Soluble intercellular adhesion molecule-1 in ovarian follicles: production by granulosa luteal cells and levels in follicular fluid

Transcription

1 FERTILITY AND STERILITY VOL. 69, NO. 4, APRIL 1998 Copyright 1998 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Soluble intercellular adhesion molecule-1 in ovarian follicles: production by granulosa luteal cells and levels in follicular fluid Paola Viganò, D.Sc.,* Francesco Fusi, M.D., Barbara Gaffuri, M.D., Viviana Bonzi, D.Sc., Augusto Ferrari, M.D., and Mario Vignali, M.D. University of Milan, Milan, Italy * II Department of Obstetrics and Gynecology, Clinica L. Mangiagalli and Istituto Auxologico Italiano. VI Department of Obstetrics and Gynecology, San Raffaele Hospital. II Department of Obstetrics and Gynecology, Clinica L. Mangiagalli. Received August 14, 1997; revised and accepted November 10, Reprint requests: Paola Viganò, D.Sc., II Department of Obstetrics and Gynecology, Clinica L. Mangiagalli, Via Commenda 12, Milan, Italy (FAX: ) /98/$19.00 PII S (97) Objective: To determine the concentration of the soluble form of intercellular adhesion molecule (ICAM)-1 in granulosa luteal cell conditioned media and in follicular fluid (FF). Design: Granulosa cells and FF samples were obtained at the time of oocyte retrieval for IVF. In 10 women, a total of 33 fluids were obtained from individual follicles, whereas in 70 women, the follicular aspirates were pooled. Setting: Clinica L. Mangiagalli and Reproductive Center, San Raffaele Hospital, Milan, Italy. Patient(s): Eighty women referred for IVF for tubal factor or male factor infertility. Intervention(s): Women underwent ovarian hyperstimulation. Main Outcome Measure(s): Soluble ICAM-1 was measured by an ELISA, and its levels were correlated with follicular size, the number of retrieved oocytes, and the number of follicles with a diameter of 15 mm. Result(s): The concentration of soluble ICAM-1 in granulosa luteal cell conditioned media was ng/ cells. Interleukin-1 can stimulate soluble ICAM-1 release in a dose-dependent manner. A significant positive correlation was demonstrated between levels of soluble ICAM-1 in pooled FF and the number of retrieved oocytes or the number of follicles with a diameter of 15 mm. Conclusion(s): Soluble ICAM-1 can be released by granulosa luteal cells and can be detected in FF after ovarian hyperstimulation. Levels of soluble ICAM-1 in FF correlate directly with some indices of ovarian function. (Fertil Steril 1998;69: by American Society for Reproductive Medicine.) Key Words: Soluble intercellular adhesion molecule (sicam)-1, oocyte, follicle, ovarian function Although ovarian function has been studied comprehensively and systematically from the viewpoint of steroidogenesis and gonadotropin responsiveness of ovarian cells (1), evidence has accumulated recently to suggest involvement of the structurally integrated immune system of the ovary and of the inflammation-associated phenomena within ovulating follicles as potential physiologic stimuli in inducing follicular maturation and ovulation (2 4). Indeed, the ovulatory process has been interpreted recently as a kind of physiologic inflammatory reaction characterized by infiltration of leukocytes and macrophages into ovulating follicles; ovarian release of proinflammatory cytokines such as tumor necrosis factor- (TNF- ) and interleukin (IL)-1, with their actions on granulosa and theca/interstitial cells; ultrastructural transformation of granulosa cells (GCs) into macrophagelike phagocytic cells; and an increase in prostaglandin levels at the time of follicle rupture (4). In this context, recent findings have emphasized the specific role of adhesion-promoting receptors in the regulation of lymphoid-ovarian cell contact and interaction and granulosa/luteal cell differentiation (2, 5, 6). In a previous study (7), we demonstrated that one of the adhesion molecules involved in the close relation between immune and endocrine ovarian cells is intercellular adhesion molecule (ICAM)-1, a single-chain kD sialoglycoprotein, which is a member of the immuno- 774

