| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of Zoology (L.W.T.C., A.S.T.W.), Pathology (A.N.Y.C.), and Obstetrics and Gynecology (H.Y.S.N.), University of Hong Kong, Hong Kong; Department of Physiology (S.C.L.A.), Chinese University of Hong Kong, Hong Kong; Division of Pharmaceutical Sciences (J.T.-T.), University of Missouri, Kansas City, Missouri 64110; Department of Ophthalmology (J.T.-T.), Yale University School of Medicine, New Haven, Connecticut 06520; and Department of Obstetrics and Gynecology (N.A.), University of British Columbia, Vancouver, British Columbia, Canada V6H 3V5
Address all correspondence and requests for reprints to: Alice S. T. Wong, University of Hong Kong, Department of Zoology, 4S-14 Kadoorie Biological Sciences Building, Pokfulam Road, Hong Kong. E-mail: awong1{at}hku.hk.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Pigment epithelium-derived factor (PEDF) is a secreted 50-kDa glycoprotein first described as an ocular neurotrophic protein synthesized by fetal retinal pigment epithelial cells (4, 5). Subsequent studies demonstrate that it is widely expressed in most normal tissues (neuronal and nonneuronal). PEDF action appears to be cell specific. To neurons, PEDF is neurotrophic, promoting cell survival and differentiation (6, 7, 8). However, when endothelial cells are exposed to PEDF, their cell migration and proliferation are inhibited (9), and they undergo apoptosis (10, 11). Although the specific functions of PEDF in cancer are not known, there is some evidence that PEDF could have inhibitory effects on tumor growth. First, the levels of PEDF are highest in normal tissue (prostate, liver, and melanocyte), and they decrease during carcinogenesis (12, 13, 14, 15). Second, PEDF has been reported to have a direct antitumor effect on melanoma (13), osteosarcoma (16), endometrial (17), and prostate (12, 14) cancer cells by inhibiting angiogenesis, cell growth, and migration. Finally, in PEDF-knockout mice, vascular density is markedly increased, and epithelial cell hyperplasia is evident in the prostate and exocrine pancreas (12).
PEDF appears to play an important role in determining the tissue growth of many organs, and this may include the ovary, where a high basal level of PEDF mRNA expression is found (18). Interestingly, allelic loss in the 17p13.3 region, where the PEDF gene is located, is frequently encountered in early-stage ovarian carcinoma (19), implicating PEDF in this region as a potential tumor suppressor gene regulating the behavior of ovarian cancers. As yet, however, no study has examined the regulation of PEDF expression in normal ovaries or ovarian tumors.
The mechanism underlying the regulation of PEDF is largely unknown. Several lines of evidence prompted us to investigate that one of the mechanisms regulating PEDF expression in OSE or ovarian cancer cells might involve estrogen. 1) Estrogen is the major female hormone with mitogenic activities on a variety of tissues including the ovary. Normal OSE and most ovarian carcinomas have specific receptors for estrogen (20, 21, 22). A role of estrogen in ovarian carcinogenesis has been suggested; however, the epidemiological data are inconsistent. Some past studies have linked the use of postmenopausal estrogen replacement therapy to an increased risk, whereas other reports detected an unchanged or reduced risk of developing the cancer. Estrogen taken as oral contraceptives during premenopausal years seems to offer protection (23, 24, 25). Although further investigations are required to resolve how exogenous estrogens affect cancer risk, numerous studies have demonstrated the mitogenic action of estrogen in ovarian cancer. For instance, experimental ovarian tumor could be induced by estrogens (26, 27), and estrogen stimulated the growth of several estrogen receptor (ER)-positive ovarian carcinoma cell lines in vitro (21, 28, 29). However, little is known about the estrogen-regulated genes in ovarian tissue. 2) It is of interest to note that a putative estrogen-response element has been identified in the 5'-flanking region of the PEDF gene, further implying that PEDF may be linked to the regulation of hormone-dependent growth in the ovary and that in the ovarian cancer (18).
