R. Scott Lucidi, M.D., Craig A. Witz, M.D., Michelle Chrisco, M.S., Peter A. Binkley, M.S., Sydney A. Shain, Ph.D., and Robert S. Schenken, M.D.

Transcription

1 A novel in vitro model of the early endometriotic lesion demonstrates that attachment of endometrial cells to mesothelial cells is dependent on the source of endometrial cells R. Scott Lucidi, M.D., Craig A. Witz, M.D., Michelle Chrisco, M.S., Peter A. Binkley, M.S., Sydney A. Shain, Ph.D., and Robert S. Schenken, M.D. Division of Reproductive Endocrinology and Infertility, Department of Obstetrics and Gynecology, University of Texas Health Science Center at San Antonio, San Antonio, Texas Objective: To characterize the source of variability in endometrial stromal cell (ESC) binding to peritoneal mesothelial cells (PMC). Design: In vitro study. Setting: University medical center. Patient(s): Reproductive-age women without endometriosis undergoing surgery for benign conditions. Intervention(s): None. Main Outcome Measure(s): Binding of ESCs (n 9) to PMCs collected from the anterior abdominal wall (AAW) (n 5), a commercially available mesothelial cell line (LP9) (three different passages) and normal ovarian surface epithelium (NOSE) (n 5). Result(s): There were no differences in the binding of same-source ESCs to mesothelial cells obtained from the AAW of different women, to different passages of LP9s or to NOSE of different women. There was a trend toward increased binding of ESCs to NOSE compared to AAW PMCs. In contrast, there were significant differences in the ability of ESCs obtained from different women to bind to same-source PMCs. Conclusion(s): There is significant variability in ESC binding to PMCs. This variation is dependent primarily on the source of the ESCs. The ESC binding to LP9 PMCs was similar to AAW PMCs and NOSE. (Fertil Steril 2005;84: by American Society for Reproductive Medicine.) Key Words: Endometriosis, mesothelium, endometrial stromal cell, attachment Received May 25, 2004; revised and accepted October 26, Supported by National Institutes of Health grant 1R01 HD Presented at the 51st Annual Meeting of the Society for Gynecologic Investigation, Houston, Texas, March 24 27, Reprint requests: R. Scott Lucidi, M.D., University of Texas Health Science Center at San Antonio, Department of Obstetrics and Gynecology, 7703 Floyd Curl Drive, San Antonio, Texas (FAX: ; lucidi@uthscsa.edu). Endometriosis is defined as the ectopic location of endometrial tissue outside the uterine cavity. Sampson s (1) theory proposes that fragments of menstrual endometrium pass retrograde through the fallopian tubes, then attach and grow on peritoneal surfaces. There is mounting evidence that Sampson s theory explains the development of the majority of endometriotic lesions (2). However, this theory fails to explain why only some women develop endometriosis. Although most women (90%) with patent fallopian tubes have reflux menstruation into the peritoneal cavity (3, 4), the incidence of endometriosis is much lower affecting approximately 5% 10% of reproductive-age women and up to 30% of infertile women (5). The specific factors involved in the implantation of refluxed endometrium to the peritoneal surface remain unknown. Some investigators have questioned whether an intact mesothelial cell layer prevents adhesion of endometrial fragments to the peritoneum (6 8). However, we recently demonstrated that proliferative, secretory, and menstrual endometrial fragments rapidly attach to intact peritoneal mesothelium in culture (9 12). These studies demonstrated that both endometrial stromal cells (ESCs) and endometrial epithelial cells attach to peritoneal mesothelial cells (PMCs) within 1 hour and that transmesothelial invasion rapidly occurs within hours. These studies also showed that the percentage of ESCs bound to PMCs varied. The source of this variation, however, is not clear. Because these studies used ESCs and PMCs from different sources, the variability could be due to differences in the ESCs, PMCs, or both. The purpose of this study was to examine the source of variability in ESC binding to PMCs. Using the model described herein, we compared the attachment of ESCs from several women to PMCs of different sources. The PMC sources used included the anterior abdominal wall (AAW), a commercially available mesothelial cell line (LP9), and normal ovarian surface epithelium (NOSE). MATERIALS AND METHODS Approval for this study and collection of endometrium and peritoneum was obtained from the Institutional Review 16 Fertility and Sterility Vol. 84, No. 1, July /05/$30.00 Copyright 2005 American Society for Reproductive Medicine, Published by Elsevier Inc. doi: /j.fertnstert

