Molecular Human Reproduction, Vol. 6, No. 7, 602-609,
July 2000
© 2000 European Society of Human Reproduction and Embryology
Uterine physiology |
Kappa opioids and TGFß1 interact in human endometrial cells
1 Department of Pharmacology, 2 Department of Clinical Chemistry, 3 Department of Experimental Endocrinology and 4 Department of Oncology, Medical School, University of Crete, Heraklion GR-711 10, Crete, Greece
Abstract
The transforming growth factor ß1 (TGFß1) is a major regulator of human endometrial function. Human endometrium possesses specific opioid binding sites, the majority of which belong to the kappa type, for which the prodynorphin-derived opioids are the endogenous ligands. Since these two systems interact in several other tissues we postulated that opioids may affect the production of TGFß1 in human endometrium. We have found that kappa opioids exerted a time- and dose-dependent inhibitory effect on TGFß1 production from endometrial stromal and epithelial cells and from the Ishikawa human endometrial adenocarcinoma cell line. This effect was reversible by the specific opioid antagonist diprenorphine. To examine if this effect represents a paracrine endometrial response to locally produced kappa opioids we searched for the presence of the endogenous kappa opioid receptor ligands. Indeed, the prodynorphin transcript was detectable on Northern blots from normal and tumoral human endometrial cells; its size was that of the pituitary transcript, i.e. ~2.4 kb long. Most immunoreactive dynorphin from human endometrium had a molecular weight of 8 kDa. Finally, immunofluorescence staining of normal and tumoral human endometrial cells revealed the presence of dynorphin-positive cytoplasmic secretory granules. Taken together, our data suggest that in human endometrium, kappa opioids and the TGFß1 form a paracrine network which appears to be retained by the Ishikawa human endometrial adenocarcinoma cell line.
dynorphins/endometrium/human/opioids/TGFß
Introduction
Multiple lines of evidence suggest that the transforming growth factor (TGF) is an important modulator of endometrial physiology involved in the regulation of endometrial growth, proliferation, and differentiation (Godkin and Dore, 1998
). Indeed, endometrial TGFß1 affects the proliferation of normal human epithelial and stromal endometrial cells as well as the proliferation of the human endometrial adenocarcinoma cell line Ishikawa (Croxtall et al., 1992; Tang et al., 1994; Albright and Kaufman, 1995). Following ovulation, TGFß1 stimulates stromal cell proliferation, while at the same time inhibiting endometrial epithelial cellular mitosis promoting their differentiation to the characteristic mature cells of the secretory endometrium (Marshburn et al., 1994). In addition, TGFß1 promotes apoptosis of endometrial stromal cells at the implantation site in an effort to accommodate the developing trophoblast in a hospitable implantation chamber (Moulton, 1994).
Human endometrium produces opioids and their receptors suggesting that endometrial opioids exert local, autocrine/paracrine, effects. Indeed, transcripts of endogenous opioid precursors proopiomelanocortin (POMC) and proenkephalin (PENK) and their end products have been reported in human endometrial cells (Walhstrom et al., 1985, Petraglia et al., 1986, Makrigiannakis et al., 1992). Multiple types of opioid binding sites have been identified in human endometrial cells, with the kappa (
)-opioid receptor (1, 2, 3) being the predominant type, while -opioid receptors are present in lower numbers and the -type of receptors are not detectable (Hatzoglou et al., 1995a).
Since these two systems interact in several other tissues we postulated that opioids may affect the production of TGFß1 in human endometrium. Indeed in rat brain, -opioids are co-localized and interact with TGF
and ß (Code et al., 1987; Ramirez-Ordonez et al., 1999). Furthermore, -opioids stimulate the release of TGFß from human peripheral blood mononuclear and porcine immune cells (Chao et al., 1992; Zhou et al., 1992). The effects of opioids on TGF are not unique but they are rather part of a more generalized effect of opioids on growth factors. Thus, it has been shown that synthetic -opioid agonists can interact with nerve growth factor (NGF) in the rat pheochromocytoma cells (Margioris et al., 1992); inhibit macrophage-colony stimulating factor (M-CSF) in mouse bone marrow cells (Roy et al., 1991); and modulate the biological effects of epidermal growth factor (EGF) in monkey kidney cells (Belcheva et al., 1998).
In the first part of this study, we examined the effect of synthetic opioid agonists and antagonists on the production of TGFß1 by normal human endometrial cells and the Ishikawa human endometrial adenocarcinoma cell line. In the second part (following our data showing a specific effect of -opioid agonists on endometrial TGF), we examined whether the endogenous ligands for the -opioid receptors are produced in normal and tumoral human endometrium. We have found that human endometrium produces dynorphins. Taken together our data suggest that -opioids and TGF interact in a paracrine mode at the level of human endometrium.
