| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Anatomical Institute, University of Munich, D-80802 Munich, Germany
Address all correspondence and requests for reprints to: Lars Kunz, Anatomical Institute, University of Munich, Biedersteiner Str. 29, D-80802 Munich, Germany. E-mail: lars.kunz{at}lrz.uni-muenchen.de.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Objective: We studied whether KATP channels are present in the human ovary and luteinized granulosa cells (GCs). Human GCs were examined regarding functionality and physiological role of the channel.
Patients and Intervention: Human GCs were obtained from in vitro fertilization patients.
Results: KATP channels are involved in membrane potential generation in human GCs because application of the KATP blocker glibenclamide resulted in depolarization as monitored by fluorescence microscopy. Furthermore, glibenclamide significantly attenuated human chorionic gonadotropin-induced progesterone production. The channel pore is composed of Kir6.1, but not Kir6.2, as indicated by RT-PCR. Kir6.1 subunit protein was detected in human follicular and luteal cells by immunohistochemistry and localized to the plasma membrane of human GCs by immunogold staining. RT-PCR experiments revealed the sulfonylurea receptor subunit SUR2B as part of the KATP channel. In addition, mRNAs encoding SUR1 and SUR2A were detected in some preparations. There is no evidence for mitochondrial KATP channels in human GCs because we detected neither Kir6.1 protein in mitochondrial membranes nor alterations of mitochondrial membrane potential by glibenclamide or the KATP opener diazoxide.
Conclusions: Endocrine cells of the human ovary possess functional KATP channels, which are linked to both plasma membrane potential generation and progesterone production. Our results may provide new insights into human ovarian physiology and raise the possibility of pharmacological targeting.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
KATP channels are targets for many drugs, which bind to the SUR subunits (2, 5, 10, 11, 12). The sensitivity of KATP channels to different substances varies, depending on the type of SUR subunit. KATP channels are inhibited by sulfonylureas such as the oral antidiabetic glibenclamide resulting in depolarization of the plasma membrane of the pancreatic ß-cell. On the other hand, they are activated by potassium channel openers (KCOs) such as diazoxide. KCOs hyperpolarize the plasma membrane and reduce cellular excitability of vascular smooth muscle cells and are therefore potential antihypertensives.
Recently mRNA for Kir 6.1/SUR 2B was detected in rat corpus luteum (CL) and placenta (13). As for primates, ovarian KATP channels are not yet described; we intended to study this channel in a human system. Other types of ion channels have been reported to be present and of physiological relevance in endocrine ovarian cells of the human (14, 15, 16, 17, 18) and other mammals (19, 20, 21, 22, 23). To study the ovarian KATP channel in the human, we used cultured luteinizing granulosa cells (GCs) from patients undergoing in vitro fertilization and human ovarian tissue sections. We applied cell and molecular biological methods to characterize the KATP channel and elucidate its subunit composition, role in steroidogenesis, and subcellular localization. Functionality of KATP channels was studied by monitoring membrane potential changes caused by channel modulation.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human GCs were isolated from follicular aspirates of women undergoing in vitro fertilization (Assisted Reproductive Technologies Bogenhausen, Munich, Germany). Experiments on these cells were approved by the local ethics committee and the patients. The cells were cultured in DME/F12 (10% fetal calf serum; Sigma-Aldrich, Munich, Germany) under a humidified atmosphere at 37 C-5% CO2 for at least 3 d before the experiments were performed (17).
Tissue samples
Human CL tissue samples were provided by consenting patients undergoing gynecological surgery (Frauenklinik, Klinik am Eichert, Göppingen, Germany). After fixation in Bouin’s fixative, the tissue samples were embedded in paraffin. Apart from this, we used paraffin-embedded ovarian samples, which contained follicles, from the tissue archive of the Women’s Hospital in Munich, which had been taken from premenopausal women during autopsies. All procedures concerning use of human materials were approved by the respective local ethics committees.
Chemicals and solutions
Glibenclamide and diazoxide were dissolved in dimethyl sulfoxide (DMSO; all from Sigma-Aldrich) and diluted into cell culture medium to their final concentration. Extracellular (EC) solution contained 140 mM NaCl, 3 mM KCl, 1 mM CaCl2, 1 mM MgCl2, 10 mM HEPES, and 10 mM glucose (pH 7.4).
