Gap Junction Communication and Connexin 43

5 downloads 0 Views 296KB Size Report
Cx 50 [6], Cx 57 [7], and Cx 60 [8]. ..... which was measured at 490 nm at 6, 8, 9, 14, and 24 h after the addition of .... columns. OD, Optical density (arbitrary units).
BIOLOGY OF REPRODUCTION 63, 1661–1668 (2000)

Gap Junction Communication and Connexin 43 Gene Expression in a Rat Granulosa Cell Line: Regulation by Follicle-Stimulating Hormone1 Britta Sommersberg,3 Andreas Bulling,3 Ulrich Salzer,4 Ulrike Fro¨hlich,3 Robert E. Garfield,6 Abraham Amsterdam,5 and Artur Mayerhofer2,3 Anatomisches Institut3 and Institut fu¨r Medizinische Mikrobiologie, Immunologie und Hygiene,4 Technische Universita¨t Mu¨nchen, D-80802 Mu¨nchen, Germany Department of Molecular Cell Biology,5 Weizmann Institute of Science, Rehovot 76100, Israel Department of Obstetrics and Gynecology,6 University of Texas, Galveston, Texas 77555 ABSTRACT

INTRODUCTION

Follicle-stimulating hormone is the major regulator of growth and development of antral follicles in the ovary. Granulosa cells (GCs) in these follicles are coupled via gap junctions (GJs) consisting of connexin 43 (Cx 43). Because we and others have found that Cx 43 and GJs, respectively, are more abundant in large antral follicles compared with small antral and preantral follicles, we hypothesized that FSH may control Cx 43 gene expression, GJ formation, and intercellular communication. To directly address these points, we chose a rat GC line (GFSHR-17) expressing the FSH receptor and the Cx 43 gene. The functionality of FSH receptors was shown by the effects of porcine FSH, namely cell rounding, reduced cellular proliferation, and stimulation of progesterone production of GFSHR-17 cells, which are effects that were detectable within hours. Treatment with FSH also statistically significantly increased Cx 43 mRNA levels, as shown after 6 to 9 h in Northern blots. These effects were antedated by altered GJ communication, which was observed within seconds. Using a single-cell/whole-cell patch clamp technique, we showed that FSH rapidly and reversibly enhanced electrical cell coupling of GFSHR-17 cells. Increased GJ communication was associated with statistically significantly decreased phosphorylation of Cx 43, which was observed within 10 min after FSH addition, during immunoprecipitation experiments. Our results demonstrate, to our knowledge for the first time, that the gonadotropin FSH acutely and directly stimulates intercellular communication of GFSHR-17 cells through existing GJs. Moreover, FSH also increases levels of Cx 43 mRNA. These changes are associated with reduced proliferation and enhanced differentiation of GFSHR-17 cells. In vivo factors in addition to FSH may be involved in the regulation of GJ/GJ communication between GCs in the follicle, but our results suggest that improved cell-to-cell coupling, enhanced Cx 43 gene expression, and possibly, formation of new GJs are direct consequences of FSH receptor activation and may antedate and/or initiate the pivotal effects of FSH on GCs.

In the ovary, follicular development is under the tight control of FSH. Depending on the presence of FSH receptors, antral follicles have the potential to grow if FSH is present [1]. However, only a few follicles mature and ovulate; most undergo atresia [2]. What regulates the fate of an individual follicle is not clear, but compartmentalization in thecal and granulosa cells (GCs) and intercellular communication via gap junctions (GJs) within these compartments are assumed to be crucial for the mediation of signals initiating growth and/or atresia, respectively [3]. Gap junctions connect adjacent cells and enable them to communicate by exchanging inorganic ions and molecules with a molecular weight less than 1000 Da. The GJ channels consist of two hemichannels (i.e., connexons), which are built from six radially arranged, transmembrane proteins called connexins (Cx). End-to-end attachment of two Cx from adjacent cells results in one intercellular channel. Once a GJ is formed, the degree of intercellular communication can be controlled by diverse gating mechanisms and initiated by changes in pH, free intracellular Ca21, and phosphorylation of Cx molecules [4]. To date, 16 different forms of Cx have been described. They are classified according to their expected molecular weight. Several forms are present in the ovary, including Cx 46, Cx 45, Cx 30.3, Cx 32, Cx 37, Cx 26, Cx 43 [5], Cx 50 [6], Cx 57 [7], and Cx 60 [8]. Most ovarian Cx are restricted to defined locations, such as Cx 40 to the endothelia of blood vessels or Cx 37 to the oocyte [9]. In contrast, that Cx 43 is the most abundant ovarian Cx and expressed by several endocrine ovarian cells, including GCs, theca cells, and luteal cells, is generally accepted [4, 5, 10]. Recent evidence from the organ culture of ovaries from Cx 43 knockout mice underscores the pivotal role of Cx 43 in follicular development and reproductive function [11]. Despite this well-documented presence and indispensability of Cx 43, little is known regarding how Cx and GJ intercellular communication of GCs or thecal cells are regulated. Treatment of rats with eCG increased levels of Cx 43 in the ovary and shortly before ovulation, downregulation of Cx 43 in GCs by LH was shown to be vitally important in the GC layer for preparation of ovulation [12, 13]. Also, around the time of ovulation, nerve growth factor induces serine phosphorylation of Cx 43 in thecal cells and decreases their cellular communication [14]. What regulates Cx 43 and GJ communication in GCs of the follicle during follicular development and before ovulation is not known. However, results of immunocytochemical analysis [15], Western blotting [16], Northern blotting [17], and in situ hybridization (present study) indicate enhanced Cx 43 protein and mRNA expression in larger fol-