Materials and methods
Primary culture of isolated epithelial and stromal cells from human endometrium
Endometrial specimens were obtained from patients undergoing diagnostic curettage or hysterectomy. Full ethics committee approval had been granted for this study. The histological type and pathology of the biopsies was confirmed by conventional pathological examination. Purified epithelial and stromal primary endometrial cell cultures were established as described previously (Chatzaki et al., 1994; Makrigiannakis et al., 1995). Briefly, the tissues were collected in Dulbecco's minimal essential medium (DMEM; Gibco BRL, Bethesda, MD, USA) supplemented with 10% fetal calf serum (FCS) and 1% antibiotic/antimycotic solution, trimmed and minced mechanically and digested for 90 min at 37°C using 0.25% of Type I collagenase (Sigma Chemical Co, St Louis, MO, USA). Separation of the epithelial and stromal fraction was carried out by filtration through a 45 µ stainless steel sieve and backwashing of the glands from the sieve, followed by pelleting by centrifugation. The two different cell types were then plated in 24-well plates (Corning Inc, New York, NY, USA) with 1 ml of culture medium and incubated in a humidified atmosphere of 5% CO2 at 37°C, in the above mentioned medium for 2 days before experimentation. At the beginning of the experiment, cultures had reached 100% confluency and at least 90% purity in epithelial or stromal cell content, as judged by light microscopy. Following this, they were pre-incubated in Phenol Red- and serum-free Roswell Park Memorial Institute (RPMI) 1640 medium (Biochrom Co, Berlin, Germany) supplemented with 0.25% bovine serum albumin (BSA) fraction V (Sigma), insulin from bovine pancreas (5 mg/l), transferin (5 mg/l) and sodium selenite (5 ng/l) (ITS cell culture supplement, Sigma), 2 mmol/l L-glutamine and 1% antibiotic/antimycotic solution for 24 h. The 1-opioid receptor agonist U69593 (Upjohn Co, Kalamazoo, MI, USA) and/or the general opiate antagonist diprenorphine (Sigma) were then added in the above medium in duplicate or triplicate wells, in parallel with non-treated controls.
Culture of the Ishikawa endometrial cell line
Ishikawa cells were established as a permanent cell line from a well differentiated human endometrial adenocarcinoma (Nishida et al., 1979; Gravanis and Gurpide, 1986). The cells were routinely cultured in 75 ml cell culture flasks (Corning) in Earle's minimal essential medium (Gibco) supplemented with 15% fetal bovine serum (FBS; Gibco), 10 mmol/l L-glutamine and 1% antibiotic–antimycotic solution (Gibco) to a final concentration of 100 IU/ml penicillin and 100 µg/ml streptomycin, in a humidified atmosphere of 5%CO2 at 37°C. Cells growing exponentially were seeded into 24-well plates (Corning) (25x103 cells/well) and incubated for 24 h in 1 ml growth medium. Before drug treatment, cells were placed in serum-free, Phenol Red-free medium as mentioned above.
Measurement of TGFß1
At the end of the treatment period, cell culture supernatants were collected and stored at –70°C for TGFß1 measurement, whereas cells were harvested in a standard trypsin solution (Gibco) and stored at –20°C for estimation of protein content. TGFß1 was measured using a human TGFß1 Quantikine immunoassay kit (R&D Systems, Oxon, UK), following the manufacturer's protocol. This includes an acidification step that activates latent TGFß1 to immunoreactive TGFß1 and a standard quantitative sandwich enzyme immunoassay technique. The sensitivity of the assay is 7 pg/ml. Results were expressed as pg/ml per µg of total cellular protein, determined on whole cellular homogenates by the Bradford method (Bradford, 1976), using BSA as standard.
Northern blot analysis
Total RNA was extracted from frozen tissues or cultured cells by the guanidine thiocyanate method (Maniatis et al., 1989). Following size-fractionation of RNA (50 µg/lane) by electrophoresis through 1.5% agarose gels containing 6% formaldehyde and 2 µg/ml ethidium bromide, gels were viewed under UV light to assess the integrity. After transfer to GeneScreen nylon filters (New England Nuclear, Boston, MA, USA), they were prehybridized and hybridized as previously described (Maniatis et al., 1989). A synthetic 48-mer oligonucleotide against bases 36–83 of the rat prodynorphin (PDYN) mRNA (this area of mRNA is identical in both humans and rats) (Douglass et al., 1989) was used as a probe that was labelled at the 3' end with [
-32P]-deoxyadenocine triphosphate (800 Ci/mmole; Amersham, Arlington, IL, USA) and terminal deoxynucleotidyl transferase (BRL, Bethesda, MD, USA) to a specific activity of ~108 dpm/mg. Blots were washed in 0.2x SSC, 0.1% SDS for 30 min at 60°C. Autoradiography using Kodak XR film took place at –70°C in the presence of an intensifying screen. The approximate size of mRNAs was determined relatively to 18S and 28S rRNAs.