In this study, we aimed to elucidate the role of PEDF in the tumorigenesis of ovarian cancer and in particular to investigate the possible regulation of PEDF by 17ß-estradiol (E2) in ovarian epithelial cells. Our findings indicate that PEDF plays an important role in regulating the normal OSE cell function, and its decrease or loss in ovarian tumors could contribute to tumor development and progression. In addition, we demonstrate that estrogen-mediated down-regulation of PEDF is a novel way of reducing PEDF levels in OSE and mediates the effect of E2 on proliferation of these cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and 5–7.
|
|
|
|
Treatments
To study the regulation of PEDF expression by E2, 5 x 105 cells were plated onto six-well dishes. After 48 h preincubation in phenol red-free medium supplemented with 2.5% charcoal/dextran-treated FBS (for removal of endogenous steroids in the serum), the cells were treated with 0.01–100 nM E2 (Sigma) in fresh medium for 24 h. To test the specificity of E2 effect, some cells were exposed to a blocker of classical ER, ICI 182,780 (ICI; Tocris Cookson Ltd., Bristol, UK), 30 min before and during treatment with E2. To study the stability of PEDF mRNA, cells were treated with 100 nM E2 for 3 h, and cells without pretreatment served as controls. Actinomycin D (ActD; 4 µg/ml) (Calbiochem, San Diego, CA) was then added to the cultures, and total RNA was prepared at the times indicated for up to 8 h. To inhibit protein synthesis, cells were incubated with 4 µg/ml cycloheximide (CHX) (Sigma) for 1 h followed by the addition of 100 nM E2 for 24 h. Control cultures were treated with the vehicle (i.e. ethanol) alone. To examine the effect of E2 and recombinant PEDF (rPEDF) (Upstate Biotechnology Inc., Lake Placid, NY) on cell growth, cells were plated for 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and apoptosis assays. Twenty-four hours later, the cultures were treated with E2 in the presence or absence of rPEDF for 5 d, with fresh hormone added every other day.
Immunohistochemistry
Formalin-fixed, paraffin-embedded tissues that include three normal ovarian tissues (ages, 23–47 yr; mean age, 31.3 yr), 13 benign tumors (ages, 20–67 yr; mean age, 36.8 yr), 12 borderline tumors (ages, 24–57 yr; mean age, 35.9 yr), and 24 ovarian carcinomas (ages, 30–69 yr; mean age, 49.6 yr) (10 serous, six endometrioid, four mucinous, and four clear cell) were obtained with Internal Review Board approval from the archives of the Department of Pathology of Queen Mary Hospital, the University of Hong Kong. Patients were identified anonymous reference numbers. Five-micrometer sections were deparaffinized and rehydrated in a graded series of ethanol following standard protocol. Sections were treated with 3% H2O2 for 10 min to eliminate endogenous peroxidase activity and then incubated with an anti-PEDF antibody (1:200) (Bioproducts, Middletown, MD) for 30 min at room temperature before exposure to biotinylated secondary antibodies for 10 min. Antigen-antibody complexes were visualized using the substrate-chromogen mixture (Zymed Laboratories Inc., San Francisco, CA) and counterstained with hematoxylin. Omission or substitution of the primary antibody with preimmune serum was used as a negative control. Immunoreactivity was assessed by the intensity and percentage of positive staining. Staining intensity was scored as 0 (negative), 1 (faint), 2 (moderate), and 3 (strong). A case was considered to be negative if there was no staining or weak staining involving less than 10% cells.
Silencing of PEDF by small interfering RNA (siRNA)
The Smart-pool siRNA for silencing PEDF (catalog no. M-010153-00) was purchased from Dharmacon (Lafayette, CO). Transfection of the PEDF-specific siRNA (20 nM) was performed using SilentFect (Bio-Rad, Carlsbad, CA) per the manufacturer’s instructions. As a nonspecific siRNA control, scrambled siRNA (Dharmacon) was used. To determine the efficiency of siRNA transfection, RT-PCR and Western blot analyses were performed to detect PEDF expression level at 5 d posttransfection. To examine the role of PEDF in cell proliferation and viability, transfected cells were plated for MTT and apoptosis assays 24 h after siRNA transfection.