follicular development, FSH, granulosa cells, ovary Supported by DFG (Graduiertenkollleg 333), in parts by DFG Ma 1080/ 12-1, and by Volkswagen-Stiftung. A.A. is the incumbent of the Joyce and Ben B. Eisenberg Professional Chair of Endocrinology and Cancer Research at the Weizmann Institute of Science. 2 Correspondence: Artur Mayerhofer, Anatomical Institute TU Mu¨nchen, Biedersteiner Str. 29, D-80802 Mu¨nchen, Germany. FAX: 49 89 397035; e-mail: [email protected] 1

Received: 14 February 2000. First decision: 8 March 2000. Accepted: 7 July 2000. Q 2000 by the Society for the Study of Reproduction, Inc. ISSN: 0006-3363. http://www.biolreprod.org

1661

1662

SOMMERSBERG ET AL.

licles compared with smaller ones, suggesting an upregulation of Cx 43 gene expression during follicular growth [5]. Although these observations give circumstantial evidence for a stimulatory direct and/or indirect role of FSH on Cx 43, its precise effect on GJ formation and communication has not yet been examined. To address these points and to study the effect of FSH on Cx 43/GJ communication of GCs directly, we used a stabily transformed rat GC line that expresses the rat FSH receptor gene (GFSHR-17) [18] and Cx 43 GJs. The biological effects of FSH on cell morphology, differentiation, and proliferation of these cells were evaluated and correlated with alterations in the Cx 43 expression pattern. The Cx 43 phosphorylation was evaluated by immunoprecipitation and subsequent immunoblotting. Importantly, electrophysiological measurements via single-cell/whole-cell patch clamp enabled us to test for acute effects of FSH on GJ intercellular communication. MATERIALS AND METHODS

Chemicals

All chemicals for the cell culture were purchased from Sigma (Deisenhofen, Germany) unless otherwise indicated, and chemicals for buffers were from Merck (Darmstadt, Germany). Reagents for in situ hybridization were purchased from Boehringer Mannheim (Mannheim, Germany). Animals and Tissue Sections

Adult, cycling, female Sprague-Dawley rats (2–6 mo of age) were derived from the breeding colonies of the University of Ulm and Technische Universita¨t Mu¨nchen. Most of them had been used for previous studies, and all animal care requirements were fulfilled. No attempt was made to select animals from different phases of the estrus cycle. While under deep CO2 anesthesia, rats were decapitated, and the ovaries were removed and immersed into 4% paraformaldehyde. Frozen, 6-mm sections were cut, and sections from five rats were used for in situ hybridization as described elsewhere [14]. Cell Culture

The GFSHR-17 cells [18] were grown on 35- or 60-mm culture plates (for protein and RNA), on Labteks (for immunofluorescence), on 96-well plates (for proliferation assays; all three purchased by NUNC, Wiesbaden, Germany) or on cover glasses (for electrophysiology; Kindler, Freiburg, Germany). Cells were cultured with Dulbecco modified Eagle medium Ham F12 (1:1 v:v; Biochrom KG, Berlin, Germany) containing 5% fetal calf serum (Biochrom KG, Berlin, Germany). Cells were maintained in culture for up to 25 passages and were treated as indicated with porcine (p) FSH (0.5 IU/ml). Western Blot Analysis

Immunoblotting of protein from GFSHR-17 was performed as described elsewhere [15]. In brief, plates of confluent grown cells were rinsed two times in 10 mM Pipes buffer (pH 7.2) containing 150 mM NaCl and 1 mM EDTA. They were resolved in a 62.5 mM Tris buffer (pH 6.8) with 2% SDS and 10% saccharose and then homogenized by sonication. The lysate was assayed for protein (BCA; Pierce, Rockford, IL, USA). After boiling (1008C) for 5 min, samples were subjected to a 12.5% SDS-PAGE (15

mg of protein per lane) and separated electrophoretically. Proteins were blotted onto nitrocellulose (Schleicher und Schuell, Dassel, Germany) and incubated overnight with either a previously described Cx 43 antiserum (1:1000) [15] or a commercial monoclonal Cx 43 antibody (1:1000; Sigma) at 48C, followed by a peroxidase-conjugated goat secondary antibody (Dianova, Hamburg, Germany). Immunoreactivity was detected using an enhanced chemiluminescence kit (Amersham Buchler, Braunschweig, Germany).

