| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of Reproductive Biology (Q.W., X.L., R.Z., G.G.), Physiology and Biophysics (G.G.), and Oncology (G.G.), Case Western Reserve University School of Medicine, Cleveland, Ohio 44106; and Department of Pharmacology (L.W., Y.-H.F.), Uniformed Services University of the Health Sciences, Bethesda, Maryland 20814
Address all correspondence and requests for reprints to: George I. Gorodeski, M.D., Ph.D., University MacDonald Women’s Hospital, University Hospitals of Cleveland, 11100 Euclid Avenue, Cleveland, Ohio 44106. E-mail: gig{at}cwru.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Growth of ectocervical epithelial cells is maintained by a balance between proliferation of the cells in the basal layer and death of cells in upper layers of the epithelium. Cells in the basal layer can either replicate or cease proliferation and exit from the mitotic cycle, stratify, and undergo terminal differentiation into cornified envelopes (superficial cells) (1). We recently reported that growth of human cervical epithelial cells (hECEs) is also regulated by apoptosis (2, 3), but the mechanisms and regulation of the apoptosis are unknown. Because apoptosis functions to eliminate abnormal cells (4), and dysregulation of apoptotic cell-death has been implicated in the premature death of cells, loss of tissue, aging, states of disease, and neoplastic transformation (5), elucidation of these data could be important for our understanding of cervical cell biology and tumorigenesis.
Most previous studies of apoptosis in the female lower genital tract looked into effects in cancer cells. Only few investigated regulation of apoptosis in normal human cervical-vaginal cells, but these studies provide conflicting results relative to the role that apoptosis could have for cancer protection (e.g. Refs.6 , 7). We have recently shown that in normal hECEs, activation of the P2X7 receptor induces apoptosis. The P2X7 receptor is a member of the P2X receptor subfamily of the ATP-dependent P2 nucleotide receptors, which are membrane-bound, ligand-operated K+-, Na+-, Ca2+-permeable channels. Studies in other types of cells showed that ligation of the P2X7 receptor can initiate signaling cascades that lead to apoptosis (reviewed in Ref.8). In cervical cells the submaximal effect occurred at supraphysiological concentrations of ATP, but the threshold of the effect was about 0.5 µM, which is at the range of extracellular ATP activity in cervical cultures (3). Interestingly, blocking the effect of endogenously secreted ATP lowered the baseline apoptosis (3), suggesting ATP-depended autocrine-paracrine regulation of apoptosis in the cervix in vivo by activation of the P2X7 receptor mechanism.
Previous studies suggested that estrogen exerts an apoptosis protective effect (e.g. Refs.9, 10, 11). However, some of these studies used cancer cells (e.g. Ref.12) and looked into apoptosis that was induced with noxious stimuli or under toxic conditions (e.g. Ref.13). We recently reported that in normal hECEs, treatment with estrogen abrogates baseline apoptosis as well as apoptosis induced by treatment with ATP or the P2X7 receptor-specific agonist 2',3'-0-(4-benzoylbenzoyl)-ATP (BzATP) (2, 3). Whereas the mitogenic effect of estrogen on cervical cells is well documented (1), these preliminary data suggest that estrogen regulates growth of cervical cells by preventing apoptosis, independent of its mitogenic potential. At present relatively little is known about the mechanisms that are involved in the estrogen effect in the cervix and to what degree normal and cancer cervical cells share similar mechanisms of action. The objectives of the present study were to understand the pharmacodynamics of estrogen antiapoptotic effect in the cervix and the mechanisms involved. Our results show a more efficient antiapoptotic effect of estrogen in normal cells, compared with cancer cervical cells, and involvement of disparate apoptosis-protective mechanisms in the two types of cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Preliminary experiments were conducted on cells plated on culture plates, and definitive experiments were repeated on cells plated on filters. The latter method improves culturing conditions and promotes epithelial cell polarization and differentiation (14). Anocell-10 filters (Anocell, Oxon, UK, obtained through Sigma Chemicals, St. Louis, MO) are ceramic-base filters with a pore size of 0.02 µm width and 50 µm depth. Filters were coated on their upper (luminal) surface with 3–5 µg/cm2 collagen type IV and incubated at 37 C overnight. The remaining collagen solution was aspirated, and the filter was dried at 37 C. Before plating, both sides of the filters were rinsed three times with Hanks’ balanced salt solution. Cells were plated on the upper surface of the filter at 3 x 105 cells/cm2. By plating at this high density, the cultures become confluent and polarized within 12 h after plating (14).