Progesterone assay
Human GCs were cultured in 24-well plates. On d 3 they were treated in triplicates with the respective compounds. After 24 h the supernatants were collected and the progesterone concentrations were measured using an ELISA test (DRG Instruments, Marburg, Germany). Mean values were normalized to the untreated control value, and the normalized mean values were averaged over several independent experiments (15).
Cell morphology and cytotoxicity assay
To assess morphological changes, human GCs were cultured in 35-mm plastic dishes and treated with different stimulants for 24 h on d 3 of culture. The cells were fixed with 2.5% glutaraldehyde in 100 mM sodium cacodylate buffer (pH 7.4), postfixed with 1% osmium tetroxide/potassium hexacyanoferrate (II) and embedded in Epon. Semithin sections (1 µm) were cut and inspected by light microscopy after staining with Azure II/methylene blue (1:1). Ultrathin sections were cut and treated with lead citrate (2.7%)/uranyl acetate (2%) and analyzed by electron microscopy (EM; EM10, Carl Zeiss, Göttingen, Germany). Potential cytotoxicity of glibenclamide was further evaluated by using a commercial nonradioactive cell proliferation 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium assay (Promega, Mannheim, Germany). A total of three wells per treatment were incubated for either 24 h on d 3 of culture or from the first day on for 3 and 6 d, respectively.
RT-PCR
Total RNA was isolated from several human GC preparations and reverse transcribed to identify the expression of messenger RNA coding for Kir 6.x and SUR subunits (24). Human ovary, heart, smooth muscle, and pancreas cDNA (BD Clontech, Palo Alto, CA) served as a positive control as PCR amplification was carried out with oligodeoxynucleotide primer pairs corresponding to human sequences (Table 1
) (9), which spanned at least one intron of the genomic sequence (except for Kir 6.2). The PCR protocol for both PCR and nested PCR consisted of 35 cycles of denaturation at 94 C (60 sec), annealing at 55 C (120 sec), and extension at 72 C (180 sec) using a PTC-200 Peltier thermal cycler (MJ Research, Watertown, MA). PCR products were separated on an agarose gel and visualized by ethidium bromide staining and UV illumination. All products were sequenced and were found to be identical with published sequences.
|
Immunohistochemical methods were used to examine the localization of Kir 6.1 subunit protein. Deparaffinized human ovarian sections (5 µm) were subjected to an additional microwave treatment for improved antigen retrieval (25) and treated with 3% H2O2 in methanol to block endogenous peroxidase. Then slices were incubated overnight at 4 C with Kir 6.1 antiserum (rabbit antirat; 1:2.000; containing 5% normal goat serum), which was a kind gift of W. A. Coetzee (New York University School of Medicine, New York, NY) (26, 27). The next day the slices were incubated with goat antirabbit antibody (1:500) and immunoreactivity was visualized by the avidin biotinylated enzyme complex/diaminobenzidine staining reaction (Vectastain Elite kit, Vector Laboratories, Burlingame, CA) (24, 25). To assure the specificity of the immunoreaction the first antiserum was either preadsorbed with a specific peptide antigen, or replaced by normal rabbit serum, or completely omitted. Kir 6.2 (rabbit antirat; Alomone Labs, Jerusalem, Israel) and SUR 2B (goat antihuman; Santa Cruz Biotechnology, Santa Cruz, CA) antisera were used in different concentrations.
Immunocytochemistry
Human GCs grown on glass were fixed with 4% paraformaldehyde [0.01 M PBS (pH 6.8)], washed with PBS, and then incubated with the Kir 6.1 antiserum (1:500) at 4 C overnight. Then cells were incubated with a fluorescein-conjugated goat antirabbit antibody (1:200) in the dark, washed and finally covered using ProLong antifade kit (Molecular Probes, Eugene, OR) (24). Subsequently fluorescence was imaged using a confocal microscope (Leica Microsystems, Wetzlar, Germany). Controls performed were the same as for immunohistochemistry.