Gel filtration chromatography and radioimmunoassay for dynorphins
Peptides in Ishikawa cell culture medium and endometrial homogenates were concentrated by a C-18 reverse phase column (Sep-Pak; Waters Associates, Milford, MA, USA). Briefly, culture media and cellular homogenates were acidified by 10 volumes of 0.1 N HCI and centrifuged at 5000 g for 10 min. The supernatants were extracted by activated Sep-Pak cartridges, washed with 20 ml 0.1 N HCl, eluted with 3 ml acetonitrile 80%–0.01% HCI, then dried under vacuum (Speed-Vac). The recovery of synthetic dynorphin diluted in medium using this method was >90%. Samples were then reconstituted in 0.5 ml 10% formic acid containing 0.5% defatted BSA and 6 mol/l urea and chromatographed on a Sephadex G-50 column (0.9x60 cm, bed volume 40 ml). The G-50 column was calibrated with blue dextran and dynorphin-(1–13). With flow rate 1.5 ml/h, fractions of 1 ml were collected, dried under vacuum and reconstituted for radioimmunoassay using an antiserum raised against synthetic porcine dynorphin-(1–13), which is identical to human dynorphin-(1–13)(Suda et al., 1983; Margioris et al., 1992). The antiserum cross-reacts with human dynorphin-(1–13) and dynorphin A. It exhibits no cross-reaction with synthetic human ß-endorphin, neo-endorphin, or met- or leu-enkephalin. The sensitivity of the assay was 1 pg/tube.
Indirect immunofluorescence
Immunofluorescence staining was performed as described previously (Fostinis et al., 1992). Briefly, cells grown on 22x22 mm coverslips in absence of serum were fixed with acetone/methanol (9:1) for 20 min. The cells were then incubated for 1 h at 4°C with rabbit antiserum against human dynorphin, diluted 1:250 in phosphate-buffered saline (PBS) 0.1% BSA (Margioris et al., 1992) in parallel with a negative control (PBS 0.1% BSA). After washing, an anti-rabbit antibody conjugated to fluorescein isothiocyanate (FITC) was added (1:150, Amersham) for 1 h at 4°C. Specimens were visualized in a Zeiss Axioscope microscope and photographed using TMZ 135–36 Kodak film.
Statistical analysis
Data of TGFß1 concentrations are presented either as pg/ml per µg of protein or as percentage of non-treated controls. For the statistical analysis of the results we used analysis of variance followed by post-hoc comparison of means and least significant difference test or planned comparison. All analyses were carried out on combined results from at least three identical experiments, performed in duplicate or triplicate.
Results
Effects of opioid agonists and antagonists on TGFß1 production from primary cultures of human endometrial cells
Purified primary cultures were established after isolation of the respective cellular fraction from endometrial biopsies, obtained from patients undergoing diagnostic curettage. As we have previously reported, our method provides a pure population of isolated epithelial or stromal monolayers, with little contamination from other uterine cell types (Makrigiannakis et al., 1995). We investigated TGFß1 production from epithelial and stromal cell primary cultures of human endometrium together with the effect of -opioid peptides on endometrial TGFß1. For this purpose, the use of synthetic opioid receptor ligands was preferred to dynorphin peptides, due to their stability in vitro. U69593 is a non-peptidic synthetic molecule that exhibits specific agonist activity on 1-opioid receptors and was therefore suitable for the purpose of these experiments. The concentration of TGFß1 was measured in the culture supernatant after 2 days of treatment with U69593 (10–7 mol/l) in parallel with non-treated controls and values were corrected for total protein content.
The concentration of TGFß1 in media under basal conditions or after opioid agonist application is presented in Figure 1. TGFß1 was secreted from all tissues studied and from both epithelial and stromal cell fraction. The mean values of the basal concentration of TGFß1 from each cell type cultured for 2 days were expressed as pg/µg of protein ± SEM, and were 14.97 ± 5.27 for proliferative epithelium (PE, n = 7), 10.62 ± 2.66 for secretory epithelium (SE, n = 5), 146.73 ± 52.21 for proliferative stroma (PS, n = 4) and 61.26 ± 14.15 for secretory stroma, (SS, n = 5) (Figure 1A). It is noted that the stromal fraction secreted higher levels than the glandular in both phases of the cycle. This finding corroborates with other reports showing that stroma is the principal source of TGFß1 in the uterine cavity (Chegini et al., 1994; Marshburn et al., 1994).