2 Board of the University of Texas Health Science Center at San Antonio. Endometrial Cell Culture Proliferative phase endometrium was obtained immediately after hysterectomy or by aspiration biopsy using a Pipelle (Unimar Inc., Prodimed, Neuilly-en-Thelle, France) during surgery performed for benign conditions (such as pelvic organ prolapse, stress urinary incontinence, menorrhagia) in women (n 9) not taking hormonal medication. Tissue samples were placed in Cellgro (Mediatech, Herndon, VA) complete serum-free medium for transport to the laboratory. We chose proliferative phase endometrium because it is readily available from women undergoing surgery and because it is more easily cultured. Endometrium was mechanically dispersed with a scalpel. The ESCs were separated from the endometrial samples as previously described by Kirk and Irwin (13). Briefly, endometrium was enzymatically digested with 0.1% collagenase type 1 and 0.05% DNAse. Endometrial epithelial cells were separated from ESCs by gravity sedimentation. The stromal cell-rich supernatant was placed in a culture flask and cells were allowed to adhere for 20 minutes then washed with medium. Adherent ESCs were cultured as monolayers in flasks with Dulbecco s modified Eagle Medium (DMEM)/ F-12 (1:1) (Sigma, St. Louis, MO) containing antibiotics/ antimycotics, 5 g/ml insulin (Sigma), and 10% fetal calf serum (HyClone, Logan, UT). After the second passage, ESCs were cultured on eightwell chamber slides. Purity of cultures was assessed morphologically after hematoxylin and eosin (H & E) staining. Stromal cells were fusiform, but more rounded than fibroblasts. Cells were also evaluated immunohistochemically using monoclonal antibodies to human cytokeratin, vimentin, von Willebrand factor, and CD45. The ESCs expressed vimentin, and did not express cytokeratin, von Willebrand factor, or CD45. Using these techniques of ESC separation, and as defined by these criteria, there is greater than 97% purity of ESCs (14, 15). Peritoneal Mesothelial Cell Culture Peritoneum from the AAW was obtained from reproductiveage women (n 5) without endometriosis undergoing surgery for benign conditions. At the time of surgery, there was no evidence of infection. Peritoneum was obtained from locations free of adhesions. Care was taken not to abrade the mesothelium or allow the tissue to air dry. The tissue was immediately placed in DMEM (Sigma) plus 10% defined fetal calf serum (FCS) and transported to the laboratory. The PMC cultures were established as previously described (16). Briefly, PMCs were enzymatically dispersed from sections of peritoneum using 0.1% collagenase, type 1 and 0.05% DNAse. Cells were plated in 25-cm 2 tissue culture flasks and grown in MCDB-131/Medium 199 (1:1) (Sigma) supplemented with epidermal growth factor (20 ng/ml), L-glutamine (2 mm), hydrocortisone (400 ng/ml), 1% antibiotic/antimycotic, HEPES buffer, and 15% FCS. The NOSE was collected from ovaries removed electively from reproductive-age women (n 5) undergoing surgery for benign disease. Collection and culture of NOSE cells was similar to that described by other investigators (17, 18). Briefly, the ovary was placed in trypsin EDTA solution (Sigma) for 5 minutes and then transferred to a 6-cm petri dish containing DMEM plus 10% defined FCS. The ovarian surface was gently scraped with a cell scraper (Nunc, Naperville, IL) and the medium containing NOSE cells was transported to the laboratory. The NOSE cells were cultured in the same medium as PMCs from the AAW. At first passage, AAW PMCs and NOSE were plated on eight-well chamber slides (Nunc). Morphologic assessment of these replicates of monolayer cultures was performed after H& E staining. The