Semiquantitative RT-PCR
Total RNA was isolated from cultured cells using the TRIzol reagent (Invitrogen) according to the manufacturer’s procedure. First strand cDNA synthesis was performed from 2.5 µg total RNA using SuperScript Reverse Transcriptase (Invitrogen). cDNA was amplified in a 15-µl PCR mixture containing 1 mM dNTPs, 1x PCR buffer, 2.5 mM MgCl2, and 1 U DNA Taq polymerase (Promega, Madison, WI) with 5 pmol each primer for human PEDF (sense, 5'-CATTCACCGGGCTCTCTAC-3'; antisense: 5'-GGCAGCTGGGCAATCTTGCA-3'). The condition in the logarithmic phase of PCR amplification was as follows: 5 min initial denaturation at 94 C, 1 min denaturation at 94 C, 35 sec annealing at 67 C, and 1.5 min extension at 72 C for 30 cycles. The number of amplification cycles during which PCR product formation was limited by template concentration was determined in pilot experiments. PCR products were analyzed by agarose gel electrophoresis. ß-Actin was used as the internal control (sense, 5'-TCACCGAGGCCCCTCTGAACCCTA-3'; antisense, 5'-GGCAGTAATCTCCTTCTGCATCCT-3'). The reproducibility of the quantitative measurements was evaluated by three independent cDNA syntheses and PCR amplification from each preparation of RNA. Densitometric analysis was performed using Scion Image software (Scion Corp., Frederick, MD), and the relative PEDF mRNA expression levels were determined as the ratio of the signal intensity of PEDF to that of ß-actin.
Real-time PCR
Real-time PCR was performed using the iCycler iQ Real-Time detection system and the IQ SYBR Green Supermix (Bio-Rad). PCR primers for PEDF were as described above. ß-Actin was analyzed in the same run as internal control. Fluorescent measurements were recorded during each annealing step. The PCR quality and specificity were verified by melting curve analysis and gel electrophoresis. For quantitative comparison, the relative abundance of the PEDF mRNA was determined by dividing the threshold of each sample by the threshold of the internal control ß-actin. These experiments were carried out in duplicate and independently repeated three times.
Western blot analysis
Cellular extracts were prepared using RIPA [1% Triton X-100, 50 mM Tris-HCl (pH 7.4), 0.1% SDS, 150 mM NaCl, and 5 mM EDTA, supplemented with protease inhibitors containing 1 mM phenylmethylsulfonyl fluoride, 1 µg/ml aprotinin, 1 µg/ml leupeptin, and 1 µg/ml pepstatin A]. Conditioned media were collected and clarified by centrifugation. Total protein content in each sample was determined using the Bradford assay (Bio-Rad). Forty micrograms extracted proteins from each cell culture or 100 µg total proteins from the conditioned media was separated by 7.5% SDS-PAGE and then transferred to nitrocellulose membrane. Membranes were blocked using 5% nonfat dry milk in PBS containing 0.05% Tween 20 and incubated overnight at 4 C with rabbit polyclonal human PEDF antibody (1:1000) (Bioproducts) or rabbit polyclonal ß-actin antibody (1:1000) (Sigma) as a loading control. For conditioned media, parallel gels were stained with Coomassie blue to confirm equal loading. The immunocomplex was detected with horseradish peroxidase-conjugated antirabbit IgG (Bio-Rad) and visualized using an enhanced chemiluminescence detection system (Amersham Biosciences, Little Chalfont, UK).