2 globulin gene superfamily (8). In particular, we showed that freshly aspirated GCs and GCs luteinized in culture expressed ICAM-1 gene and protein and that the binding of lymphoid cells to steroid-producing cells of the ovary was, at least in part, mediated by ICAM-1 (7). Constitutive surface expression of ICAM-1 varies with the cells examined, but activation studies in humans demonstrated that inflammatory mediators including IL-1, interferon (IFN)-, and TNF- (9) caused strong induction of ICAM-1 in a wide variety of tissues and greatly increased lymphocyte and monocyte adhesion by binding to its ligand, lymphocyte function-associated antigen-1 (10). A soluble form of ICAM-1 has been described recently that not only binds to lymphocyte function-associated antigen-1 but that is also believed to be released as an indirect consequence of inflammation or tissue damage in certain inflammatory and autoimmune diseases, including asthma, diabetes, rheumatoid arthritis, and malignant melanoma (11 13). Moreover, it has been suggested that the fast detection and measurement of circulating ICAM-1 would not only facilitate early diagnosis of many immunologic disorders but also would allow monitoring of therapy and intervention (14). Given the strict resemblance of the ovulatory process to an inflammation-associated reaction, we conducted this study to investigate the possibility of local ovarian release of soluble ICAM-1 during folliculogenesis and ovulation. To this aim, we measured the levels of soluble ICAM-1 in follicular fluid (FF) and in conditioned medium obtained from both cytokine-stimulated and unstimulated GCs luteinized in culture. MATERIALS AND METHODS Isolation of GCs and FF We obtained GCs and nonbloody FFs from women undergoing IVF-ET procedures at the Reproductive Center, San Raffaele Hospital, for tubal factor or male factor infertility. These women were pretreated with GnRH analogue (buserelin acetate, Suprefact; Hoechst, L Aquila, Italy), which was started in the midluteal phase of the preceding cycle and continued for 14 days. Follicle-stimulating hormone (Metrodin; Serono Laboratories, Inc., Rome, Italy) was started thereafter at a dose of three ampules per day, with dose variations depending on individual responses. Follicular development was monitored by daily measurements of serum estradiol and by ultrasonographic measurements of follicle diameter. Human chorionic gonadotropin (10,000 IU, Profasi; Serono Laboratories, Inc.) was administered when the size of at least two leading follicles reached 15 mm or serum estradiol levels were 1 nmol per mature follicle. Transvaginal ultrasonographic guided oocyte retrieval was performed 36 hours later. The contents of visible follicles were aspirated, centrifuged, separated from cells, and frozen at 70 C until assayed. Care was taken to puncture the follicles in avascular regions to avoid contamination with blood. Oocytes were identified and separated. In 70 women the follicular aspirates were pooled, whereas in 10 women, 33 samples of FF were obtained from individual follicles. This study was approved by the local Human Institutional Investigation Committee. Cell Cultures Human GCs were harvested with FF from the preovulatory follicles. Cells were then separated from FFs, and GCs were cultured as described previously (7, 15, 16). Briefly, the cells were washed and then dispersed at 37 C for 30 minutes in medium 199 with Earle s balanced salt solution (EBSS; BioWhittaker, PBI, Milan, Italy) containing 0.1% collagenase (Boehringer-Mannheim, Milan, Italy). Thereafter, the dispersed cells were layered on 3.5 ml of lymphocyte separation medium (Flow Laboratories, Milan, Italy) and centrifuged at 400 g for 10 minutes to separate GCs from debris and contaminating red blood cells. Cells collected from the interphase were resuspended in medium 199 with EBSS supplemented with 10% fetal bovine serum (FBS; BioWhittaker, PBI) and antibiotics. To reduce contaminating adherent peripheral blood mononuclear cells (PBMC), we then incubated the cells for 60 minutes at 37 C in humidified air containing 5% CO 2 on a Petri culture dish (Falcon, Becton Dickinson, Milano, Italy). Nonadherent cells were collected, washed in culture medium, and plated in Petri dishes (3 ml/dish, 60 mm in diameter). The medium was changed every other day. Granulosa cells in culture resembled the typical morphology of luteinizing cells, as reported previously by McAllister et al. (17). To confirm the purity of our cell cultures, we showed by flow