AAW PMCs and NOSE were polygonal with irregular borders (16). Purity of cultures was determined by incubation with monoclonal antibodies to human cytokeratin (Oncogene Science, Uniondale, NY), vimentin (Oncogene Science), CD45 (The Binding Site, San Diego, CA), von Willebrand factor (Dako, Carpinteria, CA) and epithelial membrane antigen (Dako) (16, 18, 19). Immunohistochemistry was performed as described for the ESC culture. The PMCs from the AAW and NOSE expressed cytokeratin and vimentin (19), and did not express CD45 and von Willebrand factor. The NOSE cells did not express epithelial membrane antigen (18). The LP9 peritoneal mesothelial cell line, obtained from the National Institutes of Health Aging Cell Repository (Coriell Institute for Medical Research, Camden, NJ), was grown in the same media as AAW PMCs and NOSE. Characterization of this cell line by immunohistochemistry was similar to AAW PMCs. ESC Labeling and Fluorescence Readings After culture, ESCs were collected using nonenzymatic cell dissociation media (Sigma) and washed with stromal cell complete medium. Cells were centrifuged at 650 g for 2 minutes. The pellet was labeled with 5 M calcein-am (Molecular Probes, Inc., Eugene, OR) for 20 minutes at 37 C, followed by washing in Hank s balanced salt solution (Invitrogen, Grand Island, NY). Calcein-AM is converted by acetoxymethyl ester hydrolysis in viable cells. Peak absorption and emission for calcein are at the wavelengths 494 nm and 517 nm, respectively. Calcein-AM treated ESCs were initially plated on 96-well plates at various concentrations obtained by serial dilution. Fluorescence readings were taken for each well using a fluorescence bioassay reader (model HTS-7000, Perkin-Elmer, Norwalk, CT) with filters for absorption and emission of 485 nm and 538 nm, respectively. Fluorescence readings obtained were linearly proportional to the number of cells in each well over a range of ,000 Fertility and Sterility 17

3 cells (r ). Experiments using serial dilutions of cells were repeated by plating the calcein-labeled ESCs over monolayers of LP9 PMCs. The monolayer of LP9s did not change the total fluorescence results (data not shown). Light microscopy demonstrated that 20,000 stromal cells per well produced a uniform distribution of ESCs without crowding or stacking of cells. Greater than 50,000 cells led to crowding of ESCs. Thus, we chose to place 20,000 cells per well on the AAW, LP9, and NOSE monolayers. Attachment Assay The AAW PMCs and NOSE at the second to third passage, and LP9 cells at the 4th, 7th, and 13th passage were placed on 96-well plates and grown to confluence. After the second to third passage, ESCs were collected using nonenzymatic cell dissociation media (Sigma), washed with stromal cell complete medium and labeled with calcein- AM. Cells were resuspended in ESC culture medium. Endometrial cells were plated over the confluent PMCs in 96-well plates at 20,000 cells per well in a total volume of 100 L. Plates were then cultured at 37 C for 1 hour in 5% CO 2 in air. After 1 hour of co-culture, wells were submerged and inverted in a 37 C bath of phosphate buffered saline (PBS) containing calcium and magnesium (Invitrogen). The bath was placed on an orbital mixer (Barnstead/Thermolyne, Dubuque, IA) set at 20 rpm and incubated at 37 C in 5% CO 2 in air for 15 minutes. During this time, nonadherent ESCs precipitated under gravity from the 96-well plate to the bottom of the PBS bath. Fluorescence readings were taken for each well before and after washing. Each assay was run in triplicate and each data point was calculated as the average of the triplicate. The percent of attached ESCs was calculated for each well ([Fluorescence value after washing/fluorescence value before washing] 100). The experiment was repeated three times using the same passage of cells. Confocal Microscopy The AAW PMCs, NOSE, and LP9 cells were cultured on 35-mm petri dishes with 10-mm glass microwell bottom (MatTek Corporation, Ashland, MA) until confluent monolayers were obtained. The ESCs labeled with 5 M calcein-am at a concentration of 300,000 cells/ml were added to the MatTek plate in a total volume of 1 ml. After 1 hour of