Immunoprecipitation Studies

Immunoprecipitation of Cx 43 was performed as described elsewhere [14]. In brief, cells grown on plates were treated with or without pFSH, harvested, and homogenized (Ultra Turrax, IKA Labortechnik, Staufen, Germany). Membrane lysate was concentrated in a sucrose gradient. Membrane lysate in 10 mM PBS containing 0.5% deoxycholate (Na), 0.02% SDS, 1% polyethoxyethanol, 1 mM Na3VO4, 0.1% PMSF, and 30 ml/ml of aprotinin was incubated with protein A/sepharose for 2 to 4 h coupled to the same Cx 43 antiserum used for immunofluorescence. This antibody recognizes phosphorylated as well as nonphosphorylated Cx 43. Immunoprecipitated protein was electrophoretically separated by SDS-PAGE, and immunodetection was performed as described earlier using monoclonal anti-phosphoserine, -tyrosine, and -threonine (1:500; Sigma) and commercial anti-Cx 43 antibodies (1: 1000; Sigma). Signals obtained were digitized and analyzed densitometrically using a computer program created by A. Bulling and M. Rumitz (staff of the Anatomical Institute). Optical densities were statistically analyzed. Immunofluorescense and In Situ Hybridization

The method of immunofluorescence detection used has been described elsewhere [19]. In brief, cells were rinsed in 10 mM PBS (pH 7.4) and fixed in 4% paraformaldehyde. After fixation, cells were incubated with Cx 43 antiserum [15] and diluted 1:200 in a 0.02 M potassium PBS (pH 7.4) with 0.3% Triton X-100 and 2.5% normal goat serum overnight at 48C. Immunoreaction was detected with a fluorescein-coupled goat-anti-rabbit antibody (Dianova). Cells were mounted using the ProLong Antifade Kit (Molecular Probes, Go¨ttingen, Germany) according to manufacturer’s instructions, and immunoreaction was viewed using a Zeiss Axiovert 135 microscope (Jena, Germany) equipped with appropriate filter sets. Cryostat sections of rat ovaries were probed with a digoxygenin (DIG) uridine triphosphate-labeled rat Cx 43 cRNA antisense probe transcribed from a cDNA complementary to nucleotides 226–564 of the rat sequence as described elsewhere [14]. Proteinase K digestion was performed with a concentration of 0.25 mg/ml. Solution for hybridization contained 43 saline-sodium citrate (SSC), 5% dextran sulfate, 13 Denhardt solution, 50% deionized formamide, yeast-tRNA (0.25 mg/ml), and denatured Herrings sperm (0.5 mg/ml). Hybridization with Cx 43 cRNA was detected using an alkaline phosphatase-coupled, antiDIG-antibody (1:200), followed by visualizing with BCIP (5-bromo-4-chloro-3-indolyl-phosphate) and NBT (4-nitroblue-tetrazolium-chloride) dissolved in 50 and 70 mg/ml dimethylformamide, respectively. For control purposes, consecutive sections were incubated with a Cx 43 cRNA sense riboprobe.

GRANULOSA CELL Cx 43 AND REGULATION BY FSH

Northern Blot Analysis

Northern blot analysis was performed as described elsewhere [14]. Cellular RNA was extracted using the Rneasy kit (Quiagen, Hilden, Germany). A total of 10 mg of RNA per lane was electrophoretically separated on 1% formaldehyde agarose gels and blotted onto a nylon membrane (Amersham Buchler), followed by ultraviolet cross-linking. Prehybridization was performed for 2 h with 0.1% SDS containing 50% deionized formamide, 7.5% Denhardt solution, 4% SSC, 1 mM EDTA, and 10% ssDNA, at 608C. Hybridization was performed overnight at 608C with a 32Plabeled Cx 43 riboprobe transcribed from the same rat Cx 43 cDNA as for in situ hybridization. After hybridization, blots were washed five times in 0.1% SSC containing 0.1% SDS at 658C before autoradiography films (Hyper MP, Amersham Buchler) were developed after 1 to 3 days. In addition, blots were probed with a b-actin riboprobe to account for loading differences. The optical density of the signals was measured as described for immunoprecipitation. Electrophysiology: Whole-Cell Patch Clamp Recording

Junctional resistance between two adjacent cells coupled via GJs is a quantitative indicator of cell coupling and, thus, can be used to assess the degree of intercellular communication. As recently shown by Bigiani and Roper [20] and described by Postma et al. [21], the single-cell/whole-cell patch clamp technique can be applied in such a way that total membrane resistance reflects junctional resistance. For our study, we selected pairs of cells and patch clamped one of the two adjacent cells in the whole-cell configuration using an EPC 9 amplifier (HEKA Electronic, Lambrecht, Germany). Pipettes were pulled and fire polished in a DMZ Universal Puller (Zeitz, Augsburg, Germany) to obtain an approximate resistance of 5 to 6 MV in extracellular solution containing 140 mM NaCl, 3 mM KCl, 1 mM CaCl2, 10 mM HEPES, and 10 mM glucose (pH 7.4). Pipette solution (i.e., intracellular medium) contained 130 mM KCl, 10 mM HEPES, 5 mM EGTA, 3 mM CaCl2, and 0.85 mM MgCl2 (pH 7.4) with a calculated amount of 100 nM free Ca21. The patched cell was clamped to a potential near the measured resting membrane potential (245 6 15 mV, n 5 20 cells). We stimulated the cells with a sinusoidal voltage pulse of 270 Hz (amplitude 5 20 mV, offset 5 270 to 290 mV). The frequency of this pulse enabled us to assess the charge of the total membrane capacitance in response to the sine wave of the pulse. Any change in this calculated membrane capacitance implies changes in cellular coupling, with increasing membrane capacitance correlating with increased cellular coupling and vice versa. In addition, total membrane resistance was calculated, reflecting the junctional resistance of the cells. Because the values for cell capacitance and resistance were calculated by the software using a single-cell model, special attention had to be paid to the effects of different stimulation parameters on the measured values in case of cells coupled with the patch clamped cell. Simulations using a multicell circuit model showed that low stimulation frequencies would lead to marked changes in both capacitance and resistance values in the single-cell model when the coupling of cells changes. In contrast, changes of membrane resistance of the patched cell (i.e., because of activity of the ion channels) would only result in changes in the overall measured resistance, not in the total membrane capacitance. Using this technique, we tested the effects of pFSH (0.5 IU/ml) both alone