|
Effect of opioid agonists and antagonists on TGFß1 production from human endometrium cells in culture
Incubation of various endometrial cell types with 10–7 mol/l U69593 caused a significant reduction in the concentration of TGFß1 in the culture supernatant of almost all primary cultures (Figure 1B
). Indeed, this effect was observed in seven out of seven of PE, three out of five SE, four out of four PS and four out of five SS. As shown in Figure 1, the inhibitory effect of the opioid agonist was at much the same level (52 ± 9 to 69 ± 5% of non-treated controls) in all types of cells. We further explored the observed effect in time- and dose-response experiments, using cells from both phases of the cycle (Figure 2). Time–response curves (Figure 2A) showed that the statistically significant reduction of TGF-ß1 secretion by U69593 was abolished after 6 days of culture of epithelial cells and 4 days of culture for stromal cells. Thus, 2-day treatment was chosen for all the following experiments. Also, it was noticed that basal release of TGFß1 decreased during the 6 days of culture, probably reflecting a decrease in the number of viable cells, due to prolonged incubation in serum-free medium with no oestrogen supply that cannot support endometrial growth for long. However, the use of this medium was necessary, because serum contains measurable levels of TGFß1, as well as other growth factors and steroid residues that could affect the regulation of TGFß1 expression (Gutierrez et al., 1998). Dose–response curves were biphasic for both stromal and epithelial cells (Figure 2B). An inhibitory effect of U69593 was observed at 10–10 mol/l (0.1 nmol/l) and at 10–8 and 10–7 mol/l (10 and 100 nmol/l), but not at 10–9 mol/l (1 nmol/l). When further dilutions of the drug were tested, it was shown that the inhibitory effect was abolished in the nanomolar concentrations of U69593 ranging from 2x10–10 to 10–9 mol/l, whereas in 2x10–9 mol/l the effect was statistically significant (data not shown). As discussed later, this interesting profile probably reflects interaction of the ligand with multiple types of opioid or other receptors. In order to test whether the observed effect is mediated through specific opioid receptors, we measured TGFß1 release after 2 days of treatment of various endometrial cell types with U69593 (10–7 mol/l) in the presence or absence of a ten-fold excess of the opioid antagonist diprenorphine (10–6 mol/l). Diprenorphine was chosen due to its general opioid receptor antagonistic activity compared with the more often used naloxone, a µ-preferential antagonist (Billington et al., 1990). Naloxone has also been reported to exhibit also agonist activity (Cruz et al., 1996). As shown in Figure 2C, the inhibition caused by U69593 was reversed in the presence of the antagonist, diprenorphine, suggesting action through opioid receptors. It is interesting to note that treatment with diprenorphine alone increased TGFß1 secretion in a statistically significant way, suggesting that the action of exogenously produced opiods is blocked.
|
Effect of opioid agonists and antagonists on TGFß1 production from Ishikawa human endometrial adenocarcinoma cells in culture
Ishikawa cells were treated with U69593 (10–7 mol/l) and the concentration of TGFß1 was measured in the culture media after 2, 4 and 6 days of treatment. The combined results of three identical experiments are illustrated in Figure 3A
. A statistically significant reduction of the concentration of TGFß1 was observed, with a maximum of 52.02 ± 5.26% of non-treated control (P < 0.05) on the second day of treatment. The effect remained similarly profound after 4 and 6 days of treatment (55.57 ± 6.30 and 56.64 ± 7.53 respectively, P < 0.05). The dose–response curve of this effect, measured after 2 days of treatment, is presented in Figure 3B. The profile of the curve was biphasic, confirming the results from the primary cultures of endometrial cells. Indeed, although we were able to detect a significant decrease in concentrations as low as 10–12 mol/l (0.001 nmol/l), at the concentration of 10–9 mol/l (1 nmol/l), the effect was abolished. Finally, we measured TGFß1 release after 2 days of treatment with U69593 (10–7 mol/l) in the presence or absence of a 10-fold excess of the opioid antagonist diprenorphine (10–6 mol/l). As shown in Figure 3C, the inhibition caused by U69593 was reversed in the presence of the antagonist, suggesting action through opioid receptors. It is interesting to note that treatment of Ishikawa cells with diprenorphine alone increased TGFß1 secretion in a statistically significant way (148.05 ± 8.48% of non-treated control, P < 0.05), suggesting that the antagonist blocked the action of endogenously produced opioids.
|
Detection of the PDYN transcript in human endometrial cells
In order to examine the expression of the PDYN gene by endometrial cells, preparations of total RNA from normal human secretory endometrial biopsy and Ishikawa cells, were subjected to Northern blot hybridization analysis using a PDYN mRNA probe. In parallel, total RNA isolated from rat pituitary and rat liver was used as positive and negative controls respectively. Figure 4A
depicts the autoradiography of Northern blots revealing the presence of PDYN transcripts in RNA extracted from endometrial sources (lanes 3 and 4), the size of which was ~2.4 kb long, similar or identical to this detected in the rat pituitary (lane 1, positive control). The probe used did not hybridize with total RNA isolated from rat liver (lane 2, negative control).