co-culture nonadherent ESCs were removed using the adherence assay protocol. Attachment of ESCs to confluent PMCs was documented using confocal laser scanning microscopy. Confocal laser scanning microscopy was performed using an Olympus IX70 inverted microscope using a N.A. oil immersion objective and the Olympus Fluoview System (Nagano, Japan). Calcein was excited with an argon laser at 488 nm (Melles Griot, Carlsbad, CA) and emission was detected using a 505- to 525-nm BP filter. Simultaneous bright field images using differential interference contrast were collected. Digital images of 512 by 512 pixels were recorded using a speed of 1.69 s/scan. Calcein emissions were pseudocolorized green and images were processed for printing using Adobe Photoshop (San Jose, CA) on a Dell personal computer (Austin, TX). Statistical Analysis Comparisons of adherences within and across ESC lines and mesothelial cell lines were made by repeated measures oneway analysis of variance (ANOVA) using SPSS for Windows statistical software package version 10.0 (SPSS, Inc., Chicago, IL). Post hoc multiple range testing was performed using the Student-Newman-Keuls test. Linear correlation was performed using Pearson product moment correlation coefficient. RESULTS Validation of Adhesion Assay The ability of the adhesion assay to accurately measure the percentage of bound cells was verified by several methods. First, preliminary studies demonstrated that the maximum number of unattached ESCs was removed by gravity sedimentation by 15 minutes. Longer periods of time, up to 1 hour, did not result in a greater number of ESCs becoming detached (data not shown). Periods longer than 1 hour lead to detachment of the PMC monolayer from the 96-well plate. Furthermore, manual washing of individual wells with PBS after the 15-minute inversion in PBS and placement on an orbital shaker did not result in lower fluorescence counts. A separate approach involved collection of the nonadherent cells after manual washing of the individual wells. These unbound cells were then plated in empty wells on the same 96-well plate. Fluorescence readings showed that the percent bound cells plus the percent unbound cells totaled 101%. To confirm that we were measuring cells and not fluorescence in the media, the unbound cells were counted manually on a hemacytometer. The percent bound measured by fluorescence plus the percent unbound measured by manual counting totaled 99.5%. The intra-assay coefficient of variation among triplicate fluorescence data averaged Finally, confocal laser scanning microscopy confirmed attachment of fluorescent ESCs to confluent monolayers of AAW PMCs, LP9, and NOSE. Source of Variation in Adherence There were no differences in binding of ESCs to the five sources of AAW PMCs, three passages of LP9s, or five sources of NOSE (P.51,.54,.08, respectively) (Fig. 1). Post hoc analysis revealed a power of 0.98 to demonstrate a 50% difference in the rate of ESC binding to any of the five 18 Lucidi et al. Endometrial stromal cell attachment to mesothelial cells Vol. 84, No. 1, July 2005

4 FIGURE 1 Average endometrial stromal cell (ESC) binding to five different sources of anterior abdominal wall (AAW) mesothelium (A), three different passages of LP9 mesothelial cells (B), and five different sources of normal ovarian surface epithelium (NOSE) mesothelium (C). There were no significant differences in binding of ESCs to peritoneal mesothelial cells of the same type (i.e., AAW, LP9, and NOSE; error bars SEM). rate of binding to the LP9 cells (Fig. 2). For all ESCs, binding to LP9 cells correlated with binding to AAW PMCs and NOSE (r and 0.86, respectively). The ESCs obtained from different women demonstrated a significant variability in the rate of PMC binding. This variation in binding was greater than twofold when comparing ESCs with the greatest rate of binding to those with the lowest rate of binding. Post hoc multiple range testing revealed that, when controlling for variability in the rate of binding to PMCs from different sources (i.e., AAW, LP9s, and NOSE), the ESCs could be divided into overlapping groups from lowest to highest binding (P.001) (Fig. 3). DISCUSSION The aim of