1663

and together with heptanol (5 mM) on GFSHR-17 cells. Results in this study were from eight experiments using eight pairs of cells, in which pFSH and heptanol were used subsequently. Stimulants were added via a flow system equipped with valves (Valvelink, Automate Scientific Inc., San Francisco, CA), by which the cells were constantly rinsed with the extracellular medium containing the respective stimulant. Proliferation Assays

Two different assays were used, the CellTiter 96 AQUeous One Solution Cell Proliferation Assay (Promega, Mannheim, Germany) and 3H-thymidine incorporation. For both methods, cells were plated on 96 wells (NUNC, Wiesbaden Germany). For the CellTiter 96 AQUeous One Solution Cell Proliferation Assay, cells were incubated with dimethylthiazolyl-carboxymethoxyphenyl-sulfophenyl-tetrazolium (MTS) and phenazine ethosulfate for 1 to 4 h. The MTS is reduced by viable cells to a formazan product, which was measured at 490 nm at 6, 8, 9, 14, and 24 h after the addition of FSH. Measurement of absorption was performed using an ELISA reader from DYNEX (Hamburg, Germany). A calibration curve of different cell densities was performed for every 96-well plate. In addition, 3H-thymidine incorporation in cellular DNA was measured as described elsewhere [22]. In brief, 3Hmethylthymidine (NEN, Ko¨ln, Germany) was diluted in RPMI medium (Biochrom KG, Berlin Germany) to obtain 50 mCi/ml, and then 20 ml of this solution were added to the cells for 6 h. Cells were subsequently harvested and lysed, and cellular DNA was precipitated on glass-fiber filters. The radioactivity of the filter pads, indicating 3H-thymidine incorporation, was counted using a Matrix 96 Scintillation Counter (Packard Instruments, Meriden, CT). Progesterone Assay

The content of progesterone in the culture medium was measured using a progesterone ELISA assay (Biochem Immunosystems, Freiburg, Germany) according to the manufacturer’s instructions, as described elsewhere [19]. Statistics

Results of proliferation assays and analyses of optical densities of Northern blots and immunopreciptation experiments were statistically analyzed using t-tests. RESULTS

Rat Ovarian Follicular GC and GFSHR-17 Cells Contain Cx 43

In accordance with the results of previous studies [5, 12– 15], in situ hybridization of rat ovaries showed that Cx 43 is readily detectable in the GCs of follicles (Fig. 1). The GCs in larger (i.e., antral follicles) showed a more intense labeling compared with that in smaller, preantral ones. Thus, we hypothesized that Cx 43 expression may be subject to regulation by FSH. To address this point, we studied the effect of FSH on a rat GC line stabily transformed to express the FSH-receptor GFSHR-17 [18]. Immunofluorescense (Fig. 2A), immunoblot (Fig. 2B), and Northern blot analyses (Fig. 2C) showed that GFSHR-17 cells also contain Cx 43 protein and its mRNA. Thus, GFSHR-17 are a valid and useful model for the purpose of our study.

1664

SOMMERSBERG ET AL.

FIG. 1. Detection of Cx 43 mRNA in the GC layer of rat ovarian follicles. In situ hybridization of rat ovaries with Cx 43 antisense riboprobe (A) and with Cx 43 sense riboprobe (control; B). The comparison of small follicles with larger follicles suggests an increase of Cx 43 expression in the GCs during follicular development. AF, Antral follicle; PF, preantral follicle.