cytometric analysis in a previous study that 4% of these cells expressed the CD-14 antigen, which represents the specific monocyte/macrophage marker (7). Peripheral blood mononuclear cells were obtained from buffy coats of three normal female blood-donor volunteers and were isolated on a Hypaque-Ficoll gradient as described previously (18). Monocytes were separated by incubating the mononuclear cells in RPMI-1640 (BioWhittaker, PBI) plus 10% FBS in tissue culture dishes for 1 hour twice and saving the adherent cells. Monocytes were activated by incubation with 1 g/ml of lipopolysaccharide (LPS; Sigma Chemical Co., St. Louis, MO). Both PBMC and LPS-activated monocytes were cultured in RPMI-1640 supplemented with 10% FBS and antibiotics. Collection of Media Conditioned by Cell Cultures Granulosa luteal cultures used for collection of conditioned media were allowed to proliferate for 10 days in medium supplemented with 10% FBS and antibiotics in a humidified atmosphere of 95% air and 5% CO 2 at 37 C. Afterwards, the cells were harvested, counted, and resus- FERTILITY & STERILITY 775

3 pended in Ham s F-10 medium plus FBS and were plated at a density of /ml in Petri dishes (1 ml/dish, 35 mm in diameter) for an additional 48 hours with and without IL-1. Both PBMC and LPS-activated monocytes were plated at a similar density for 48 hours. Finally, the medium was collected, centrifuged for 5 minutes at 400 g, and stored at 20 C until assayed for soluble ICAM-1. Soluble ICAM-1 ELISA The quantitative detection of soluble ICAM-1 in conditioned media obtained from different cell cultures and in FFs was performed using a commercially available ELISA kit (Bender MedSystem, Vienna, Austria). A dose of standard culture medium was also evaluated to exclude the presence of soluble ICAM-1 in the culture medium; levels of soluble ICAM-1 in culture medium alone were always 0.5 ng/ml. The concentration of soluble ICAM-1 was expressed as ng/ml, and the interassay and intraassay coefficients of variation were 8% and 4%, respectively. Flow Cytometry Immunofluorescent staining was performed with the use of an unconjugated murine anti ICAM-1 (anti-cd54; Immunotech, Marseille, France) followed by a secondary polyclonal goat anti-mouse IgG fluorescein isothiocyanate conjugate, as described previously (18). All samples were then analyzed on a flow cytometer (FACScan; Becton Dickinson & Co, Mountain View, CA) gated to exclude nonviable cells. Negative controls were stained with the fluorescent reagent alone. Markers were set according to negative controls to quantitate the percentage of positively stained cells. The percentage of positive cells was calculated on a histogram displaying log 10 of fluorescence (in arbitrary units) versus number of cells. Five thousand events were analyzed for experimental samples. Statistical Analysis Data are expressed as means SEM. Regression analysis or one-way analysis of variance and Fisher s least significant difference test as posttest were performed, as appropriate, to test for statistical significance of differences between groups. We used Statview 512 software on an Apple Macintosh computer. P 0.05 was defined as statistical significance. TABLE 1 Detection of soluble ICAM-1 in conditioned medium from different cell cultures. Cell type No. of cells/ml No. of samples Level of soluble ICAM-1 (ng/ml) PBMC BDL* LPS-activated monocytes BDL Granulosa luteal cells Note. ICAM-1 intercellular adhesion molecule-1; BDL below detectable limit. * Below the detectable limit of 6.25 ng/ml. Expressed as mean SEM. RESULTS Production of Soluble ICAM-1 and Expression of Surface ICAM-1 by Granulosa Luteal Cells Table 1 shows the concentrations of soluble ICAM-1 in eight samples of conditioned medium from granulosa luteal cells. In a previous study, we showed that 4% of our cell population expressed the CD-14 antigen, which represents the specific monocyte/macrophage marker (7). This finding allows us to exclude the possibility that soluble ICAM-1 content in conditioned media by granulosa luteal cell cultures might be attributed to contaminating monocytes/macrophages. However, we also measured soluble ICAM-1 concentrations in conditioned media by PBMC or LPS-activated monocytes cultured in similar conditions. Granulosa cells cultured (for 48 hours) in the absence