the current study was to evaluate the source of variability of ESC binding to PMCs. We wanted to determine whether the rate of ESC binding to PMCs is dependent on the source of endometrium, the source of PMCs, or both. The significant difference in ESC binding to PMCs was primarily dependent on the source of endometrium. There were no significant differences in the rate of binding of ESCs to PMCs of the same source. For example, the average binding by the nine ESC samples to the five sources of AAW PMCs is similar (range 21% 30%). Although there was variability in ESC binding to NOSE (range 30% 44%), this difference in binding was considerably lower than the twofold difference seen when considering ESC from different sources. These data suggest that the binding rate of ESCs to PMCs of the same type (i.e., AAW, LP9, or NOSE) is minimally influenced by the specific source of the PMCs. The present study also demonstrates that ESCs bind to the LP9 cell line at a rate similar to both AAW PMCs and NOSE. Moreover, the binding of ESCs to LP9 cells is consistent through multiple passages. This suggests that the commercially available LP9 line is suitable for in vitro experiments evaluating ESC PMC attachment. Lucidi. Endometrial stromal cell attachment to mesothelial cells. Fertil Steril AAW PMCs or to any of the three passages of LP9 cells with alpha Similar analysis for NOSE demonstrated a power of This decrease in power was a result of variability in NOSE binding. There was a trend (P.15) toward an increased rate of all ESC binding to NOSE compared to AAW PMCs (Figs. 2 and 3). In all cases, the ESC binding was greater to NOSE than to AAW PMCs. There was an intermediate The novel binding assay in this study using calcein-am to label ESCs offers several advantages. The labeling with this fluorophore is rapid, taking 20 minutes. Calcein-AM is nonfluorescent and is cell permeant. In live cells, calcein-am is converted to fluorescent calcein by intracellular esteraces. Thus, the fluorescent readings represent the number of viable ESCs placed in each well after lifting from monolayer culture. It is advantageous to label only viable cells so that the proportion of nonviable cells present does not inaccurately lower the measured proportion of cells that have become bound. This assay also uses a plate inversion technique with gravity sedimentation of unattached ESCs cells. Compared with manual rinsing of individual wells, we believe Fertility and Sterility 19

5 FIGURE 2 Comparison of average endometrial stromal cell binding to each of three types of peritoneal mesothelial cells (error bars SEM). AAW anterior abdominal wall; LP9 commercially available mesothelial cell line; NOSE normal ovarian surface epithelium. Lucidi. Endometrial stromal cell attachment to mesothelial cells. Fertil Steril that gravity sedimentation is a more consistent method of removing unattached cells. Also, gravity sedimentation permits the PMC monolayers to remain intact, whereas manual rinsing with a pipette can shear too forcefully the attached PMC monolayer from the culture plate. This study is limited in that proliferative phase endometrial cells were used. We chose proliferative phase endometrium because it is readily available from women undergoing surgery and because it is more easily cultured. It is likely that ESC binding to PMCs will involve the same interaction of FIGURE 3 Average binding to three different peritoneal mesothelial cells (PMCs) by endometrial stromal cell (ESCs) from nine patients. There was a significant difference in the ability of ESC from different patients to bind PMCs (P.001). Letters (a, b, c, d, e) designate significantly different groups by post hoc multiple range tests. For each ESC source, the rate of binding to LP9 PMCs was highly correlated with the rate of binding to anterior abdominal wall (AAW) mesothelium and normal ovarian surface epithelium (NOSE) PMCs (r and 0.86, respectively, P.01 for both). Lucidi. Endometrial stromal cell attachment to mesothelial cells. Fertil Steril Lucidi et al. Endometrial stromal cell attachment to mesothelial cells Vol. 84, No. 1, July 2005