FSH Affects GJ Communication and Cx 43 Phosphorylation of GFSHR-17 Cells Evaluation of GJ communication. As reported in a preliminary study [23], double-cell/whole-cell patch clamp showed that GFSHR-17 cells are functionally coupled. To study the acute effects of FSH on GJ coupling, we used a

derived model of the single-cell/whole-cell patch clamp technique. Figure 3 depicts a representative experiment in which FSH caused a reversible decrease in total membrane resistance Rm. Specifically, the decrease in eight independent experiments with eight different cell pairs ranged from 8% to 55% relative to the median value of Rm (Table 1). To show that the measured effects resulted from changes in GJ coupling, we subsequently tested heptanol, which is a substance known to disrupt GJ channels, on each cell pair examined. Addition of heptanol (5 mM) caused an increase in Rm and, as a consequence of decreased GJ communication, a decrease in total membrane capacitance Cm (data not shown). Furthermore, we avoided bias in the measurements from changes of flow rate by switching channels and testing the buffer without FSH. Phosphorylation of Cx 43 in GFSHR-17 cells. Increased phosphorylation of Cx 43 at serine, threonine, and tyrosine residues has been proposed to be involved in the gating mechanism of GJ channels [24]. Thus, we examined the phosphorylation state of Cx 43 within 10 min after the addition of FSH. Immunoprecipitated Cx 43 was immunoblotted and developed with anti-phosphoserine, -threonine, and -tyrosine antibodies, respectively (Fig. 4). Phosphorylation of Cx 43 serine residues did not consistently change after FSH treatment. Thus, in seven experiments, we found no change in four cases but decreases of 63%, 68%, and 15% in three cases. Tyrosine residues of Cx 43 were, however, clearly dephosphorylated after addition of FSH. The FSH statistically significantly (P , 0.05, paired t-test) dephosphorylated Cx 43 tyrosine in seven independent experiments to 92%, 63%, 75%, 16%, 51%, and 76% of the respective control values; in one experiment, no change was observed. Furthermore, phosphorylation at threonine residues of Cx 43 decreased statistically significantly in response to FSH (n 5 8) to 95%, 88%, 83%, 25%, 52%, 48%, 68%, and 80% of control values (P , 0.05, paired ttest).

FIG. 2. Expression of Cx 43 in GFSHR-17 cells. A) Immunofluorescense detection of Cx 43 in GFSHR-17 cells. Note the typical dotted, plaque-like staining of the GJ protein and the predominant location at cell borders. B) Western blot. The anti-Cx 43 antiserum recognizes three immunoreactive bands around 43 000 Da. The two upper bands represent the phosphorylated forms of Cx 43 because of its lower electrophoretic motility, as described elsewhere [15]. C) Northern blot. A single Cx 43 transcript of approximately 3.6 kilobases is identified. The positions of the 18S and 28S RNAs are indicated by bars. Experiments with similar results were repeated five times (Northern blot) or four times (Western blots).

1665

GRANULOSA CELL Cx 43 AND REGULATION BY FSH

TABLE 1. Relative decrease of total membrane resistance (Rm) as measured in eight experiments.a Experiment 1 2 3 4 5 6 7 8

Relative decrease of Rm 8% 29.5% 14.4% 15.6% 27% 20.5% 39% 55%

a The median of R during the first 29 sec, when cells were only rinsed m with buffer, was calculated. The minima of decrease during the consecutive period of stimulation (30 sec) were subtracted from the median. In three of eight experiments, this decrease was preceded by a short peak increase of Rm (11%–13%, n 5 3).

Regulation of Cx 43 mRNA in GFSHR-17 Cells

Northern blotting of GFSHR-17 mRNA revealed that FSH increased Cx 43 mRNA levels approximately twofold after 6 to 9 h. A summary of three experiments is shown in Figure 5; note that two more experiments, with an increase of 270% or a marginal decrease of 14%, were excluded. The effect of FSH on Cx 43 mRNA levels was less pronounced when GFSHR-17 cells were treated with FSH for a longer period (24 h, data not shown). Correlation of Cx 43 Levels with Cell Proliferation and Differentiation of GFSHR-17 Cells FIG. 3. Rapid effects of FSH on cellular coupling. A) Scheme of the electrical circuit required for recording in the whole-cell mode. Two adjacent cells, each with its specific membrane resistance (R1 and R2) and capacitance (C1 and C2), are coupled via GJ channels that can be described as junctional resistance (RJ). Within this system, a sine pulse, consisting of an offset of 260 mV and an amplitude of 20 mV, is administered to a single patched cell. The subsequent changes in the current amplitude, phase, and offset are measured, and serial resistance (Rs), total membrane resistance (Rm), and total membrane capacitance (Cm) are calculated. Simulation of the electrical circuit shows that changes in Rm and Cm reflect changes in RJ. B) A typical observation of the progression of Rm measured in a cell pair in response to FSH. In this case, administration of FSH caused a small, transient increase in Rm that was followed by a large, reversible decrease in Rm. Similar results could be reproduced in eight independent experiments (Table 1).

Addition of FSH to GFSHR-17 induced a statistically significant inhibition of cell proliferation (Fig. 6) observed 9 h after addition, as shown by the results of two different assays. Results of 3H-thymidine incorporation showed that proliferation was inhibited by 36% by FSH compared with nontreated cells (n 5 14, P , 0.05). Reduction of MTS through viable cells was also inhibited by FSH by 34%, 42%, and 45% within 6, 8, and 9 h, respectively, after the addition of FSH (n 5 14 and P , 0.05 for each experiment). Furthermore, addition of FSH caused morphological changes of GFSHR-17 that are typical for GC differentiation. Cells became round, detached from the plate, and formed three-dimensional cell clusters (Fig. 7). These morphological differentiation criteria could first be seen 3 h

FIG. 4. Effects of FSH on Cx 43 phosphorylation. Untreated cells and cells treated with FSH (0.5 IU/ml) for 10 min were used for immunoprecipitation of Cx 43. Upper panels show typical results for each antibody. Lower panels summarize results from densitometric analysis of all experiments. The columns represent means 1 SEM. Untreated controls (Co) were assigned a value of 1.00. Number of experiments (different wells) are given inside the columns. OD, Optical density (arbitrary units). ,P , 0.05 (paired t-test).