|
Characterization and localization of dynorphin in human endometrial cells
To identify the type of immunoreactive material that is the end product of the PDYN mRNA expressed in endometrial cells, normal endometrial cell homogenates and culture supernatant from Ishikawa cells were concentrated and analysed by gel filtration chromatography. The bulk of immunoreactive material eluted from the chromatography exhibited the molecular weight of the 8 kDa form of dynorphins, while small quantities of a 4 kDa peptide was also detectable in culture media from Ishikawa cells (Figure 4B
). It is interesting that the chromatographic profile from both sources exhibits a very similar profile, confirming the resemblance of the Ishikawa cell line behaviour to the original tissue.
Endometrial cells growing in vitro on slides were stained with an anti-human dynorphin antibody and examined by immunofluorescence. Ishikawa cells (Figure 5B) and epithelial cells from primary cultures (Figure 5D) were strongly positive in contrast with the negative controls (cells treated without primary antibody) (Figure 5A and C). Dynorphin was localized in the cytoplasm, where well-demarcated positive granules were observed. There was a marked concentration of immunoreactive material in the peri-nuclear area in both preparations, whereas nuclei were negative. Preparations from normal and tumoral epithelial cells revealed the same pattern of expression. Isolated stromal cells from primary endometrial cultures showed also weak cytoplasmic staining for dynorphin (Figure 5F), compared with the negative control (Figure 5E).
|
Discussion
Several reports suggest that opioid peptide agonists modulate the production and/or effects of growth factors in a paracrine manner. Thus, the endogenous -opioid agonist, dynorphin, inhibits the effect of macrophage-colony stimulating factor (M-CSF) on mouse bone marrow cells (Roy et al., 1991). Similarly, the synthetic -opiod agonist U-69593 inhibits the proliferative effect of epidermal growth factor (EGF) on PC12 rat pheochromocytoma cells (Venihaki et al., 1996). U69593 also affects the EGF-dependent modulation of extracellular signal-regulated protein kinase in monkey kidney cells (Belcheva et al., 1998). µ-opioid agonists stimulate the release of TGFß from human peripheral blood mononuclear and porcine immune cells (Chao et al., 1992, Zhou et al., 1992) and in rat brain they are co-localized and interact with TGF and ß (Code et al., 1987; Ramirez-Ordonez et al., 1999). Human endometrium is a typical example of a tissue where opioids and growth factors co-exist side-by-side. TGFß1 represents a major endometrial growth factor (Godkin and Dore, 1998). Indeed, TGFß1 affects the proliferation of epithelial and stromal cells of normal and tumoral human endometrium (Croxtall et al., 1992; Marshburn et al., 1994; Tang et al., 1994, Albright and Kaufman, 1995). It should be noted that TGFß1 also enhances the apoptosis of normal endometrial stroma (Moulton, 1994). Based on the above data we postulated that opioids may affect TGFß1 production in human endometrium. The bulk of opioid binding sites in normal and tumoral human endometrium are of the 1 type. Based on this data, we used a specific 1-opioid receptor agonist, U69593. We have found that U69593 had a consistent inhibitory effect on TGFß1 production from endometrial cells obtained from biopsies of proliferative and secretory endometrium. This inhibitory effect was equally demonstrable in epithelial as well as stromal cells and on the epithelial cells of the Ishikawa cell line. This effect appeared to be specifically mediated by opioid receptors since the general opioid antagonist, diprenorphine, reversed it. Interestingly, the dose–response curves of TGFß1 were bell-shaped, showing an attenuation of this effect of U69593 at nanomolar concentrations. This type of response to opioids has been described in other systems (Yamaguchi et al., 1998). It may reflect the low affinity that each `specific' opioid agonist has towards other types of opioid receptors as well as to the competition of exogenous synthetic opioids to endogenous opioid agonists, or even other endometrial factors, e.g. interleukin-6, that can activate opioid receptors (Wang et al., 1996). Furthermore, it has been suggested that at high concentrations, opioids may also bind to completely different receptors (Hatzoglou et al., 1995b). Experiments conducted using Ishikawa cells transiently transfected with the TGFß1 promoter linked to the luciferase reporter gene have shown that treatment of the transfectants with U69593 had no effect on luciferase expression suggesting that the effect of opioids is post-transcriptional (data not shown). Our data are in agreement with reports showing that steroids and retinoids regulate TGFß1 at a post-transcriptional level (Kim et al., 1992). Indeed, it has been suggested that a stem-loop which is formed in the 5'-untranslated region of the TGFß1 transcript may represent a possible target of the opioids which may, thus, suppress translation (Romeo et al., 1993).