MTT assay
Cell viability was assessed colorimetrically by the mitochondria-dependent reduction of MTT (Sigma) to formazan (31). Cells were seeded at 2000 per well in 96-well plates. After a 5-d incubation period, 10 µl MTT was added to each well, and plates were incubated at 37 C for 4 h. Finally, culture medium was aspirated, and 100 µl DMSO was added to extract the dye. Conversion of MTT to formazan by metabolically viable cells was monitored by the absorbance of each well at 570 nm with 630 nm as the reference wavelength. Relative cell viability was expressed as the fold change over control cultures. Assays were performed in triplicates, and data points represent the mean values ± SD from three independent experiments.
Terminal deoxynucleotidyl transferase-mediated nick-end labeling (TUNEL) staining
TUNEL assay was performed using an in situ cell death detection kit (Roche Biochemical, Indianapolis, IN) per the manufacturer’s instructions. This system end-labels the fragmented DNA of apoptotic cells. Omission of the enzyme in the TUNEL reaction was used as a negative control, and cells treated with DNAse I were used as a positive control. The number of TUNEL-positive cells was counted in five different fields under a light microscope at x40 magnification, and representative fields were photographed.
Hoechst staining
To confirm the findings of TUNEL assay, apoptotic cells were also detected by Hoechst staining. Briefly, cells were trypsinized and fixed with Carnoy’s fixative (methanol/acetic acid, 1:3) for 10 min at room temperature. After two PBS washes, DNA-specific fluorochrome Hoechst 33342 (Sigma) was added at a final concentration of 5 µg/ml, and the suspension was incubated at room temperature for 30 min in the dark. Cell suspensions were mounted on slide glasses and subjected to immunofluorescence microscopic examination. Apoptotic cells were identified by chromatin condensation and nuclei fragmentation. The percentages of apoptotic cells were calculated from the ratio of apoptotic cells to total cells counted. At minimum, 500 cells were counted in five different fields, and assays were performed in duplicate at least three times.
Promoter constructs and luciferase assay
The PEDF promoter region between positions –864 and +63 was cloned into promoterless luciferase reporter vector pGL3-Basic as previously described (32). Cells were transiently transfected with 1 µg of the construct using Lipofectamine 2000 (Invitrogen) as directed by the manufacturer. After 6 h, the cells were incubated with or without E2 (100 nM) for an additional 24 h before being harvested for luciferase activity measurement. The pSV-ß-galactosidase plasmid was cotransfected as internal control. Luciferase units were calculated as luciferase activity/ß-galactosidase activity and are presented as the mean ± SD of three individual experiments with triplication. The fold change was calculated by comparison with the promoterless luciferase vector (pGL3-Basic).
Statistical analysis
Every experiment was repeated at least twice in either duplicate or triplicate with different cell preparations to ensure consistency of the findings. Statistical analysis was carried out using Fisher’s exact test and ANOVA followed by Tukey’s post hoc test (GraphPad Software, San Diego, CA) where applicable. P < 0.05 was considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, a; and Table 1). In benign serous and mucinous cystadenomas, a comparable strong epithelial expression was detected (Fig. 1A, b; and Table 1). Of these, six showed strong expression of PEDF (46.1%), and five showed moderate expression (38.5%). In contrast, tumors of borderline malignancy exhibited no or only weak expression level of PEDF (Fig. 1A, c), and loss of PEDF was confirmed in 22 of 24 tumors (91.7%) (Fig. 1A, d–f; and Table 1). There was no significant correlation between the PEDF expression and the clinical data of the tumor histological type (P = 0.74). Omission or substitution of the primary antibody with preimmune serum did not produce any staining (data not shown). We also performed real-time PCR to verify our immunohistochemistry results with ovarian cancer tissue specimens (OC-1–14) and normal ovarian epithelium scrapings (OSE-1) or culture of OSE scrapings (OSE-2 and -3). As shown in Fig. 1B, PEDF mRNA expression was significantly higher in normal OSE (n = 3) compared with ovarian cancer (n = 14). No significant difference was seen in uncultured and cultured OSE brushings, in which all cases had relatively high levels of PEDF expression. Together, these results suggest that PEDF protein diminishes with ovarian neoplastic progression.