1666

SOMMERSBERG ET AL.

FIG. 5. FSH-dependent regulation of Cx 43 mRNA expression in GFSHR-17. This figure shows the increase in arbitrary optical density (OD) of Cx 43 mRNA bands in GFSHR-17 cells treated without (Co) or with FSH for 9 h. To account for small loading differences, results were normalized using the signal observed for b-actin in the same Northern blots. ,P , 0.05 (paired t-test, n 5 3).

after addition of FSH, and they persisted throughout the entire observation period up to 24 h. FSH also induced progesterone synthesis. Thus, 24 h after the addition of FSH to GFSHR-17, accumulation of progesterone could be detected in the medium (ø0.5 ng per 60-mm plate, n 5 5 wells). In untreated GFSHR-17 cells, which served as controls, progesterone levels remained below the detection limit of the assay. DISCUSSION

The present study provides novel evidence for a direct role of the gonadotropin FSH in the regulation of GJ communication, Cx 43 phosphorylation, and Cx 43 gene expression in a GC-derived ovarian cell line expressing the FSH receptor. Circumstantial evidence previously suggested involvement of FSH in Cx 43 gene expression in ovarian follicles, because the GCs of small follicles isolated from sheep ovary contained lower amounts of immunoreactive Cx 43 compared with GCs from larger follicles, which grow in a FSH-dependent way [16]. Also, the typical plaque-like, membrane-associated Cx 43 immunoreactivity was increased in larger follicles [15]. Importantly, treatment of rats with eCG increased Cx 43 expression in the ovary [13]. Taken together, these reports suggest either a direct or an indirect role of FSH in Cx 43 regulation. Our results obtained in vitro (in GFSHR-17 cells) are in agreement with this assumption, because FSH likewise in-

FIG. 6. FSH inhibits proliferation of GFSHR-17 after 9 h. Specifically, FSH statistically significantly inhibited proliferation of GFSHR-17 as measured 9 h after addition of FSH by 3H-thymidine incorporation. Statistical analysis was performed using paired t-test. The columns represent means 1 SEM. Number of experiments are given inside the columns. ,P , 0.01 (paired t-test).

FIG. 7. FSH affects cell morphology of GFSHR-17 cells. Phase-contrast microscopic view of GFSHR-17 cells without (A) and after (B) addition of FSH (0.5 IU/ml for 3 h). Note that GFSHR-17 responded to FSH by cell rounding and formation of cell clusters. This was a robust response observed in basically every experiment.

creased Cx 43 mRNA and protein levels (data not shown). This result indicates that where Cx 43 is concerned, GCs in the follicle and GFSHR-17 cells respond to FSH in a similar way. Importantly, our present results also allow the novel conclusion that FSH, rather than FSH-dependent products or other factors in the follicular fluid, is able to directly regulate Cx 43 and GJs. Because of the complex situation in vivo, with follicular GCs being exposed to a cocktail of potent factors, including numerous growth factors [14] and estrogen, results of previous studies failed to clearly pinpoint such a direct effect of FSH. Estrogens, for example, produced in an FSH-dependent manner by GCs have been shown to affect the expression of Cx 43 in other tissues (e.g., myometrium) [25]. The situation in the follicle in vivo may be even more complex, because interference of assumed direct/indirect FSH effects with other factors in the follicular fluid cannot be excluded. To specifically examine the possibility of a direct effect of FSH on Cx 43 and GJ, we chose GFSHR-17 cells [18], which derive from preovulatory rat follicles and stabily express the rat FSH receptor. The GFSHR-17 cells form GJs and also possess Cx 43; Cx 26, Cx 32, and Cx 37 were not detected (unpublished results). Thus, they are excellent models for the purpose of this study. Use of GFSHR-17 also allowed us to rule out a possible side effect of estrogens on Cx 43 regulation, because in contrast to primary GCs, these cells lack aromatase activity and do not produce estrogens (unpublished results). However, GFSHR-17 cells produce progesterone when stimulated with pFSH [18]. This may be relevant, because progesterone may also have effects on Cx 43 gene expression in typical target tissues bearing progesterone receptors [26]. In GFSHR-17 cells, we could not detect immunoreactive progesterone receptors by immunoblotting and immunocytochemistry with a monoclonal antibody (data not shown). Nevertheless, we selected for our study a low concentration of pFSH (0.5 IU/ml) that still causes visible changes in GFSHR-17 cells (i.e., induction of individual cell rounding and growth in clusters clearly observed within 3 h after FSH treatment) to minimize any possible involvement of progesterone. That progesterone production was indeed minimal under these conditions was shown by our measurements. However, FSH was biologically active, because in addition to the FSH-induced cell rounding and cluster formation in GFSHR-17 cells, cellular proliferation decreased significantly within 9 h after FSH addition. This indicates that the well-established function of FSH to promote differentiation of GCs is conserved in GFSHR-17 cells.