The paracrine effect of opioids on TGFß1 in human endometrium may be part of a larger network composed of several neuropeptides regulating endometrial growth and differentiation via modulation of local growth factors. TGFß1 appears to be such a factor regulating the proliferation of both epithelial and stromal endometrial cells as well as that of the Ishikawa cells (Croxtall et al., 1992; Tang et al., 1994; Albright and Kaufman, 1995). Furthermore, TGFß1 stimulates stromal cell proliferation preparing the endometrium for placental implantation while, at the same time, inhibiting epithelial mitosis and promoting their differentiation (Marshburn et al., 1994). TGFß1 also stimulates the apoptosis of endometrial stroma (Moulton, 1994), a mechanism responsible for the degeneration of antimesometrial decidua that enables trophoblast development and remodelling of the implantation chamber. Opioids have been shown to affect cell proliferation. Thus, it has been reported that they inhibit lung cancer (Maneckjee and Minna, 1990) and PC12 rat pheochromocytoma cell proliferation (Venihaki et al., 1996) while preventing their apoptosis (Dermitzaki et al., 2000), and in human prostate cancer cells they stimulate cell proliferation (Moon, 1988). In addition, opioid peptides inhibit the action of oestradiol on human myometrial cells (Kornyei et al., 1999). It is possible that at least some of these effects of opioids are mediated by locally produced growth factors affected by opioids in a paracrine manner. Indeed, in the present study, we report the expression of the main endogenous ligand for the opioid receptors, the dynorphin, in normal and tumoral human endometrial cells. More specifically, we have found that the PDYN gene transcripts and their protein end-product dynorphin are detectable in both types of cells. On Northern blot analysis the size of the transcript is 2.4 kb in size, which is similar or identical to that present in other tissues including rat pituitary and adrenals. The size of the immunoreactive material was ~8 kDa, a finding which confirmed our hypothesis. Localization of dynorphin by immunofluorescent staining showed vesicles containing this peptide in the cytoplasm of normal and malignant endometrial cells. Epithelial cells had higher concentrations while stromal cells much less. Expression of dynorphin peptides has been previously reported in the uterus of other species. Indeed, Douglass et al. showed low amounts of immunoreactive dynorphin A in rat but not guinea pig and rabbit uterus (Douglass et al., 1987). Similarly, in human endometrial cells, we could not detect dynorphin A, while the main product of the PDYN processing was the 8 kDa bioactive peptide. Multiple forms of dynorphins are also present in another intra-uterine site, human placenta (Ahmed et al., 1986, 1987). However, neither the uterus nor the placenta of mice soon after implantation and at any age of gestation express dynorphins (Zhu and Pintar, 1998), implying species-specific differences.
In summary, we report here the expression of the PDYN gene by normal human endometrium and also the synthesis and secretion of dynorphins by these cells. In addition, we show an inhibitory effect of the -opioid agonist, U69593, on endometrial TGFß1. Our results underline the significance of these neuropeptides in endometrial physiology, since they implicate their activity with the expression of a key growth factor. The cross-talk between endometrial opioids and growth factors suggest the involvement of these agents in the regulation of uterine cell proliferation and differentiation.
Acknowledgments
Dr. E.Chatzaki was supported by the Greek National Foundation of Fellowships (IKY). This work was supported by the European Union EKBAN99-66 grant to A.G.
Notes
5 To whom correspondence should be addressed at: Medical School, University of Crete, Heraklion 71110 Greece. E-mail: gravanis{at}med.voc.gr 
References
Ahmed, M.S., Randall, L.W., Cavinato, A.G. et al. (1986) Human placental opioid peptides: correlation to the route of delivery. Am. J. Obstet. Gynecol., 155, 703–706.[ISI][Medline]
Ahmed, M.S., Randall, L.W., Sibai, B. et al. (1987) Identification of dynorphin 1–8 in human placenta by mass spectrometry. Life Sci., 25, 2067–2076.
Albright, C.D. and Kaufman, D.G. (1995) Transforming growth factor-beta 1 mediates communication between human endometrial carcinoma cells and stromal cells. Pathobiology, 63, 314–319.[ISI][Medline]
Belcheva, M.M., Vogel, Z., Ignatova, E. et al. (1998) Opioid modulation of extracellular signal-regulated protein kinase activity is ras-dependent and involves Gbetagamma subunits. J. Neurochem., 70, 635–645.[ISI][Medline]
Billington, C.J., Herman, B.H., Bartness, T.J. et al. (1990) Effects of the opiate antagonists diprenorphin and naloxone and of selected opiate agonists on feeding behavior in guinea pigs. Life Sci., 46, 147–154.[ISI][Medline]
Bradford, M. (1976) A rapid and sensitive method for the quantitation of protein utilizing the principle of protein-dye binding. Anal. Biochem., 72, 248–254.[ISI][Medline]
Chao, C.C., Hu, S., Molitor, T.W. et al. (1992) Morphine potentiates transforming growth factor-beta release from human peripheral blood mononuclear cell cultures. J. Pharmacol. Exp. Ther. 262, 19–24.
