| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ovary |
Department of Obstetrics/Gynecology3
Department of Anatomy and Cell Biology,4 Sahlgrenska Academy at Göteborg University, 413 45 Göteborg, Sweden
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
calcium, developmental biology, female reproductive tract, ovary, ovulation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Epithelial cells are characterized by two major histological findings [3]. Firstly, they have the ability to form barriers between the two tissue compartments that the epithelium separates. Secondly, the plasma membrane of the cell is intrinsically polarized into apical and basolateral domains. Epithelial tightness is maintained by formation of specialized structures known as tight junctions (TJ). Apart from restricting the paracellular movement of molecules and ions across the epithelial sheet, TJs also have a role in the maintenance of the apical/basolateral polarity [4, 5].
The TJ consists of a growing number of peripheral and integral membrane proteins that build up morphologically distinguishable strands that connect neighboring cells. ZO-1 (named after zona occludens, a TJ synonym) and later ZO-2 and ZO-3, located on the cytoplasmic facet of the cell membranes, were the first TJ proteins to be identified [6–8]. They are part of a multiprotein complex and bind directly to the integral TJ protein occludin [9] and the claudins [10], thereby linking the TJ to the cytoskeleton via direct or indirect interactions with actin [11]. It has been proposed that ZO-1 might be involved in tumorigenesis [12, 13]. Occludin was originally implicated in the formation and sealing of TJ, but functional studies have revealed that occludin can be excluded and yet the cells are still able to form TJ strands, suggesting a regulatory rather than structural role for this protein [14, 15]. Occludin is widely expressed in both endothelial and epithelial cells, but does not exists in cells and tissues without TJ [9]. Whereas only one occludin gene exists, claudin consists of a multigene family with 15 or more members. Functional analyses including knock-out experiments with lethality of animals have shown that claudin-1 or -2 is essential for TJ function [10, 16].
Earlier electron microscopic studies [17] have described the presence of incomplete TJ in OSE. Expression of TJ proteins in normal human ovaries and formation of functional TJ in OSE have, however, not been specifically studied before. In the present study we have analyzed the expression and localization of the TJ proteins ZO-1, occludin, and claudin-1 in normal human OSE of ovarian biopsies and in primary and secondary OSE cultures. In addition, we have developed an in vitro method to evaluate the barrier function of TJ in normal human OSE.
|
MATERIAL AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
OSE cells and tissue biopsies from normal ovaries were obtained from women operated on for benign nonovarian diseases at the Gynecology Unit, Sahlgrenska University Hospital, Göteborg, Sweden. OSE cells for culture were obtained from a total of 32 women. Normal ovarian tissue biopsies were collected from a total of nine women. A description of age at time for operation, reproductive status, and parity are found in Table 1. The study was approved by the Ethical Committee of Göteborg University and the patients had given their informed consent.
|
Harvest of Cells and Culture Conditions
To obtain OSE, cytobrushes (Cytobrush Plus; Medscand AB, Malmö, Sweden) were used as the first step after entering the peritoneal cavity at laparotomy or laparoscopy. The brushes were simultaneously rotated and moved 2–3 times over the ovarian surface, exerting only a slight pressure to minimize the risk of damaging or disrupting the underlying basal membrane and tunica albuginea. The brushes were withdrawn and immediately placed in culture medium, taken to the laboratory, and gently rubbed against each other to release the cells. The cells were then centrifuged for 5 min at 300 x g, diluted in fresh culture medium, and equally seeded into two 30-mm diameter Petri dishes (Falcon; Becton Dickinson, Meylan, France). OSE cells were cultured in a 1:1 mixture of M199:MCDB105 medium (Sigma Chemicals, St Louis, MO) supplemented with 15% fetal bovine serum (FBS; Life Technologies Ltd., Paisley, UK) and penicillin-streptomycin (100 IU/ml-100 µg/ml; Life Technologies Ltd.). Culture medium was exchanged every 2–3 days. When the cells reached confluence, cells were split and further cultured in Transwell inserts or Petri dishes according to each experiment.