In some experiments cells were shifted to steroid-free medium to deprive cells of estrogens. This medium is composed of phenol-red-deficient DMEM/Ham’s F12 or RPMI 1640 (Sigma) containing 8% heat-inactivated fetal bovine serum that was previously treated with charcoal to remove steroids. Preparation of charcoal-treated serum was described (17); briefly, dextran-coated charcoal (Sigma) was dissolved at 8% in 0.15 M NaCl, autoclaved, mixed by stirring, spun, and the pellet resuspended as 1 g/1.25 ml in H2O. Fetal bovine serum (Hyclone, Logan, UT) was mixed with the activated charcoal-dextran at 20:1 (vol/vol) and incubated for 45 min at 55 C. At the completion of incubation, the mixture was spun twice at 800 x g for 20 min, and the supernatant (serum) was decanted and collected. Some assays were carried out in serum-free medium. Cells were shifted to modified Ringer’s solution [composed of 120 mM NaCl, 1.2 mM CaCl2, 1.2 mM MgSO4, 5 mM KCl, 10 mM NaHCO3, 10 mM HEPES, 5 mM glucose, and 0.1% BSA (pH 7.2)]. Levels of extracellular Ca2+ were controlled by adding EGTA. All treatments involved adding drugs to both the luminal and subluminal solutions.
DNA fragmentation assay
The assay was modified from Lizard et al. (18). Cells attached on plates were harvested and combined with floaters recovered from the medium. Cells were washed in ice-cold PBS lacking Ca2+ and Mg2+, resuspended in the same medium, and cellular DNA extracted using DNA extraction kit (Stratagene, La Jolla, CA). In some experiments the latter step was done by using lysis buffer [composed of 10 mM EDTA, 400 mM NaCl, 1 mg/ml proteinase K, 35 mM sodium dodecyl sulfate (SDS), and 10 mM Tris-HCl (pH 8.2)]; cells were spun briefly at 180 x g, the pellet was resuspended in the lysis buffer, transferred to an Eppendorf tube, and incubated overnight at 37 C. The tubes were spun for 5 min at Eppendorf high speed, and the supernatant containing the DNA precipitated in 2 volumes of 100% ethanol and incubated overnight at –20 C. Tubes were spun for 5 min at Eppendorf high speed, and the pellet containing the DNA was saved. For assays, the DNA-containing pellets were resuspended in 100 µl Tris/EDTA buffer [composed of 10 mM Tris-HCl, 0.2 mM Na-EDTA (pH 7.5)], and equal amounts of DNA (determined by spectrofluorometry) were separated on 1.8% agarose gel electrophoresis for 15 h at 20 V. Gels were prepared in Tris-buffered ethanolamine buffer [9 mM Tris-borate (pH 8), 2 mM EDTA] plus 0.1 µg/ml ethidium bromide and photographed under UV light.
DNA solubilization assay
Twenty-four hours before the end of the experiment, cells were labeled with [3H]thymidine (specific activity 100 Ci/mmol; 5 µCi/1 x 106 cells) for about 15 h. The medium was removed, cells were washed three times with fresh medium lacking [3H]thymidine, and incubated in the same medium for an additional 9 h in the absence or presence of ATP or BzATP. At the end of incubations, the conditioned medium was stored, cells were lysed in 0.5 ml of lysis buffer [10 mM Tris-HCl (pH 7.5) plus 1 mM EDTA-0.2% Triton X-100] for 1 h at 4 C. The intact chromatin was separated from the fragmented DNA by 5 min centrifugation at 4 C in Eppendorf microcentrifuge at 12,000 x g. The supernatant (referred to as lysate) was stored, whereas the pellet was resuspended in 0.5 ml of 1% SDS. The radioactivity contained in the supernatants lysates and pellets was counted in a scintillation counter, and DNA fragmentation was calculated as [percent solubilized DNA] = ([cpm supernatant + cpm lysates]/[cpm supernatant + cpm lysate + cpm pellet]) x 100.