of cytokine stimulation secreted readily detectable amounts of soluble ICAM-1 ( ng/ cells), whereas concentrations of soluble ICAM-1 in supernatants obtained from PBMC or LPS-activated monocytes were below detectable limits of the ELISA kit (6.25 ng/ml). This further confirms that contaminating leukocytes cannot be responsible for the release of soluble ICAM-1. Surface ICAM-1 expression was also evaluated in all IVF-GC samples analyzed. The percent of luteinized GCs positive for ICAM-1 was detected by flow cytometry. Percentages of cells positive for soluble ICAM-1 ranged from 85% to 99% in the different samples, indicating that release of soluble ICAM-1 in the culture medium was always associated with the surface expression of the protein and could not be attributed to a significant reduction on the cell membrane. Figure 1 shows a histogram of a representative flow cytometric analysis. Figure 2 shows the effect of IL-1 on soluble ICAM-1 shedding by granulosa luteal cells. A 48-hour treatment with 50 pg/ml of IL-1 induced a 28% increase in soluble ICAM-1 basal release. This IL-1 mediated stimulatory effect resulted in a 68% increase in levels of soluble ICAM-1 when a concentration of 500 pg/ml was used and in a 93% increase for a concentration of 5,000 pg/ml. These data demonstrate that granulosa luteal cells have the capacity to produce soluble ICAM-1 and that this production can be regulated by IL-1 in a dose-dependent manner. Soluble ICAM-1 Measurements in FF In a total of 33 samples of FF from individual follicles obtained from 10 patients, the mean SEM concentration of soluble ICAM-1 was ng/ml (overall range, ng/ml). Figure 3 shows the mean SEM levels of soluble ICAM-1 detected in individual follicles from each 776 Viganò et al. Soluble ICAM-1 and ovarian function Vol. 69, No. 4, April 1998

4 FIGURE 1 Histogram of a typical ICAM-1 expression profile of GCs luteinized in culture by flow cytometric analysis. Left panel represents nonbinding control. patient and divided according to follicular diameter. No statistically significant correlation was observed between intrafollicular concentrations of soluble ICAM-1 and follicular diameter (r 0.17, P 0.3). Soluble ICAM-1 levels in the FF from large follicles ( 18 mm) were ng/ml; these did not differ significantly from those in smaller (11 14 mm) or mediumsize follicles (15 18 mm) ( ng/ml and ng/ml, respectively) (Fig. 3). Indeed, wide variations in soluble ICAM-1 content were observed among size-matched FIGURE 2 Interleukin-1 modulation of soluble ICAM-1 (sicam-1) shedding by granulosa luteal cells. Results are expressed as means SEM from three different experiments. follicles obtained from different individuals. In contrast, follicles of different sizes from the same subject showed a low variability (Fig. 3). Pooled FF specimens representing 1 to 25 follicles each were obtained from 70 women. The concentration of soluble ICAM-1 in FFs showed a significant association with the number of retrieved oocytes (r 0.41, P ) (Fig. 4A) and with the number of follicles with a diameter of 15 mm, as measured by transvaginal ultrasonography at the time of ovum pickup (r 0.41, P ) (Fig. 4B). DISCUSSION Intercellular adhesion molecule-1/cd54 is an adhesion molecule expressed on cells of multiple lineages at the site of inflammation. It is a physiologic ligand for lymphocyte function-associated antigen-1 (CD11a/CD18) and Mac-1 (CD11b/CD18), and its expression may be induced or upregulated by proinflammatory cytokines such as IL-1, TNF-, and IFN- (8 10). This versatile molecule is essential for various immunologic functions, as antibodies to ICAM-1 have been shown to inhibit leukocyte adhesion to endothelial cells, granulocyte migration, mitogen- and antigen-induced lymphocyte proliferation, lymphocyte-mediated cytotoxicity, and mixed lymphocyte reactions in vitro (19). Circulating ICAM-1 can be detected in serum and other fluids in different pathologic situations, including pulmonary fibrosis, diabetes, allograft rejection, and many other inflammatory, autoimmune, and neoplastic diseases. Moreover, levels of soluble ICAM-1 have been shown to be elevated in malignant ascites of patients with ovarian carcinoma (12 14). We have demonstrated previously that surface ICAM-1 is expressed at both the RNA and protein levels by human GCs FERTILITY & STERILITY 777