An ovarian cancer cell line, NIH: OVCAR-3 (OVCAR), was purchased from American Type Culture Collection (ATCC; Rockville, MD). OVCAR cells were cultured in a 1:1 mixture of M199:MCDB105 medium (Sigma) supplemented with 10% FBS (Life Technologies Ltd.) and penicillin-streptomycin (100 IU/ml-100 µg/ml; Life Technologies Ltd.). OVCAR cells were grown in culture inserts and trans epithelial resistance (TER) was measured. As positive control in the immunoblot experiments, Madin-Darby canine kidney (MDCK) cells (a kind gift from Dr. Nillson, Goteborg, Sweden) were used. MDCK cells were cultured in Eagle minimal essential medium (Sigma Chemical Co.), supplemented with 10% fetal bovine serum (Life).
Immunofluorescence
Cells in first or second passage were grown on 18 x 18-mm glass cover slips (Histolab; Histolab Products AB, Göteborg, Sweden) until confluence. Cover slips were then removed from the culture dish, washed in PBS, fixed in cold acetone for 15 min, air dried at room temperature, and frozen at –20°C until analysis. Fresh-frozen tissue biopsies were cryosectioned and fixed as described for cultured cells. Cultured cells and tissue sections were incubated with 5% nonfat milk for 30 min, followed by incubation with primary antibodies against TJ proteins overnight: rabbit polyclonal antibody to ZO-1, occludin, and claudin-1 (1:500; Zymed Laboratories, San Francisco, CA). Mouse monoclonal antibodies against cytokeratin 8 (1:400; DAKO, Copenhagen, Denmark), cytokeratin AE1/AE3 (1:50; Boehringer Mannheim, Mannheim, Germany), E-cadherin (1:100; Transduction Laboratories, Nottingham, Aylesbury, UK), and N-cadherin (1:100; Zymed Laboratories) were used to further characterize epithelial features of the cells. Bound antibodies were visualized by biotinylated secondary horse anti-rabbit (1:200; Vector Laboratories, Burlingame, CA) followed by streptavidin-fluorescein isothiocyanate (1:200; Amersham, Buckinghamshire, UK) while the nuclei were counterstained with Hoechst 33342 (1 µg/ml; Molecular Probes Inc., Eugene, OR). In the control sections, which showed only negligible signals, the first antibody was replaced by 5% nonfat milk in PBS. Sections and cells were mounted in 4-diazabicyclo-2, 2, 2-octane (Dabco; Fluka, Buchs, Switzerland) and photographed with a Nikon ECLIPSE E600 fluorescence microscope (Nikon Corporation, Tokyo, Japan) using a digital cytoversion program (Applied Imaging Corporation, Santa Clara, CA).
Immunoblot Analysis
Confluent secondary cultures were solubilized in lysis buffer consisting of 62.5 mM Tris, 20% glycerol, and 2% SDS. The protein concentration of the supernatant was determined with the Micro BCA protein assay kit according to the manufacturer's instructions (Pierce, Rockford, IL). Samples were diluted in SDS-sample buffer and heated at 70°C for 10 min. Fifty micrograms of total protein from each sample were loaded into each lane of a SDS-polyacrylamide gel (NuPAGE 4–12% Bis-Tris Gel; Invitrogen Ltd., Paisley, UK) and separated by electrophoresis. Proteins were transferred to polyvinyldifluoride membrane (Amersham) using a blotting system (Novex miniblot; Invitrogen Ltd.) and incubated with primary antibodies: occludin (1:500) and claudin-1 (1:500). MDCK cells were used as positive control for TJ-expressing epithelia. Prestained standards (see blue; Invitrogen Ltd.) were used as molecular weight markers. Immunoreactive protein was visualized by chemiluminescence using alkaline phosphatase-conjugated secondary goat anti-rabbit (1:30 000) (Tropix, Bedford, MA) and CDP Star (Tropix) as substrate. The membrane was exposed to enhanced chemiluminescence film (Amersham).
Resistance Measurements
OSE cells were seeded with a density of 9.0 x 104/cm2 on 6.5-mm transparent Transwell filter inserts with a membrane pore size of 0.4 µm (Transwell-Clear; Corning COSTAR, Costar Europe Ltd., Badhoevedorp, The Netherlands). Culture medium was added into both chambers and routinely exchanged every 2–3 days. TER was measured every 2–3 days by a combined Millicell Electrical Resistance System (ERS; Millipore Corp., Bedford, MA) according to the manufacturer's instructions. TER was measured in OSE secondary cultures from seven different women. For each culture, cells were seeded in triplicate wells. TER (see Figs. 5 and 6) was determined by subtracting the background resistance of a blank filter and multiplying by the area of the monolayer (0.33 cm2 for the 6.5-mm inserts).
|
|
To further evaluate TJ function, filter-cultured OSE were subjected to chelation of extracellular Ca2+ with 2 mM ethyleneglycol-bis-(
-aminoethyl ether)-N, N'-tetraacetic acid (EGTA; Sigma) to rapidly disrupt cell-cell contacts. Confluent OSE was treated with EGTA for 30 min and then transferred back to EGTA-free media. TER was measured every 5 min during the EGTA experiment, and after 45 min, 12 h, and 24 h during the recovery period. Confluent OSE from the same patient, continuously cultured with ordinary medium, was measured at the same time points as controls. For each culture, TER was measured in triplicate wells. This experiment was performed on cultures from three different patients.