|
|
) and Western blot analyses (Fig. 2B). In contrast, the levels were minimal to absent in ovarian cancer cell lines Caov-3 and SKOV-3, suggesting that the decreased PEDF protein expression may result from decreased mRNA synthesis. Media conditioned by these cells were tested for PEDF expression by Western blot. Equal loading was confirmed by parallel gels stained with Coomassie blue (Fig. 2C). Consistently, OSE cells secreted the most PEDF, and ovarian cancer cells secreted much less to no detectable PEDF. The presence of a specific 50-kDa band indicated that PEDF was appropriately synthesized and processed.
|
) (P < 0.05). The growth-inhibitory effect of rPEDF was dose dependent because the effect (decreased by 42.2, 61.2, and 78.9% respectively) increased with concentration of treatment (1, 10, and 100 nM) (Fig. 3A). The antiproliferative effect of rPEDF was blocked by the addition of antibodies specific to PEDF (P < 0.05), confirming the effect of PEDF was specific (Fig. 3A). The antibody by itself had no effect on cell number (Fig. 3A). The ability of rPEDF to decrease cell number could result from either decreased cell cycle progression or increased apoptosis. Our analyses did not detect a significant change in cell cycle progression as assessed by propidium iodide staining (data not shown); however, exposure to rPEDF resulted in a concomitant increase of apoptosis with 55.7% of the cells being TUNEL positive (at 100 nM rPEDF) when compared with 16.8% in untreated controls (Fig. 3B). The apoptotic activity was greatly reduced by the addition of PEDF antibodies. The ability of rPEDF to induce apoptosis was examined also by counting Hoeschst-stained cells with fragmented DNA. Similar to the TUNEL staining experiments, a significant increase in rPEDF-induced apoptosis was observed (Fig. 3B).
RNA interference assay was conducted to further verify the role of endogenous PEDF in the regulation of cell proliferation and viability. siRNA oligos that specifically target human PEDF were transiently transfected into OSE cells. As revealed by real-time PCR and Western blot analyses (Fig. 3C), transfection of PEDF siRNA efficiently repressed PEDF expression by over 80% at both mRNA and protein levels, whereas nonspecific siRNA had no effect. Importantly, inhibition of PEDF expression by RNA interference resulted in a significant increase (98.6% increase) in cell proliferation (Fig. 3D). In addition, PEDF siRNA suppressed apoptosis (Fig. 3E). No inhibition was observed for nonspecific siRNA (Fig. 3E).
To examine whether PEDF might also target tumor epithelial cells, we treated Caov-3 and SKOV-3 ovarian cancer cells with rPEDF proteins in vitro. As shown in Fig. 4A, rPEDF at 100 nM substantially decreased cell proliferation in both cancer cell lines (P < 0.05) when compared with results obtained with untreated controls, and this response was blocked by treatment with an antibody against PEDF (Fig. 4A). We also observed a 3.5-fold increase in the rate of apoptosis in rPEDF-treated cells, with two apoptosis assays confirming the same finding (P < 0.005) (Fig. 4B). Consistent with these results, siRNA-mediated down-regulation of PEDF increased proliferation and decreased apoptosis of Caov-3 and SKOV-3 cells (Fig. 4, C–E). No effects were observed for nonspecific siRNA (Figs. 4, C–E). These findings indicate a functional role for PEDF in OSE and cancer cell growth and survival.
|
, a). The inhibition was dose-dependent, with more than 90% decrease at 100 nM. This significant down-regulation of gene expression at the mRNA level is consistent with the changes in protein levels (Fig. 5A, b and c). To examine the specificity of the effects imposed by E2, the studies were repeated in the presence of a pure ER antagonist, ICI. Treatment with increasing concentrations (10 nM, 100 nM, or 1 µM) of ICI attenuated E2-induced down-regulation of PEDF mRNA expression by 20, 60, and 100%, respectively, reflecting the mediation of E2 effect through classical ER (Fig. 5B, a). The compound alone did not alter PEDF mRNA expression, even at a concentration (1 µM) that caused complete abolition of the E2 effect (Fig. 5B). Similar results were obtained by Western blotting with antibodies to PEDF (Fig. 5B, b and c).