5 FIGURE 3 Concentrations of soluble ICAM-1 (sicam-1) in FF from individual follicles. Results are shown as means SEM of soluble ICAM-1 levels of individual follicles of different sizes obtained from each woman (left) or divided according to follicle diameter (right). Levels of soluble ICAM-1 in FF from follicles of diameter 18 mm (n 13) showed no significant difference from levels in follicles of mm (n 9) or mm (n 11). and granulosa cells luteinized in culture (7). No information is currently available on the possibility of local production of soluble ICAM-1 by the ovary. Indeed, we provide herein the first evidence that FF contains immunoreactive soluble ICAM-1 and that ICAM-1 expressing human granulosa luteal cells are also able to release the glycoprotein. Conditioned media obtained from activated monocytes or PBMC cultured in conditions similar to IVF-GCs contained undetectable levels of soluble ICAM-1. This is consistent with data from a previous study, in which supernatants from 2-day PBMC cultures had to be concentrated eightfold only to approach a detectable value of 1 ng/ml (14). These findings, together with the low percentage of monocytes/ macrophages demonstrated in our cultures (7), allowed us to exclude that constitutive shedding of soluble ICAM-1 from granulosa luteal cell cultures might be attributed to contaminating leukocytes. Our study did not exclude the possibility that a certain amount of soluble ICAM-1 in FF might originate in the circulation. However, our observation that in vitro cultured granulosa luteal cells secreted the protein into the medium argues that the ovarian follicle is an important source of soluble ICAM-1 in FF. The amount of soluble ICAM-1 in individual follicles did not correlate with follicular size; indeed, most variations in FF content could be observed between different individuals rather than between follicles in the same individual. This observation led us to investigate whether the individual ovarian production of soluble ICAM-1 could be related to the functional activity of the ovaries. We chose to correlate FF levels of soluble ICAM-1 with the number of retrieved oocytes and with the number of larger follicles because these indices represent the definitive functional purpose of the ovary (4). Our results demonstrated that FF concentrations of soluble ICAM-1 increased with follicular recruitment and with the number of retrieved oocytes. This is of particular interest if we consider that ICAM-1 expression and release can be modulated by various cytokines (9) and that cytokines have been reported to display considerable actions on folliculogenesis, ovulation, and corpus luteal function (2 4). Specific binding sites for IL-1 and TNF- on GCs have been detected lately, and both cytokines have been shown to affect gonadotropin-induced steroidogenesis in cultured IVF-GCs (20, 21). Although surface ICAM-1 has been shown to be involved in GC binding to lymphoid cells (7), it is unclear whether the presence of soluble ICAM-1 in conditioned media from granulosa luteal cells and in FF simply represents a passive shedding of ICAM-1 from the cell surface as a consequence of ovarian hyperstimulation or whether the molecule might act as a mediator of intrafollicular signals. The correlation between levels of soluble ICAM-1 and the number of recruited follicles, but not with follicular size, suggests that if there is a specific functional significance for the molecule in the ovary, this should be in the early stages of follicle development. Even in the absence of a specific physiologic role for soluble ICAM-1, quantitative assessments of the soluble 778 Viganò et al. Soluble ICAM-1 and ovarian function Vol. 69, No. 4, April 1998