Chatzaki, E., Gallagher, C.J., Iles, R.K. et al. (1994) Characterization of the differential expression of marker antigens by normal and malignant endometrial epithelium. Br. J. Cancer, 69, 1010–1014.[ISI][Medline]
Chegini, N., Zhao, Y., Williams, R.S. and Flanders, K.C. (1994) Human uterine tissue throughout the menstrual cycle expresses transforming growth factor-beta 1 (TGF beta 1), TGF beta 2, TGF beta 3, and TGF beta type II receptor messenger ribonucleic acid and protein and contains [125I]TGF beta 1-binding sites. Endocrinology, 135, 439–449.[Abstract]
Code, R.A., Seroogy, K.B. and Fallon, J.H. (1987) Some transforming growth factor-alpha connections and their co localization with enkephalin in the rat central nervous system. Brain Res., 421, 401–405.[ISI][Medline]
Croxtall, J.D., Jamil, A., Ayub, M. et al. (1992) TGF-beta stimulation of endometrial and breast-cancer cell growth. Int. J. Cancer, 12, 822–827.
Cruz, S.L., Villarreal, J.E. and Volkow, N.D. (1996) Further evidence that naloxone acts as an inverse opiate agonist: implications for drug dependence and withdrawal. Life Sci., 58, 381–389.
Dermitzaki, I., Chatzaki, E., Gravanis, A. and Margioris, A. (2000) Opioids transiently prevent activation of apoptotic mechanisms following short periods of serum deprivation. J. Neurochem., 74, 960–969.[ISI][Medline]
Douglass, J., Cox, B., Quinn, B. et al. (1987) Expression of the prodynorphin gene in male and female mammalian reproductive tissues. Endocrinology, 120, 707–713.[Abstract]
Douglass, J., McMurray, C.T., Garrett, J.E. et al. (1989) Characterization of the rat prodynorphin gene. Mol. Endocrinol., 3, 2070–2078.[Abstract]
Fostinis, Y., Theodoropoulos, P., Gravanis, A. and Stournaras, C. (1992) Heat shock protein HSP90 and its association with the cytoskeleton: a morphological study. Biochem. Cell Biol., 70, 779–786.[ISI][Medline]
Godkin, J.D. and Dore, J.J. (1998) Transforming growth factor beta and the endometrium. Rev. Reprod., 3, 3–6.
Gravanis, A. and Gurpide, E. (1986) Effects of estradiol on DNA polymerase activity in a human endometrial adenocarcinoma cell line (Ishikawa cells). J. Clin. Endocrinol. Metab., 63, 356–359.[Abstract]
Gutierrez, M., Menendez, L., Brieva, R. et al. (1998) Different types of steroids inhibit [3H]diprenorphine binding in mouse brain membranes. Gen. Pharmacol., 31, 747–751.[ISI][Medline]
Hatzoglou, A., Gravanis, A., Margioris, A.N. et al. (1995a) Identification and characterization of opioid-binding sites present in the Ishikawa human endometrial adenocarcinoma cell line. J. Clin. Endocrinol. Metab., 80, 418–423.[Abstract]
Hatzoglou, A., Ouafik, L., Bakogeorgou, E. et al. (1995b) Morphine cross-reacts with somatostatin receptor SSTR2 in the T47D human breast cancer cell line and decreases cell growth. Cancer Res., 102, 5632–5636.
Kornyei, J.L., Vertes, Z., Oszter, A. et al. (1999) Opioid peptides inhibit the action of oestradiol on human myometrial cells in culture. Mol. Hum. Reprod., 5, 565–572.
Kim, S.J., Park, K., Koeller, D. et al. (1992) Post-transcriptional regulation of the human transforming growth factor-beta 1 gene. J. Biol. Chem., 5, 13702–13707.
Makrigiannakis, A., Margioris, A., Markogiannakis, E. et al. (1992) Steroid hormones regulate the release of immunoreactive ß-endorphin from the Ishikawa human endometrial cell line. J. Clin. Endocrinol. Metab., 75, 584–589.[Abstract]
Makrigiannakis, A., Zoumakis, M., Margioris, A. et al. (1995) The Corticotropin-Releasing Hormone in normal and tumoral epithelial cells of human endometrium. J. Clin. Endocrinol. Metab., 80, 185–189.[Abstract]
Maneckjee, R. and Minna, J.D. (1990) Opioid and nicotine receptors affect growth regulation of human lung cancer cell lines. Proc. Natl Acad. Sci. USA, 87, 3294–3298.