Data and Statistics for Resistance Measurements
The results from TER measurements were shown as median, top, and lowest range of the data from three measured wells for each time point. The possible statistical difference between EGTA test group and control group within the same patient were examined by the Mann-Whitney test. Differences were assumed to be significant at P < 0.05.
Electron Microscopy
Filter-grown OSE primary and secondary cultures from four women (2 pre- and 2 postmenopausal; see Table 1) were fixed for 1 h in 2.5% glutaraldehyde in 0.05 M sodium cacodylate, pH 7.4, followed by postfixation for 1 h in 1% OsO4. After dehydration in ethanol series, the specimens were embedded in epoxy resin. Ultrathin sections were contrasted with uranyl acetate and lead citrate and examined in a Zeiss electron microscope.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
By immunofluorescence microscopic analysis, we could determine that the main localization of the TJ-associated protein ZO-1 and the integral membrane proteins occludin and claudin-1 was at the cell borders of normal OSE (Fig. 1, A through C). The underlying stroma were without staining, indicated by stars. TJ protein immunofluorescent staining was the same in sections from both pre- and postmenopausal women. The histological appearance of the ovarian surface and the underlying stroma is shown with hematoxylin and eosin staining (Fig. 1D). In the negative control, where primary antibody was replaced by nonfat milk, no staining was seen (Fig. 1E).
|
Characterization of Normal Ovarian Surface Epithelium in Culture
To further investigate TJ in human, OSE cells, isolated by the brush technique from ovaries in pre- and postmenopausal women (n = 32), were cultured and used for immunohistochemistry, immunoblotting, and measurement of TER in Transwell experiments (Table 1). Primary cultured OSE cells typically formed a monolayer with a characteristic cobblestone-like appearance, and this growth pattern was maintained during the first three passages in approximately 45% of the patient samples (Fig. 2A). Cultures that consisted of a mixture of epithelial cells and great numbers of cells that were elongated and spindle shaped, suggesting fibroblast contamination (n = 18), were excluded from further studies [18]]. Cultured OSE cells stained positively for cytokeratin 8 (Fig. 2B) and cytokeratin AE1/AE3 (data not shown) for at least three passages, indicating that the epithelial features were maintained. In conformity with previous findings [19], the cell-cell adhesion molecule N-cadherin was found in cultured human OSE but staining for E-cadherin was negative (Fig. 2C). N-cadherin staining was localized to the cell borders (Fig. 2D).
|
Immunofluorescence staining of cultured OSE showed that the TJ-associated proteins ZO-1, occludin, and claudin-1 were distributed along the cell borders (Fig. 1, G through I), suggesting that OSE cells likely develop TJ also in culture. In the negative control where the primary antibody was replaced by nonfat milk, no staining was seen (Fig. 1F). With immunoblot, we could confirm expression of occludin and claudin-1 in cultured OSE (Fig. 3). The expression of both these proteins were equally strong in both pre- (lanes 1 and 4) and postmenopausal (lanes 2, 3, 5–8) women. Bands of approximately 22 kDa (claudin-1) and 65 kDa (occludin) were identified in all OSE samples analyzed. The migration pattern of the bands corresponded well with the band expressed by MDCK cells (lane 10) used as positive controls. In some OSE samples, an unknown band of approximately 10–14 kDa was found with the claudin-1 antibody. Ultrastructural examination confirmed that the cultured OSE cells were organized as a monolayered epithelium (Fig. 4A). Moreover, the cells showed a polarized appearance with an apical surface that contained microvilli and a basolateral surface that was delimited by a junctional complex located at the apical end of the intercellular space (Fig. 4B).