GRANULOSA CELL Cx 43 AND REGULATION BY FSH

In our study, we found that FSH strongly affected GJ communication in GFSHR-17 in a direct manner. A direct promoting effect on GJ coupling became evident in our patch clamp experiments. Within seconds, FSH decreased the junctional resistance between cell pairs, which is indicative of improved GJ transfer of ions. In contrast, cells did not change their resistance when buffer alone was used, whereas subsequent application of a disrupting agent of GJs, heptanol [27], used for control purposes increased the resistance irreversibly. The molecular events leading to different gating of GJ channels in GFSHR-17 are not known. In general, gating is regulated by various parameters, including intracellular Ca21, pH, and phosphorylation of Cx [28]. Some evidence exists that FSH may affect intracellular Ca21 levels in Sertoli cells [29], but we did not observe any increase of intracellular Ca21 when FSH was added to GFSHR-17, as determined by single-cell Ca21 measurement using FURA-2 (unpublished results). Several phosphorylation sites of the Cx 43 protein at serine, threonine, and tyrosine residues have been described [30], and we found that phosphorylation of Cx 43 was altered after treatment with FSH. In particular, phosphorylation at tyrosine and threonine significantly decreased, whereas phosphorylation at serine was not significantly altered or decreased in some experiments. Thus, the overall phosphorylation state of Cx 43 decreased or, in some cases, was unaffected. This is important considering that for Cx 43, increased phosphorylation is reported to be associated with lower unitary conductance in cardiomyocytes [24]. That increased serine phosphorylation leads to GJ communication breakdown was previously described after treatment of theca cells with nerve growth factor [14]. Likewise, epidermal growth factor has been shown to attenuate GJ communication via increased phosphorylation of Cx 43 at serine sites [14]. Moreover, evidence exists that increased phosphorylation at tyrosine residues is associated with reduced GJ communication [21, 31]. The opposite effects, namely reduced phosphorylation at threonine and tyrosine residues and increased GJ communication, as found in our present study are, therefore, in agreement with these reports. Our results suggest the following sequence of events induced by FSH in GFSHR-17 cells: FSH acutely regulates gating of existing GJ and increases GJ communication. This involves decreased phosphorylation of Cx 43. In addition, FSH induces differentiation of cells, slows down proliferation, and simultaneously increases Cx 43 gene expression, most likely leading to increased GJ coupling in differentiated cells. The GJs are thought to be essential for cell-to-cell communication within the GC layer of the ovarian follicle, and active communication may determine the fate of the follicle. Our results indicate that FSH, which regulates follicular growth, directly regulates GJs in the GC-derived cell line of GFSHR-17 cells, which share important traits with GCs in vitro. The experimental model used and the results obtained in the present study, therefore, provide the basis for future studies that can address both whether and how steroids and/or growth factors are involved in the regulation of Cx 43 gene expression, Cx 43 phosphorylation, and overall gating of GJ channels in cells of a follicle. ACKNOWLEDGMENTS The authors would like to thank M. Rumitz and Dr. J. Grosse for their help.

