate), M; EGF, ng; and forskolin, M on MMP-9 TIMP-1 ratio, secretion by luteinized human granulosa cells in normal ovulatory and PCOD patients. Control culture in serum-free medium. Data are presented as mean SEM from four independent experiments. * Significant difference compared with control. 698 Ben-Shlomo et al. Regulation of MMP-9 and TIMP-1 in granulosa cells Vol. 79, Suppl 1, March 2003

6 FIGURE 4 The effects of TPA (12-O-tetradecanoylphorbol-13-acetate), M; EGF, ng; and forskolin, M on progesterone, secretion by luteinized human granulosa cells in normal ovulatory and PCOD patients. Control culture in serum-free medium. Conditioned media were analyzed for progesterone by ELISA. Data are presented as mean SEM from four independent experiments. * Significant difference compared with control. (26 28). In luteinized granulosa cells, TPA did not bring about any significant reduction of progesterone production but it did significantly increase MMP-9 as well as TIMP-1 production in both PCOD and normal cells, whereas, because of unequal increases, MMP-9 TIMP-1 ratio increased significantly in normal cells and remained constant in PCOD cultures. The increased ratio in normal cells was the only experimental permutation in which MMP-9 TIMP ratio increased beyond one. It is noteworthy that previous studies were not concordant on the effect of TPA in ovarian MMP production models. Some indicated increase, whereas others found decrease or no effect (29, 30). FIGURE 5 The relation between progesterone secretion and MMP-9 TIMP-1 ratio in media conditioned by luteinized human granulosa cells normal ovulatory and PCOD patients. Progesterone concentrations in conditioned media were plotted against MMP- 9 TIMP-1 ratio for normal ovulatory (triangles) and PCOD women (quadrangles). FERTILITY & STERILITY 699