E2 directly regulates PEDF transcription
The down-regulation of PEDF was time-dependent; mRNA was dramatically reduced by 50% at 3 h, tapering gradually afterward, and reaching a maximum at 24 h after E2 treatment (Fig. 6A). Decreased PEDF mRNA levels lasted at least 48 h after E2 addition (Fig. 6A). This rapid effect of E2 on PEDF mRNA levels suggests that E2 could directly regulate the expression of PEDF.
To evaluate whether the observed decrease in PEDF mRNA expression was a direct effect of E2, cells were stimulated with E2 in the absence or presence of the protein synthesis inhibitor CHX. Treatment of OSE cells with E2 alone or in combination with CHX resulted in a strong reduction of PEDF mRNA expression, suggesting that the inhibition of PEDF gene expression by E2 was not dependent upon de novo synthesis of an intermediate protein (Fig. 6B). Next, to test whether the effect of E2 on PEDF mRNA expression was the result of decreased mRNA stability upon E2 treatment, we performed ActD chase experiments to determine the half-life and stability of PEDF mRNA. OSE cells were preincubated with 100 nM E2 for 3 h and then treated with ActD. Total RNA was isolated at 0, 1, 2, 4, 6, and 8 h after the addition of ActD and analyzed by semiquantitative RT-PCR. The apparent half-life of PEDF mRNA was approximately 2–3 h in the absence of E2 and was not changed upon E2 treatment (Fig. 6C). Thus, the decreased PEDF mRNA level in response to E2 was not due to an effect on PEDF mRNA stability.
To assess whether E2 causes transcriptional repression at the PEDF promoter, we transfected OSE cells with a luciferase reporter construct of the PEDF promoter region +63 to –864 and measured luciferase activity after treatment with 100 nM E2. The luciferase reading of the promoterless pGL3-Basic vector was arbitrary shown as 1. The results showed that the activity of PEDF promoter strongly reduced (10-fold) after the addition of E2, which could be reversed by the addition of ICI (Fig. 6D) (P < 0.05 compared with untreated control). These results demonstrate a direct effect of E2 on the human PEDF gene promoter.
PEDF abrogates the proliferative effect of E2 on OSE cells
To evaluate the functional role of E2 in OSE cells, the cells were treated with E2 at a dose range between 0.01 and 100 nM for 5 d and assessed by MTT dye conversion. As shown in Fig. 7A, a dose-dependent increase in cell growth was clearly observed in E2-treated OSE cultures, with an approximately 3-fold increase in cell growth at 100 nM E2. Exposure of cells to ICI abolished the E2-induced cell growth enhancement (Fig. 7B). ICI by itself had no effect on cell growth (Fig. 7B). TUNEL staining showed significant decreases of apoptosis in E2-treated cells (5.3%) compared with untreated controls (21.8%) (at 100 nM E2) (P < 0.005) (Fig. 7C). The E2-induced apoptosis was found to be dose-dependent and reversible by cotreatment of cells with ICI (Fig. 7C). The percentage of apoptotic cells determined by Hoechst nuclear staining is approximately the same as determined by TUNEL assay (Fig. 7C).
To demonstrate that the regulation of PEDF expression could explain some of the effects of E2 on OSE cell growth, we tested the ability of E2 to trigger proliferation of OSE cells in the absence or presence of rPEDF proteins. Cotreatment of cells with E2 and rPEDF blocked the estrogenic action on growth stimulation, suggesting a negative role of PEDF in estrogen-induced OSE cell proliferation (Fig. 7, B and C).