|
|
Barrier Function of Filter-Cultured OSE
To investigate if the expressed TJ proteins assembled into functional TJs, we cultured OSE cells from seven women on porous filters in culture inserts in which the build up of transepithelial resistance (TER) by the cell monolayer was measured. For this purpose, the cell number had to be expanded and therefore the cells were seeded after first passage into the inserts. TER was measured every second to every third day. Four out of seven secondary cultures developed constantly increasing TER in a time-dependent pattern. Three cultures did not reach confluence after seeding into the small inserts and were discarded (Fig. 5). Maximum TER, which never exceeded 200 
·cm2, was reached after approximately 14 days (Fig. 5). The ovarian cancer cell line (OVCAR) reached maximum TER levels after approximately 8–10 days. Maximum TER was found to be four to five times higher in OVCAR compared with OSE (Fig. 5). Two cultures that were followed for an extended time period maintained TER at a steady-state level for over 30 days (data not shown).
It was confirmed in Ca2+ switch experiments that the TER recorded in OSE cultures on filter indeed reflected epithelial tightness. A normal extracellular Ca2+ level is necessary for establishment of firm intercellular contact, including the assembly of TJ elements and a reduction of Ca2+ in the culture medium by adding a Ca2+ chelator will break up the TJ and lower TER dramatically [20]. Five minutes after adding 2 mM EGTA to OSE, TER was rapidly decreased (P < 0.05) when compared with a parallel culture maintained in normal medium (Fig. 6). TER decreased to <5 ·/cm2 during the first 30 min. After changing from Ca2+-depleted to regular medium, TER was partially recovered and fully restored after 24 h, indicating resealing of TJs.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
TJs are essential structures in normal epithelia, being responsible for the maintenance of apical-basal polarity and creation of a physiological barrier that prevents inappropriate leakage through the paracellular pathway between the separated tissue compartments [4]. Identification of TJ proteins and understanding of their assembly into functional TJs have been subject to intense studies [6, 9, 10]. The involvement of TJ proteins in disease is also increasingly recognized [12, 13, 23]. Earlier electron microscopy studies on normal human ovarian biopsies suggest that OSE has incomplete TJs [17], but to our knowledge, there is no information on TJ protein expression and barrier function in these cells. Immunohistochemical staining revealed that OSE expressed both transmembrane (occludin and claudin-1) and submembranous (ZO-1) TJ proteins at a typical location, i.e., at cell-cell contacts close to the apical pole of the cells. Counterstaining with cytokeratin antibody indicated that most TJ immunoreactivity was confined to OSE, although ZO-1 was also detected in the ovarian stroma. The expression levels of claudin-1 and occludin, estimated with semiquantitative immunoblotting of cell lysates, were found to be comparable with those of other epithelial cells. Thus, OSE fulfill biochemical and morphological criteria of epithelia that possess a functional barrier.
In the present study, we successfully cultured human OSE that retained an epithelial morphology for 2–3 passages, after which the cells acquired a more elongated fibroblast-like shape, consistent with previous reports [1, 18, 24]. Similar to OSE in vivo, the cultured cells expressed occludin, claudin-1, and ZO-1 that delineated the entire cell perimeter. The immunostaining pattern was not different from that regularly observed in other cultured epithelial cells that form functional TJs [25]. We therefore investigated the ability of OSE cells to establish a tight barrier, and for that purpose, cells were grown to confluence in Transwell bicameral chambers. Indeed, both primary and secondary OSE cultures developed a small but significant resistance across the cell monolayer. The transepithelial resistance was first recorded approximately 2 wk after plating, and it did not rise above 200 ·cm2 even though the cells were cultured for more than 4 wk. For comparison, primary thyroid epithelial cells are known to establish a resistance of >5000 ·cm2 within 6–7 days [26]. This indicates that OSE can be classified as a low-resistance epithelium similar to, e.g., the mucosa of the small intestine [27]. The results further illustrate the slow growth capacity of OSE cells in culture [18].