Maniatis, T., Fritsch, E.F. and Sambrook, J. (1989) Molecular cloning. In Maniatis, T. (ed.), A Laboratory Manual. Vol 2. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, USA, pp. 37–43.
Margioris, A.N., Markogiannakis, E., Makrigiannakis, A. and Gravanis, A. (1992) PC12 rat pheochromocytoma cells synthesize dynorphin. Its secretion is modulated by nicotine and nerve growth factor. Endocrinology, 131, 703–709.[Abstract]
Marshburn, P.B., Arici, A.M. and Casey, M.L. (1994) Expression of transforming growth factor-beta 1 messenger ribonucleic acid and the modulation of deoxyribonucleic acid synthesis by transforming growth factor-beta 1 in human endometrial cells. Am. J. Obstet. Gynecol., 170, 1152–1158.[ISI][Medline]
Moon, T.D. (1988) The effect of opiates upon prostatic carcinoma cell growth. Biochem. Biophys. Res. Commun., 153, 722–727.[ISI][Medline]
Moulton, B.C. (1994) Transforming growth factor-beta stimulates endometrial stromal apoptosis in vitro. Endocrinology, 134, 1055–1060.[Abstract]
Nishida, M., Kasahara, K., Kaneko, M. et al. (1979) Establishment of a new human adenocarcinoma cell line, Ishikawa cells, containing oestrogen and progesterone receptors. Acta Obstet. Gynecol. Jpn., 37, 1103–1111.
Petraglia, F., Facchinetti, F., M'Futa, K. et al. (1986) Endogenous opioid peptides in uterine fluid. Fertil. Steril., 46, 247–251.[ISI][Medline]
Ramirez-Ordonez, R., Barreto-Estrada, J.L. and Garcia-Arraras, J.E. (1999) Growth factors effects on the expression of morphological and biochemical properties of avian embryonic sympathetic cells. Emphasis on NGF. Brain Res. Dev. Brain Res., 12, 27–36.
Romeo, D.S., Park, K., Roberts, A.B. et al. (1993) An element of the transforming growth factor-beta 1 5'-untranslated region represses translation and specifically binds a cytosolic factor. Mol. Endocrinol., 7, 759–766.
Roy, S., Loh, H.H. and Lee, N.M. (1991) Dynorphin blocks opioid inhibition of macrophage-colony stimulating factor-induced proliferation of bone marrow cells. Eur. J. Pharmacol., 202, 355–359.[ISI][Medline]
Suda, T., Tozawa, T., Tachibana, S. et al. (1983) Multiple forms of immunoreactive dynorphin in human pituitary and pheochromocytoma. Life Sci., 32, 865–870.[ISI][Medline]
Tang, X.M., Zhao, Y., Rossi, M.J. et al. (1994) Expression of transforming growth factor-beta (TGF beta) isoforms and TGF beta type II receptor messenger ribonucleic acid and protein, and the effect of TGF beta s on endometrial stromal cell growth and protein degradation in vitro. Endocrinology, 135, 450–459.[Abstract]
Venihaki, M., Gravanis, A. and Margioris, A.N. (1996) Kappa opioids exert a strong antiproliferative effect on PC12 rat pheochromocytoma cells. Peptides, 17, 413–439.[ISI][Medline]
Walhstrom, T., Laatikainen, K., Salminen, J. and Leppaluoto, J. (1985) Immunoreactive ß-endorphin is demonstrable in the secretory but not in the proliferative endometrium. Life Sci., 36, 987–991.[ISI][Medline]
Wang, Y., Pei, G., Cai, Y.C. et al. (1996) Human interleukin-2 could bind to opioid receptor and induce corresponding signal transduction. Neuroreport., 20, 11–14.
Yamaguchi, T., Kitagawa, K. and Kuraishi, Y. (1998) Itch-associated response and antinociception induced by intracisternal endomorphins in mice. Jpn. J. Pharmacol., 78, 337–343.[Medline]
Zhou, Y., Scamurra, R., Molitor, T.W. and Murtaugh, M.P. (1992) Characterization of transforming growth factor-beta 1 gene expression in porcine immune cells. Mol. Immunol., 29, 965–970.[ISI][Medline]
Zhu, Y. and Pintar, J.E. (1998) Expression of opioid receptors and ligands in pregnant mouse uterus and placenta. Biol. Reprod., 59, 925–932.
Submitted on December 21, 1999; accepted on April 26, 2000.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  E. Chatzaki, A. Makrigiannakis, A.N. Margioris, E. Kouimtzoglou, and A. Gravanis The Fas/FasL apoptotic pathway is involved in {kappa}-opioid-induced apoptosis of human endometrial stromal cells Mol. Hum. Reprod., September 1, 2001; 7(9): 867 - 874. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||