Effect of PEDF on estrogen-induced cell proliferation in ovarian cancer cells
Next, we asked whether the effect of estrogen on the regulation of PEDF seen in OSE could also be found in ovarian cancer cells. As shown in Fig. 8A, E2 treatment decreased the expression of PEDF mRNA in the ER-positive Caov-3 cells (2.25-fold). This effect was blocked by ICI (Fig. 8A). In contrast, E2 did not affect PEDF mRNA expression in the estrogen-insensitive SKOV-3 cell line (20, 33) (data not shown). To determine the effects of PEDF on estrogen-induced cell proliferation, we performed MTT assays. The results indicated that treatment of Caov-3 cells with E2 caused an approximately 4-fold increase in cell proliferation, which could be reversed by the addition of ICI (P < 0.05) (Fig. 8B). The presence of rPEDF proteins completely attenuated the proliferative effect of estrogen, suggesting the growth-inhibitory role of PEDF in modulating the mitogenic action of estrogen in ovarian cancer cells. Parallel TUNEL and Hoechst staining demonstrated a decrease in number of apoptotic cells (from 13.7 to 3.8%) upon estrogen treatment (Fig. 8C). This effect was blocked by ICI and rPEDF.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The cause for the loss of PEDF expression in cancer cells is unknown, but it appears that estrogen may be one of the mechanisms in this down-regulation. We showed here that E2 treatment resulted in a decrease of PEDF in both OSE and ovarian cancer cell lines and that this repression was reversed by the addition of a specific ER antagonist ICI, suggesting that ER is involved. It is not known whether this is a mechanism common to all ER-expressing ovarian tumors. However, in the study of ovarian cancer cell lines, at least one (SKOV-3) that is ER positive but estrogen insensitive (20, 33) showed greatly reduced constitutive expression of PEDF, suggesting that other mechanisms might also contribute to the loss of PEDF expression. Notably, the PEDF gene was mapped at the 17p13.3, a chromosomal region suspected to harbor tumor suppressor genes and recurrently deleted in ovarian carcinomas (19, 34, 35), but whether the ovarian carcinoma cells and cancer lines that we studied exhibit allelic loss or somatic mutation of the PEDF gene has yet to be defined. Epigenetic changes such as promoter hypermethylation may be an alternative mechanism (our preliminary observation). It is also possible that other factors (cytokines and growth factors) may modulate PEDF expression. Irrespective of the underlying mechanisms, the findings of frequent loss of PEDF expression in epithelial ovarian carcinomas suggest that this change in PEDF may confer a selective advantage to the cells and could be an important event in the pathogenesis of ovarian cancer.
The OSE that surrounds the ovary is the source of epithelial ovarian carcinomas. Our results indicate that OSE produces a large amount of PEDF. Moreover, we show that PEDF is a potent growth inhibitor and/or inducer of apoptosis in OSE cells. Although the regulation of in vivo PEDF activity in the ovary has not been established, it is tempting to speculate that the OSE’s response to PEDF is intimately linked to the apoptosis and mitotic activity of OSE before and after ovulation. In this regard, our data would suggest that lower levels of PEDF, which are found in response to high levels of estrogens released with the follicular fluid at the time of ovulation, may play a role in the survival and/or proliferation of OSE cells next to the rupture site (1). Conversely, high levels of PEDF, which are present before ovulation, may induce apoptosis in OSE cells surrounding the apex of the follicle to facilitate the release of oocyte (2). The process of apoptosis is also critical in eliminating OSE that are sequestered in inclusion cysts, thereby removing potential sites of ovarian tumors (36). Therefore, one could speculate that the loss of PEDF in these cells, and the subsequent survival of damaged cells, could increase the risk of developing cancer. Intriguingly, PEDF is also found to be a potent inducer of apoptosis in ovarian cancer cells. Although the number of cell lines being examined is too small for definitive conclusion to be drawn, the fact that PEDF is equally potent to induce apoptosis in both estrogen dependent (Caov-3) and independent (SKOV-3) ovarian cancer cell lines would suggest this protein has potential as new therapeutic strategies for treatment of estrogen-resistant ovarian tumors. A 32-bp deletion in exon 1 of ER transcript has been found in the SKOV-3 cell line, which is ER positive but insensitive to E2 with respect to cell proliferation and induction of gene expression (20, 33). This may explain the lack of responsiveness and resistance to E2 in some ovarian cancer. Of particular interest, our findings may also be pertinent to human breast cancer. In ER-negative breast cancer cells, exogenous PEDF also reduces cell proliferation (our unpublished data), supporting a general growth suppressor role for this protein, in addition to its well-known effect on the tumor vasculature.