6 cell adhesion molecules irrespective of whether the endometrium is proliferative, secretory, or menstrual. Hence, it is unlikely that the menstrual phase of the ESCs would alter our observation that the source of PMCs minimally influences the attachment of ESCs. Moreover, our previous studies have demonstrated significant variability in attachment using secretory phase endometrium as well. In summary, the present study demonstrates that the source of endometrium, rather than the source of PMCs, has the greatest effect on ESC binding to PMCs. We hypothesize that the ability of endometrial cells to bind PMCs may influence the development of endometriosis. As most women with patent fallopian tubes experience retrograde menstruation, altered rates of endometrial binding to the peritoneum might help explain why only some women develop endometriosis. Accordingly, comparison of attachment of endometrium from different groups of women, such as women with and without endometriosis, deserves evaluation. We believe that the attachment model used in this study is suitable to accomplish this goal. REFERENCES 1. Sampson J. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;14: Witz CA. Current concepts in the pathogenesis of endometriosis. Clin Obstet Gynecol 1999;42: Halme J, Hammond MG, Hulka JF, Raj SG, Talbert LM. Retrograde menstruation in healthy women and in patients with endometriosis. Obstet Gynecol 1984;64: Liu DT, Hitchcock A. Endometriosis: its association with retrograde menstruation, dysmenorrhoea and tubal pathology. Br J Obstet Gynaecol 1986;93: Eskenazi B, Warner ML. Epidemiology of endometriosis. Obstet Gynecol Clin North Am 1997;24: Groothius P, Koks CA, de Goeij AF, Dunselman GA, Arends JW, Evers JL. Adhesion of human endometrial fragments to peritoneum in vitro. Fertil Steril 1999;71: Koks CA, Groothuis PG, Dunselman GA, de Goeij AF, Evers JL. Adhesion of shed menstrual tissue in an in-vitro model using amnion and peritoneum: a light and electron microscopic study. Hum Reprod 1999;14: Dunselman GA, Groothuis PG, de Goeij AF, Evers JL. The mesothelium, Teflon or Velcro? Mesothelium in endometriosis pathogenesis. Hum Reprod 2001;16: Witz CA, Monotoya-Rodriguez IA, Schenken RS. Whole explants of peritoneum and endometrium: a novel model of the early endometriosis lesion. Fertil Steril 1999;71: Witz CA, Allsup KT, Montoya-Rodriguez IA, Vaughn SL, Centonze VE, Schenken RS. Culture of menstrual endometrium with peritoneal explants and mesothelial monolayers confirms attachment to intact mesothelial cells. Hum Reprod 2002;17: Witz CA, Thomas MR, Montoya-Rodriguez IA, Nair AS, Centonze VE, Schenken RS. Short-term culture of peritoneum explants confirms attachment of endometrium to intact peritoneal mesothelium. Fertil Steril 2001;75: Witz CA, Cho S, Centonze VE, Montoya-Rodriguez IA, Schenken RS. Time series analysis of transmesothelial invasion by endometrial stromal and epithelial cells using three-dimensional confocal microscopy. Fertil Steril 2003;79 Suppl 1: Kirk D, Irwin JC. Normal human endometrium in cell culture. Methods in Cell Biology 1980;21B: Simon C, Piquette GN, Frances A, el-danasouri I, Irwin JC, Polan ML. The effect of interleukin-1 beta (IL-1 beta) on the regulation of IL-1 receptor type 1 messenger ribonucleic acid and protein levels in cultured human endometrial stromal and glandular cells. J Clin Endocrinol Metab 1994;78: Kirk D, King RJ, Quigley MM, Gwatkin RB, Irwin JC. Hormonal regulation of human endometrial stromal cells in culture: an in vitro model for decidualization. Fertil Steril 1989;52: Witz CA, Montoya-Rodriguez IA, Miller DM, Schneider BG, Schenken RS. Mesothelium expression of integrins in vivo and in vitro. J Soc Gynecol Invest 1998;5: Kruk PA, Maines-Bandiera SL, Auersperg N. A simplified method to culture human ovarian surface epithelium. Lab Invest 1990;63: Ohtake H, Katabuchi H, Matsuura K, Okamura H. A novel in vitro experimental model for ovarian endometriosis: the three-dimensional culture of human ovarian surface epithelial cells in collagen gels. Fertil Steril 1999;71: La Rocca P, Rheinwald J. Coexpression of simple epithelia keratin and vimentin by human mesothelium and mesothelioma in vivo and in culture. Cancer Research 1984;44: Fertility and Sterility 21