TJs are highly dynamic structures that may open or close in response to a multitude of bioactive agents, such as hormones, growth factors, cytokines, and various biochemical compounds that interfere with kinase activities. A simple way to test the functionality of TJs is to investigate if Ca2+ is required to maintain its stability. It is well established that TJs rapidly dissociate and become leaky when Ca2+ is removed from the culture medium. Conversely, readdition of Ca2+ makes TJ proteins reassemble and reseal the barrier. In fact, such Ca2+ switch experiments are widely used to elaborate the regulatory mechanisms of TJ formation and breakdown [25, 28, 29]. When a Ca2+-chelating agent (EGTA) was administered in excess to the culture medium of filter-grown OSE, we found that the resistance disappeared within 30 min. Moreover, the resistance was fully restored although at a slow rate after washout of EGTA and replacement with ordinary medium. This proves that the low resistance found in OSE is functionally significant and that its properties are not different from those of other epithelia. It is noteworthy, however, that OSE does not express E-cadherin [22], which actually is believed to be the primary target of EGTA in this kind of experiment. E-cadherin is a Ca2+-dependent adhesion molecule expressed in most epithelia, and extraction of Ca2+ from its extracellular domain by EGTA treatment immediately causes loss of adhesion [30]. The resulting retraction of adherens junctions then secondarily affects the integrity of adjacently located TJs. Altogether, this suggests that another cadherin may substitute for E-cadherin to functionally link TJ to the adherens junction. This function may be manufactured by N-cadherin, which is expressed in OSE in vivo [19] and also in the cultured OSE cells in the present study. There might be a correlation between low-resistance epithelium and the expression of N-cadherin because the transepithelial resistance in the ovarian cancer cell line (OVCAR-3), which express both E- and N-cadherin [31], is five times higher then in normal OSE (data not shown). This could also be the situation for the E- and N-cadherin-expressing epithelium lining the inclusion cysts.
One of the primary functions for TJs in OSE covering the surface might be to protect the interior of the ovary from substances in the peritoneal cavity. As in other organs, they would work as a fence between the apical and basolateral membrane domains and allow vectorial transport of materials between adjacent compartments [4]. At human ovulation, OSE cells disappear from the apical surface of the follicle to facilitate ovulation [32]. Cytokines are released in the stroma just beneath the protrusion point [33], with a possible effect on OSE TJ barriers. Increased TJ permeability by cytokines has earlier been shown in, e.g., human intestinal epithelium [34]. This monthly process includes breaking up both occludin and adherens junctions and indicates also that TJs are hormonally regulated. In line with this, estradiol increases TJ permeability of the cervical epithelium [35], while dexamethasone increases TJ formation in mammary epithelium [36, 37]. The TJ-associated protein ZO-1 is expressed in the corpora lutea, but not the atretic follicles, of the baboon ovary, indicating influence of gonadotrophins and/or steroids [38]. This is in agreement with the E-cadherin pattern of expression during corpus luteum formation [39]. In this study, we have presented a model with cultured OSE in a bicameral system (Transwell) where this hypothesis may be tested.
In conclusion, the results of this study strongly suggest that the somewhat undecided mesothelial-like human OSE are polarized and have tight junctions like other epithelial cells both in vivo and in vitro. The above characteristics reflect the complexity of OSE and show the importance of detailed studies on the normal biology of OSE and its relationship to ovarian cancer. The exact role for TJs in ovarian cancer biology is still uncertain and it would be interesting to study the function and expression of TJ proteins in a large study of ovarian tumors. This study also shows that you can study the function of TJs in cultured OSE.
|
FOOTNOTES |
|---|
2 Correspondence: Karin Sundfeldt, Department of Obstetrics/Gynecology, KK-Spec. lab. Gula straket 5, SU/Sahlgrenska Hospital, SE-413 45 Goteborg, Sweden. FAX: 46 31 418717; karin.sundfeldt{at}obgyn.gu.se
Received: 16 October 2003.
First decision: 29 October 2003.
Accepted: 13 February 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. A. Adissu, E. K. Asem, and S. A. Lelievre Three-Dimensional Cell Culture to Model Epithelia in the Female Reproductive System Reproductive Sciences, December 1, 2007; 14(8_suppl): 11 - 19. [Abstract] [PDF]
|
|
|
|
 |
|
 T. Groten, H.M. Fraser, W.C. Duncan, R. Konrad, R. Kreienberg, and C. Wulff Cell junctional proteins in the human corpus luteum: changes during the normal cycle and after HCG treatment Hum. Reprod., December 1, 2006; 21(12): 3096 - 3102. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.-H. Chou, L.-H. Wei, M.-L. Kuo, Y.-J. Huang, K.-P. Lai, C.-A. Chen, and C.-Y. Hsieh Up-regulation of interleukin-6 in human ovarian cancer cell via a Gi/PI3K-Akt/NF-{kappa}B pathway by lysophosphatidic acid, an ovarian cancer-activating factor Carcinogenesis, January 1, 2005; 26(1): 45 - 52. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