Immunoelectron microscopy
The intracellular localization of the Kir 6.1 protein was studied by immunoelectron microscopy as previously described (28). Briefly, human GCs were embedded in Lowicryl (K4M, Polysciences Europe, Eppelheim, Germany) and sectioned for EM. The EM grids were incubated with the Kir 6.1 antiserum (1:1.500) buffer at 4 C overnight. Thereafter the grids were incubated for 2–4 h with a secondary gold-labeled antiserum (1:20; diameter of gold particles: 10 nm; Aurion, Wageningen, The Netherlands). Finally, they were washed, fixed with 2% glutaraldehyde in PBS, and analyzed by EM (EM10; Carl Zeiss).
Monitoring of plasma membrane potential
To measure the functional activity of KATP channels in the plasma membrane, we monitored changes in the plasma membrane potential using the bisoxonol dye bis-(1,3-dibutylbarbituric acid) trimethine oxonol [DiBAC4(3), Molecular Probes] (29, 30, 31). This anionic fluorescent dye is distributed across the plasma membrane, depending on the membrane potential. Cultured human GCs (d 5–10) on glass coverslips were put into a recording chamber mounted on a TCS SP2 confocal microscope (Leica Microsystems). DiBAC4(3) (500 nM in EC solution) was applied for about 15 min to ensure dye distribution across the cell membrane. Afterward the KATP inhibitor glibenclamide [10 µM in EC with 500 nM DiBAC4(3)] was applied. Changes in fluorescence intensity were monitored for 30 min by sampling every 5 sec at excitation and emission wavelengths of 488 and 520 ± 20 nm, respectively.
Detection of mitochondrial membrane potential
Human GCs grown on coverslips were loaded with 5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolylcarbocyanine iodide (DePsipher; R&D Systems, Minneapolis, MN) in culture medium for 30 min at 37 C/5% CO2. Cells were washed with EC solution and transferred into a microscopical recording chamber (see paragraph above) containing EC solution. Because the cationic dye exhibits potential-dependent accumulation in mitochondria, fluorescence emitted on excitation at 488 nm was monitored every 10 sec. In one channel the punctuate orange-red fluorescence (605 ± 15 nm) of dye aggregates in mitochondria with normal polarization was monitored, whereas in the other channel, the diffuse green monomer fluorescence (530 ± 15 nm) from depolarized mitochondria was recorded (32, 33). The intensities were quantified over single cells, and changes in mitochondrial membrane potential were assessed from alterations in the red to green fluorescence intensity ratio. Real-time changes were monitored during application of glibenclamide (10 µM) and diazoxide (100 µM). In the end, the protonophore and uncoupler of oxidative phosphorylation carbonyl cyanide 3-chlorophenylhydrazone (CCCP; Sigma-Aldrich) was applied to achieve complete depolarization.
Data analysis
Data were analyzed and depicted by using Prism 4 (GraphPad Software, San Diego, CA). Progesterone production data were statistically analyzed by a repeated-measures ANOVA followed by Newman-Keuls multiple comparison posttest. For data of membrane potential measurements, a one-sample t test was used to test for significant alterations.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Progesterone production by human GCs is known to be stimulated by hCG (Fig. 1; repeated-measures ANOVA, P < 0.0001; posttest, P < 0.001). This elevation over basal production levels was diminished by the specific KATP channel blocker glibenclamide (10 µM; posttest, P < 0.01), which alone had no effect on basal production (posttest, P > 0.05). Thus, human GCs possess a KATP channel, and its opening is required during the hCG-mediated mechanisms stimulating progesterone production. There is also a KATP-independent component of hCG-stimulated progesterone production because the value for glibenclamide alone was significantly different from the one for hCG + glibenclamide (posttest, P < 0.05). Because glibenclamide had to be dissolved in DMSO, the solvent was added to all solutions used for treatment in a final concentration of 0.1% (vol/vol). However, DMSO did not reduce the basal or the hCG-induced progesterone production. In fact, it even increased it by 10% (basal) and 29% (hCG-induced; results not shown). Thus, the attenuating effect of glibenclamide on hCG-induced progesterone production can be unambiguously assigned to the KATP blocker. To detect potential toxic effects of the substances or the solvent DMSO, we examined cell morphology after 24 h treatment by light and electron microscopical inspection of sections. The cells did not show any obvious differences in their appearance in comparison with the untreated control cells (not shown). Furthermore, cell number as assessed by 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium assay was not affected by long-term treatment with 10 µM glibenclamide (for 3 and 6 d, respectively) or 24 h treatment in parallel to hCG on d 3 of culture (not shown).