1667

REFERENCES 1. McGee EA, Perlas E, LaPolt PS, Tsafriri A, Hsueh AJ. Follicle-stimulating hormone enhances the development of preantral follicles in juvenile rats. Biol Reprod 1997; 57:990–998. 2. Amsterdam A, Hosokawa K, Yoshida Y, Sasson R, Jung Y, Kotsuji F. Cross talk among multiple signaling pathways controlling ovarian cell death. Trends Endocrinol Metab 1999; 10:255–262. 3. Amsterdam A, Selvaraj N. Control of differentiation, transformation, and apoptosis in granulosa cells by oncogenes, oncoviruses, and tumor suppressor genes. Endocr Rev 1997; 18:435–461. 4. Goodenough DA, Goliger JA, Paul DL. Connexins, connexons, and intercellular communication. Annu Rev Biochem 1996; 65:475–502. 5. Grazul-Bilska AT, Reynolds LP, Redmer DA. Gap junctions in the ovaries. Biol Reprod 1997; 57:947–957. 6. Zampighi GA, Loo DD, Kreman M, Eskandari S, Wright EM. Functional and morphological correlates of connexin50 expressed in Xenopus laevis oocytes. J Gen Physiol 1999; 113:507–524. 7. Manthey D, Bukauskas F, Lee CG, Kozak CA, Willecke K. Molecular cloning and functional expression of the mouse gap junction gene connexin-57 in human HeLa cells. J Biol Chem 1999; 274:14716– 14723. 8. Itahana K, Tanaka T, Morikazu Y, Komatu S, Ishida N, Takeya T. Isolation and characterization of a novel connexin gene, Cx-60, in porcine ovarian follicles. Endocrinology 1998; 139:320–329. 9. Simon AM, Goodenough DA, Li E, Paul DL. Female infertility in mice lacking connexin 37. Nature 1997; 385:525–529. 10. Khan-Dawood FS, Yang J, Dawood MY. Expression of gap junction protein connexin-43 in the human and baboon (Papio anubis) corpus luteum. J Clin Endocrinol Metab 1996; 81:835–842. 11. Juneja SC, Barr KJ, Enders GC, Kidder GM. Defects in the germ line and gonads of mice lacking connexin43. Biol Reprod 1999; 60:1263– 1270 12. Wiesen JF, Midgley AR Jr. Changes in expression of connexin 43 gap junction messenger ribonucleic acid and protein during ovarian follicular growth. Endocrinology 1993; 133:741–746. 13. Granot I, Dekel N. Cell-to-cell communication in the ovarian follicle: developmental and hormonal regulation of the expression of connexin43. Hum Reprod 1998; 13(suppl 4):85–97. 14. Mayerhofer A, Dissen GA, Parrott JA, Hill DF, Mayerhofer D, Garfield RE, Costa ME, Skinner MK, Ojeda SR. Involvement of nerve growth factor in the ovulatory cascade: trkA receptor activation inhibits gap junctional communication between thecal cells. Endocrinology 1996; 137:5662–5670. 15. Mayerhofer A, Garfield RE. Immunocytochemical analysis of the expression of gap junction protein connexin 43 in the rat ovary. Mol Reprod Dev 1995; 41:331–338. 16. Huet C, Monget P, Pisselet C, Hennequet C, Locatelli A, Monniaux D. Chronology of events accompanying follicular atresia in hypophysectomized ewes. Changes in levels of steroidogenic enzymes, connexin 43, insulin-like growth factor II/mannose 6 phosphate receptor, extracellular matrix components, and matrix metalloproteinases. Biol Reprod 1998; 58:175–185. 17. Granot I, Dekel N. Developmental expression and regulation of the gap junction protein and transcript in rat ovaries. Mol Reprod Dev 1997; 47:231–239. 18. Keren TI, Dantes A, Sprengel R, Amsterdam A. Establishment of steroidogenic granulosa cell lines expressing follicle stimulating hormone receptors. Mol Cell Endocrinol 1993; 95:R1–R10. 19. Mayerhofer A, Hemmings HCJ, Snyder GL, Greengard P, Boddien S, Berg U, Brucker C. Functional dopamine-1 receptors and DARPP-32 are expressed in human ovary and granulosa luteal cells in vitro. J Clin Endocrinol Metab 1999; 84:257–264. 20. Bigiani A, Roper SD. Estimation of the junctional resistance between electrically coupled receptor cells in Necturus taste buds. J Gen Physiol 1995; 106:705–725. 21. Postma FR, Hengeveld T, Alblas J, Giepmans BN, Zondag GC, Jalink K, Moolenaar WH. Acute loss of cell-cell communication caused by G protein-coupled receptors: a critical role for c-Src. J Cell Biol 1998; 140:1199–1209. 22. Bendigs S, Salzer U, Lipford GB, Wagner H, Heeg K. CpG-oligodeoxynucleotides co-stimulate primary T cells in the absence of antigen-presenting cells. Eur J Immunol 1999; 29:1209–1218. 23. Sommersberg B, Bulling A, Ngezahayo A, Amsterdam A, Dantes A, Garfield RE, Kolb H, Mayerhofer A. FSH regulates gap-junctional communication and connexin 43 gene expression in rat granulosa cells

1668

24.

25.

26. 27.

SOMMERSBERG ET AL.

(GFSHR-17). In: Program and Abstracts, The Endocrine Society’s 81st Annual Meeting; June 12–15, 1999; San Diego, CA. p. 245. Moreno AP, Saez JC, Fishman GI, Spray DC. Human connexin43 gap junction channels. Regulation of unitary conductances by phosphorylation. Circ Res 1994; 74:1050–1057. Grummer R, Traub O, Winterhager E. Gap junction connexin genes cx26 and cx43 are differentially regulated by ovarian steroid hormones in rat endometrium. Endocrinology 1999; 140:2509–2516. Yu W, Dahl G, Werner R. The connexin43 gene is responsive to oestrogen. Proc R Soc Lond B Biol Sci 1994; 255:125–132. Peracchia C. Effects of the anesthetics heptanol, halothane and isoflurane on gap junction conductance in crayfish septate axons: a calcium- and hydrogen-independent phenomenon potentiated by caffeine

28. 29.

30. 31.

and theophylline, and inhibited by 4-aminopyridine. J Membr Biol 1991; 121:67–78. Bruzzone R, White TW, Paul DL. Connections with connexins: the molecular basis of direct intercellular signaling. Eur J Biochem 1996; 238:1–27. Lalevee N, Rogier C, Becq F, Joffre M. Acute effects of adenosine triphosphates, cyclic 39,59-adenosine monophosphates, and folliclestimulating hormone on cytosolic calcium level in cultured immature rat Sertoli cells. Biol Reprod 1999; 61:343–352. Dhein S. Gap junction channels in the cardiovascular system: pharmacological and physiological modulation. Trends Pharmacol Sci 1998; 19:229–241. Goldberg GS, Lau AF. Dynamics of connexin43 phosphorylation in pp60v-src-transformed cells. Biochem J 1993; 295:735–742.