Estrogen is a major regulator of growth and differentiation in normal ovaries. Despite intensive studies on estrogen-induced growth, very little is known about potential estrogen-regulated genes in ovarian tissue (21, 26, 27, 28, 29, 37). In the present study, we demonstrate that E2 lowers the expression of PEDF in OSE and ovarian cancer cells in vitro. Furthermore, this newly discovered effect of E2 appears to confer these cells with a growth advantage because in the presence of exogenous PEDF, their mitogenic response to E2 is attenuated. Thus, the present findings not only confirm others, which indicate that E2 stimulation of OSE and tumor cell proliferation (21, 26, 27, 28, 29, 37), but also reveal its possible interaction with PEDF. Interestingly, we observed that growth factors such as hepatocyte growth factor and epidermal growth factor could also regulate the expression of mRNA for PEDF and at least in part counteract the growth-inhibitory effects of PEDF on OSE (data not shown), suggesting that other mitogenic signaling pathways may converge with PEDF to modulate its effects. These factors may also be present and act to inhibit PEDF signaling in vivo, contributing to the mitogenic activity of OSE after ovulation. Indeed, the level of hepatocyte growth factor is highest in the late follicular phase and during the luteal phase (38). Although the combined effects of growth factors on OSE cell growth are starting to be investigated, it has been known for a long time that expression and action of one factor may affect the expression of other factors and the cellular response of OSE at any one time depends on the combined influences of numerous growth factors (39).
In other cell types, E2 has been shown to regulate transcriptional activation of downstream genes as well as transcript stability (40). Our current findings indicate that E2 decreases PEDF mRNA in human OSE cells through a predominantly transcriptional mechanism. 1) There is a rapid decrease of mRNA level within 3 h after E2 treatment. 2) The use of ActD suggests that the observed changes in mRNA are due to E2-regulated mRNA synthesis rather than mRNA stability. Pretreatment of cells with CHX has no effect on the E2-mediated decrease of PEDF mRNA expression, suggesting that E2 effect does not require de novo protein synthesis. 3) Furthermore, E2 causes transcriptional repression at the PEDF promoter, which could be reversed by ICI, indicating the regulation is ER mediated. This is the first demonstration that PEDF is a target gene of ER activation. Investigations are in progress to identify and characterize the mechanisms responsible for the E2-mediated transcriptional repression of PEDF.
In summary, we have shown that PEDF is consistently down-regulated in ovarian cancer. The identification of PEDF as a new ER-regulated target gene implicates a novel mechanism to regulate PEDF expression. Furthermore, our studies are the first to establish a role of PEDF in normal OSE and tumor growth by influencing cell proliferation and apoptosis. This highlights the potential of PEDF as a tumor suppressor and may provide a target for therapeutic intervention in ovarian cancer.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
The authors have nothing to disclose.
First Published Online June 15, 2006
Abbreviations: ActD, Actinomycin D; CHX, cycloheximide; E2, 17ß-estradiol; ER, estrogen receptor; FBS, fetal bovine serum; ICI, ICI 182,780; MTT, 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bromide; OSE, ovarian surface epithelium; PEDF, pigment epithelium-derived factor; rPEDF, recombinant PEDF; siRNA, small interfering RNA; TUNEL, terminal deoxynucleotidyl transferase-mediated nick-end labeling.
Received February 9, 2006.
Accepted for publication June 7, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and -ß, progesterone receptor, and androgen receptor mRNA in normal and malignant ovarian epithelial cells. Proc Natl Acad Sci USA 96:5722–5727
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