|
RT-PCR analysis was used to examine the details of KATP subunit expression, and we detected mRNA coding for Kir 6.1 (Fig. 2) in human GCs, whereas mRNA for Kir 6.2 was not found (data not shown). Of the three possible sulfonylurea-sensitive subunits, SUR 2B (Fig. 2) was always present, whereas SUR 1 and SUR 2A were also detected in some preparations (data not shown; n = 5 for each). We obtained PCR products of expected size and correct sequence with all primer pairs when used on respective-positive control tissues (data not shown).
|
), whereas Kir 6.2 protein was not found (data not shown). Human GCs were stained evenly and a certain accumulation of the dye at the outer parts of the cells was observed. To test for localization of Kir 6.1 protein in the plasma membrane, transmission electron microscopy on human GCs was applied. Using gold-labeled secondary antibodies, we found a distinct plasma membrane localization (Fig. 3D), whereas gold staining of the cytoplasm and cell organelles was at background level.
|
) and in GCs of large follicles (Fig. 4B) as well as vascular smooth muscle cells (Fig. 4A). All controls including preadsorption of Kir 6.1-antiserum antigen were negative (Fig. 4D). In accordance with our RT-PCR results, Kir 6.2 protein was not detected (data not shown).
|
). KATP channels are not present in mitochondrial membranes
We studied the presence of mitochondrial KATP (mitoKATP) channels to exclude their possible contribution to the observed reduction of progesterone production by glibenclamide. Although gold immunostaining had not indicated the presence of Kir 6.1 protein in mitochondria, functional mitoKATP channels of partially unknown composition are reported in several cell types, e.g. cardiomyocytes (34). Therefore, we monitored alterations of the mitochondrial membrane potential by means of the fluorescence probe DePsipher. Opening of mitoKATP channel would result in a K+ influx into the mitochondrion from the cytoplasm and would thereby cause a depolarization. Application of diazoxide and glibenclamide had no effect on the green to red fluorescence intensity ratio, i.e. the mitochondrial membrane potential remained unchanged (Fig. 5; n = 5). Decoupling by the protonophore CCCP in the end of the experiment caused a steep rise in the ratio, indicating an increase in number of depolarized mitochondria and decrease in number of mitochondria exhibiting normal membrane polarization.
|
Using fluorescence microscopy changes of cellular membrane potential (depolarization) due to blockage of functional KATP channels were detected. Application of glibenclamide resulted in an increase in fluorescence intensity by 17% (n = 5, P = 0.0185), corresponding to membrane depolarization and, thereby, indicating the presence of functional KATP channels. Based on an assumed change in fluorescence intensity of about 1% per millivolt (supplier information) and published data, we roughly estimated depolarization being in the range of 10–15 mV (31).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We provide evidence for the molecular nature of a KATP channel in human GCs using RT-PCR and immunochemistry. RT-PCR analysis showed that in human GCs, mRNAs coding for the Kir 6.1 and SUR 2B subunits are expressed. The absence of mRNAs encoding SUR 1 and SUR 2A subunits in some GC preparation is unlikely to be due to RT-PCR problems but more likely because of heterogeneity of the primary human samples. This might be due to variability in the in vitro fertilization patients regarding hormonal treatment or pathological aberrations and further thorough examination of this probably (patho-)physiologically interesting observation is required.
Immunocytochemistry showed that the Kir 6.1 subunit is expressed in all human GCs. Regarding the localization of the KATP channel in the ovary, immunohistochemistry showed that the Kir 6.1 subunit is expressed in vascular smooth muscle cells, as expected. Importantly, however, the Kir 6.1 subunit was also detected in GCs of large follicles and those of the CL corresponding to isolated human GCs in culture. Thus, the KATP channel is specific for not only smooth muscle cells but also endocrine cells of the human ovary.
Furthermore, we showed that the channel is pharmacologically affected. The application of the KATP inhibitor glibenclamide leads to a depolarization of the plasma membrane caused by blockage of KATP channels. Because the channel can be inhibited, it must be assumed that the KATP is at least partially open. Taken together this would imply that the KATP is involved in establishing the plasma membrane potential as in other cell types (8, 11).