7 Although in general the effects of EGF on our cell cultures were less dramatic, there were the effects of decreasing MMP-9 TIMP-1 ratio and increasing progesterone production only in cells from PCOD patients. Almahbobi et al. (31) found that in ovaries from PCOD patients there was a higher level of EGF receptors as compared with normal ovaries. The origin of this difference between normal and PCOD ovaries may be a preexisting difference but may well be an additional facet of deteriorated follicular development. The most consistent correlation which we observed in our preliminary study is between progesterone secretion and the ratio of MMP-9 to TIMP-1. This ratio reflects the main part of what can be called collagenolytic balance of the tissue. It is apparent from the results with TPA that an increase of the ratio beyond one is not accompanied by a change in progesterone production. Thus, the significant part of this phenomenon may be the strong association between reduction of the ratio and an increase in progesterone secretion. Indeed, from the graphic representation, one may hypothesize that despite basically lower progesterone secretion by cells from PCOD patients, previously described also by Doldi et al. (32), when this ratio decreases towards zero, both cell types increase their progesterone secretion to the same projected range. Theoretically there are two possibilities: either progesterone production and the collagenolytic balance are two independent phenomena which result from the same stimulus, or the two are interdependent. Two lines of evidence tend to support the latter. First, activation of different pathways led to the same association, which would make it unlikely that both are activated by the same mechanisms throughout. Second, it is neither MMP-9 nor TIMP-1 levels per se that correlate with progesterone secretion but their ratio. Thus, it is conceivable that the combined function of the two may influence steroidogenesis or that progesterone may influence their balance. The specific answer to this issue was beyond the scope of the present study, but it is difficult to ignore the potential importance of future corrective pharmacological intervention, based on such an answer, toward improvement of luteal function. References 1. Burghen GA, Given JR, Kitabchi AE. Correlation of hyperandrogen with hyperinsulinism in polycystic ovarian disease. J Clin Endocrinol Metab 1980;50: Balen AH, Conway GS, Kaltsas G, Techatraisak K, Manning PJ, West C, et al. Polycystic ovary syndrome: the spectrum of the disorder in 1741 patients. Hum Reprod 1995;10: Ben-Shlomo I, Homburg R, Shalev E. Hyperandrogenic anovulation (the polycystic ovary syndrome) back to the ovary? Hum Reprod Update 1998;4: Ashkenazi J, Farhi J, Orvieto R, Homburg R, Dekel A, Feldberg D, et al. Polycystic ovary syndrome patients as oocyte donors: the effect of ovarian stimulation protocol on the implantation rate of the recipient. Fertil Steril 1995;64: Wang JX, Davies MJ, Norman RJ. Polycystic ovarian syndrome and the risk of spontaneous abortion following assisted reproductive technology treatment. Hum Reprod 2001;16: Daya S. Issues in the etiology of recurrent spontaneous abortion. Curr Opin Obstet Gynecol 1994;6: Tsafriri A. Ovulation as a tissue remodeling process. Proteolysis, and cumulus expansion. Adv Exp Med Biol 1995;377: Liu K, Olfsson JI, Wahlberg P, Ny T. Distinct expression of gelatinase A (matrix metalloproteinase (MMP)-2), collagenase-3 (MMP-13), membrane type MMP-1 (MMP-14) and tissue inhibitor of MMPs type 1 mediated by physiological signals during formation and regression of the rat corpus luteum. Endocrinology 1999;140: Stamouli A, O Sullivan MJ, Frankel S, Thomas EJ, Richardson MC. Supression of matrix metalloproteinase production by HCG in cultures of human luteinized granulosa cells as a model for gonadotropininduced luteal rescue. J Reprod Fertil 1996;107: Pitzel L, Ludeman S, Wuttke W. Secretion and gene expression of metalloproteinase and gene expression of their inhibitors in porcine corpora lutea at different stages of the luteal phase. Biol Reprod 2000;62: Cossins J, Dudgeon TJ, Catlin G, Gearing AJ, Clements JM. Identification of MMP-18, a putative novel human matrix metalloproteinase. Biochem Biophys Res Commun 1996;228: Hulboy DL, Rudolph LA, Matrisian LM. Matrix metalloproteinases as mediators of reproductive function. Mol Hum Reprod 1997;3: Aston KE, Stamouli A, Thomas EJ, Vyas S, Iredale JP, Arthur MJ, et al. Effect of gonadotrophin on cell and matrix retention and expression of metalloproteinases and their inhibitor in cultured human granulosa cells modelling corpus luteum function. Mol Hum Reprod 1996;2: Goldman S, Dirnfeld M, Abramovici H, Kraiem Z. Triiodothyronine and follicle-stimulating hormone, alone and additively together, stimulate production of the tissue inhibitor of metalloproteinases-1 in cultured human luteinized granulosa cells. J Clin Endocrinol Metab 1997; 82: Shalev E, Goldman S, Ben-Shlomo I. The balance between MMP-9 and MMP-2 and their tissue inhibitor (TIMP)-1 in luteinized granulosa cells: comparison between women with PCOS and normal ovulatory women. Mol Hum Reprod 2001;7: Kyei-Mensah A, Zaidi J, Campbell S. Ultrasound diagnosis of polycystic ovary syndrome. Baillieres Clin Endocrinol Metab 1996;10: Shalev E, Giladi Y, Matilsky M, Ben-Ami M. Decreased incidence of severe ovarian hyperstimulation syndrome in high risk in vitro fertilization patients receiving intravenous albumin: a prospective study. Hum Reprod 1995;10: Ben-Shlomo I, Winer-Megnagi Z, Golan J, Eyali V, Geslevich J, Shalev E. Individual implantation rate: proposal for a new index for evaluation of assisted reproduction results. Fertil Steril 1997;68: Salamonsen LA, Nagase H, Suzuki R, Woolley DE. Production of matrix metalloproteinase 1 (interstitial collagenase) and matrix metalloproteinase 2 (gelatinase A: 72 kda gelatinase) by ovine endometrial cells in vitro: different regulation and preferential expression by stromal fibroblasts. J Reprod Fertil 1993;98: Rao VH, Bridge JA, Neff JR, Schaefer GB, Buehler BA, Vishwanatha JK, et al. Expression of 72 kda and 92 kda type IV collagenases from human giant-cell tumor of bone. Clin Exp Metastasis 1995;13: Goto T, Endo T, Henmi H, Kitajima Y, Kiya T, Nishikawa A, et al. Gonadotropin-releasing hormone agonist has the ability to induce increased matrix metalloproteinase (MMP)-2 and membrane type 1-MMP expression in corpora lutea, and structural luteolysis in rats. J Endocrinol 1999;161: Duncan WC, McNeilly AS, Illingworth PJ. The effect of luteal rescue on the expression and localization of matrix metalloproteinases and their tissue inhibitors in the human corpus luteum. J Clin Endocrinol Metab 1998;83: Kitahara M, Ichikawa M, Kinoshita T, Shiozawa S, Shigematsu H, Komiyama A. Prostacyclin inhibits the production of MMP-9 induced by phorbol ester through protein kinase a activation, but does not affect the production of MMP-2 in human cultured mesangial cells. Kidney Blood Press Res 2001;24: McCawley LJ, Li S, Benavidez M, Halbleib J, Wattenberg EV, Hudson LG. Elevation of intracellular camp inhibits growth factor-mediated matrix metalloproteinase-9 induction and keratinocyte migration. Mol Pharmacol 2000;58: Jalkanen J, Ritvos O, Huhtaniemi I, Stenman UH, Laatikainen T, Ranta T. Phorbol ester stimulates human granulosa-luteal cell cyclic adenosine 3, 5 -monophosphate and progesterone production. Mol Cell Endocrinol 1987;51: Johnson MD, Torri JA, Lippman ME, Dickson RB. Regulation of motility and protease expression in PKC-mediated induction of MCF-7 breast cancer cell invasiveness. Exp Cell Res 1999;25: Kupferman M.E, Fini ME, Muller WJ, Weber R, Cheng Y, Muschel RJ. Matrix metalloproteinase 9 promoter activity is induced coincident 700 Ben-Shlomo et al. Regulation of MMP-9 and TIMP-1 in granulosa cells Vol. 79, Suppl 1, March 2003

8 with invasion during tumor progression. Am JPathol 2000;157: Colandrea TD, D Armiento J, Kesari KV, Chada KK. Collagenase induction promotes mouse tumorigenesis by two independent pathways. Mol Carcinog 2000;29: McAllister JM, Byrd W, Simpson ER. The effects of growth factors and phorbol esters on steroid biosynthesis in isolated human theca interna and granulosa-lutein cells in long term culture. J Clin Endocrinol Metab 1994;79: Hori H, Uemura T, Minaguchi H. Effects of GnRH on protein kinase C activity, Ca2 mobilization and steroidogenesis of human granulosa cells. Endocr J 1998;45: Almahbobi G, Misajon A, Hutchinson P, Lolatgis N, Trounson AO. Hyperexpression of epidermal growth factor receptors in granulosa cells from women with polycystic ovary syndrome. Fertil Steril 1998; 70: Doldi N, Grossi D, Destefani A, Gessi A, Ferrari A. Polycystic ovary syndrome: evidence for reduced 3 beta-hydroxysteroid dehydrogenase gene expression in human luteinizing granulosa cells. Gynecol Endocrinol 2000;14:32 7. FERTILITY & STERILITY 701