6 FIGURE 4 Correlation between levels of soluble ICAM-1 (sicam-1) in pooled FF specimens derived from 70 women subjected to ovarian hyperstimulation and the number of retrieved oocytes (A) (correlation coefficient (r) 0.41, P ) and the number of follicles with a diameter of 15 mm (B) (r 0.41, P ). protein may be of great value in understanding the effect of combinations of different cytokines at different phases of ovarian function and, most important, in explaining the interplay between the endocrine and immune systems. Because overall ovarian function involves mechanisms common to an inflammatory reaction (4), the physiologic implications of our present findings could include the possibility to control the specific immunologic pathways that underlie this reaction. In conclusion, the data presented herein support the following new observations: [1] media conditioned by human IVF-GCs contain soluble ICAM-1, and this secretion can be up-regulated by IL-1 in a dose-dependent manner; [2] FFs derived from women subjected to ovarian hyperstimulation contain immunoreactive soluble ICAM-1; and [3] levels of soluble ICAM-1 in FFs correlate significantly with the number of retrieved oocytes and the number of larger follicles. Future studies will address the potential role of this molecule as a mediator of ovarian physiology. References 1. Gougeon A. Regulation of ovarian follicular development in primates: facts and hypotheses. Endocr Rev 1996;17: Adashi EY. The potential relevance of cytokines to ovarian physiology: the emerging role of resident ovarian cells of the white blood cell series. Endocr Rev 1990;11: Ben-Rafael Z, Orvieto R. Cytokines involvement in reproduction. Fertil Steril 1992;58: Mori T, Takakura K, Fujiwara H, Hayashi K. Immunology of ovarian function. In: Bronson RA, Alexander NJ, Anderson D, Ware Branch D, Kutteh WH, editors. Reproductive immunology. London: Blackwell Science, 1996: Mayerhofer A, Lahr G, Gratzl M. Expression of the neural cell adhesion molecule in endocrine cells of the ovary. Endocrinology 1991;129: Fujiwwara H, Ueda M, Imai K. Human leukocytes antigen-dr is a differentiation antigen for human granulosa cells. Biol Reprod 1993; 49: Viganò P, Gaffuri B, Ragni G, Di Blasio AM, Vignali M. Intercellular adhesion molecule (ICAM)-1 is expressed on human granulosa cells and mediates their interaction with lymphoid cells. J Clin Endocrinol Metab 1997;82: Springer TA. Adhesion receptors of the immune system. Nature 1990; 346: Rothlein R, Czajkowsky M, O Neill MM, Marlin SD, Merluzzi VJ. Induction of intercellular adhesion molecule-1 on primary and continuous cell lines by pro-inflammatory cytokines. J Immunol 1988;141: Marlin SD, Springer TA. Purified intercellular adhesion molecule-1 (ICAM-1) is a ligand for lymphocyte function-associated antigen 1 (LFA-1). Cell 1987;51: Becker JC, Dummer R, Hartmann AA, Burg G, Schmidt RE. Shedding of ICAM-1 from human melanoma cell lines induced by INF- and tumor necrosis factor-. Functional consequence of cell-mediated cytotoxicity. J Immunol 1991;147: Gearing AJH, Newman W. Circulating adhesion molecules in disease. Immunol Today 1993;14: Roep BO, Heidenthal E, de Vries RRP, Kolb H, Martin S. Soluble forms of intercellular adhesion molecule-1 in insulin-dependent diabetes mellitus. Lancet 1994;343: Rothlein R, Mainolfi EA, Czajkowski M, Marlin SD. A form of circulating ICAM-1 in human serum. J Immunol 1991;11: Di Blasio AM, Viganò P, Cremonesi L, Carniti C, Ferrari M, Ferrari A. Expression of the genes encoding basic fibroblast growth factor and its receptor in human granulosa cells. Mol Cell Endocrinol 1993;96:R Di Blasio AM, Viganò P, Ferrari A. Insulin-like growth factor-ii stimulates human granulosa-luteal cell proliferation in vitro. Fertil Steril 1994;61: McAllister JM, Mason JI, Byrd W, Trant JM, Waterman MR, Simpson ER. Proliferating granulosa-lutein cells in long term monolayer culture: expression of aromatase, cholesterol side-chain cleavage and 3 beta hydroxysteroid dehydrogenase. J Clin Endocrinol Metab 1990;71: Viganò P, Pardi R, Magri B, Busacca M, Di Blasio AM, Vignali M. Expression of intercellular adhesion molecule-1 (ICAM-1) on cultured human endometrial stromal cells and its role in the interaction with natural killers. Am J Reprod Immunol 1994;32: Becker JC, Termeer C, Schmidt RE, Brocker EB. Soluble intercellular adhesion molecule-1 inhibits MHC-restricted specific T cell/tumor interaction. J Immunol 1993;151: Barak V, Yanai P, Treves AJ, Roisman I, Simon A, Laufer N. Interleukin-1: local production and modulation of human granulosa luteal cells steroidogenesis. Fertil Steril 1992;58: Wang LJ, Brannstrom M, Robertson SA, Norman RJ. Tumor necrosis factor- in the human ovary: presence in follicular fluid and effects on cell proliferation and prostaglandin production. Fertil Steril 1992; 58: FERTILITY & STERILITY 779