Gold immunostaining of human GCs and EM was used to examine the localization of the Kir 6.1 subunit more closely because we wanted to address the possible presence of mitoKATP channel, which was described in cardiomyocytes (35). EM allowed localizing the Kir 6.1 subunit to the plasma membrane of human GCs. The staining of organelles like mitochondria was at background level, which indicated that the KATP is not located in the mitochondrial membrane. To further examine a potential mitoKATP, changes in the activity of the mitochondrial membrane were evaluated using a fluorescent potential probe and confocal microscopy. In these experiments there was no response of the mitochondrial membrane potential to both the KATP inhibitor glibenclamide and the KCO diazoxide. Together with gold immunostaining, these results rule out the presence of a functional mitoKATP and support the sole localization of the KATP in GCs in the plasma membrane. The intracellular immunofluorescence signal (see Fig. 3A) is thus most likely due to synthesized but not yet to the plasma membrane targeted Kir 6.1 protein.
The main hormone produced by luteal granulosa cells is progesterone, and as we found, progesterone production of human GCs was affected by glibenclamide; the KATP inhibitor reduces hCG-induced progesterone production. This means that a blockage of the KATP and a subsequent depolarization of the plasma membrane reduces hormone production. Because obvious toxic effects (cell number, morphology) of the substances are ruled out, this implies that the KATP establishes a link between the plasma membrane potential and hormone production.
Previous reports on GCs from different animals support this idea because changes of extracellular ion concentrations, a rather crude intervention, which alters the membrane potential, affects steroidogenesis as well (36, 37, 38). The underlying mechanisms are, however, not yet known. The fact that in human GCs a KATP-independent hCG-induced progesterone production component exists and that blockage of other ion channel types has also an impact on progesterone production in GCs reflects the complexity of the process (15, 16, 17, 22, 23). We assume that the roles of different channel types in steroidogenesis are not simply additive, but most likely interactions between some of them are involved, e.g. the activation of voltage-dependent channels by hyperpolarizing channels [e.g. calcium-activated potassium channel (BKCa)]. In the case of the BKCa a role in steroidogenesis in human GCs was demonstrated; however, the function of the BKCa is not linked via the membrane potential to hormone production (15).
Finally, our findings are of interest from a pharmacological point of view as well because the mere presence of KATP channels in the human ovary implies that they might be a potential target for frequently used drugs directed to KATP channels in other tissues. Moreover, activation or blockage of KATP channels by endogenous substances (e.g. peptide hormones) is known from other cell types (8), and these compounds are also acting on GCs in the human ovary (15, 16). These possibilities require further thorough investigations.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure statement: L.K., J.S.R., and A.M. have nothing to declare.
First Published Online February 14, 2006
1 L.K. and J.S.R. contributed equally to this work. 
Abbreviations: BKCa, Calcium-activated potassium channel; CCCP, carbonyl cyanide 3-chlorophenylhydrazone; CL, corpus luteum; DiBAC4(3 ), bis-(1,3-dibutylbarbituric acid) trimethine oxonol; DMSO, dimethyl sulfoxide; EC, extracellular; EM, electron microscopy; GC, granulosa cell; hCG, human chorionic gonadotropin; KATP, ATP-sensitive potassium channel; KCO, potassium channel opener; Kir, inward rectifier potassium channel; mitoKATP, mitochondrial KATP; SUR, sulfonylurea receptor.
Received October 14, 2005.
Accepted February 8, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  P. Olivero, E. Leiva-Salcedo, L. Devoto, and A. Stutzin Activation of Cl- Channels by Human Chorionic Gonadotropin in Luteinized Granulosa Cells of the Human Ovary Modulates Progesterone Biosynthesis Endocrinology, September 1, 2008; 149(9): 4680 - 4687. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Malester, X. Tong, I. Ghiu, A. Kontogeorgis, D. E. Gutstein, J. Xu, K. D. Hendricks-Munoz, and W. A. Coetzee Transgenic expression of a dominant negative KATP channel subunit in the mouse endothelium: effects on coronary flow and endothelin-1 secretion FASEB J, July 1, 2007; 21(9): 2162 - 2172. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