Western blot analysis
The postnuclear supernatant of cells was solubilized in lysis buffer [50 mM Tris-HCl (pH 6.8), 1% [(3-cholamidopropyl) dimethyl-ammonio]-1-propane-sulfonate, 5 mM EDTA (pH 8.0)] containing 50 µg/ml phenylmethylsulfonyl fluoride, 10 µg/ml benzamidine, 10 µg/ml bacitracin, 10 µg/ml leupeptin, and 2 µg/ml aprotinin. Aliquots normalized to 15 µg protein (about 45 µl), determined by protein assay solution (Bio-Rad Laboratories, Hercules, CA) were loaded on 10% polyacrylamide SDS gel, and vertical electrophoresis was conducted at 50 mA for 1.5 h. Gels were transferred onto Immobilon membrane (Millipore, Bedford, MA) at 200 V for 1.5 h, and membranes were blocked in 5% milk and exposed to the primary antibody at 4 C overnight. Membranes were washed thrice in PBS and fluorescent stained for 1 min using an enhanced chemiluminescence kit of peroxidase-conjugated secondary antibody from Amersham (Piscataway, NJ).
Isolation of RNA and RT-PCR was done as described (19). Total RNA from cultured cells was isolated with a kit (Qiagen Inc., Chatsworth, CA), using lysis buffer plus ß-mercaptoethanol at 350 µl/107 cells. The final total RNA pellets were resuspended in 50 µl diethyl pyrocarbonate-water and quantitated by measuring OD260. RT-PCR for Bcl-2 and Bax mRNA assays was done using 1.5 µg total RNA. Primer pairs were designed against Bcl-2 (forward, 5'-TCCATTATAAGCTGTCACAG-3' and reverse, 5'-GAAGAGTTCCTCCACCAC-3', GenBank accession no. U34964, 1 min at 94 C, 1 min at 55 C, and 1 min at 72 C, 35 cycles) and against Bax (5'-GCAGAGGATGATTGCTGATG-3' and 5'-CTCAGCCCATCT TCTTCCAG-3', GenBank accession no. NM 017059, 1 min at 94 C, 1 min at 55 C, and 1 min at 72 C, 35 cycles). To eliminate the measurement error from uneven sample loading and provide a semiquantitative measure of the relative changes in gene expression, RT-PCR signals were normalized to the 18S signal of the corresponding reverse transcription product. Preliminary experiments were also conducted with each gene to ensure that the number of cycles (35 cycles) represented a linear portion for the PCR OD curve for the cervical samples. All PCR products were verified by restriction digestion and by sequencing.
Mitochondria were isolated as described (20) (all steps performed at 4 C). Cells harvested off 20 50-cm2 dishes were washed in low protein buffer [150 mM NaCl, 20 mM glucose, 20 mM sodium phosphate (pH 7.4)]; resuspended at a density of 109/ml in sucrose buffer, composed of sucrose (250 mM), KCl (100 mM), mannitol (75 mM), Tris/HCl (10 mM), and EDTA (2 mM) plus 10 µl protease inhibitor cocktail (pH 7.4) (Roche, Indianapolis, IN); and sonicated at 800 x g for 10 min. Broken cells were spun at 2000 x g for 10 min and at 17,000 x g for 15 min. The pellet was washed and resuspended in mannitol buffer composed of mannitol (220 mM), sucrose (100 mM), NaCl (2 mM), EGTA (0.1 mM), KCl (135 mM), dithiothreitol (1 mM), 3[N-morpholino]propanesulfonic acid (20 mM), K2HPO4 (20 mM), and MgCl2 (5 mM) (pH 7.0); purified on a Percoll gradient to a concentration of 10 µgP/µl; and maintained at –80 C.
Mitochondrial swelling, which reflects formation of the mitochondrial permeability transition pores (21), was recorded as loss of absorption at 540 nm as described (22). Mitochondria (10 µg) were resuspended in 90 µM mannitol buffer at 37 C and energized with 5 mM glutamate (to provide electrons to complex I of the respiratory chain) and 5 mM succinate (to feed electrons into complex II). Before experiments an aliquot was used to determine the mitochondrial transmembrane potential after incubation with 1 µM tetramethylrhodamine methylester and measuring stability of fluorescence at 590 nm wavelength (23).
The method of fluorescence experiments of attached cells was described (24). Cells were grown on filters and incubated with 5–7 µM fura-2/acetoxymethyl ester plus 0.25% Pluronic F12. Measurements of fura-2 fluorescence were made in a fluorescence chamber by measuring light intensity at 340, 360, and 380 nm excitation and 510 nm emission wavelengths, and changes in cytosolic calcium Cai were calculated as described (24). The experimental conditions provide for continuous perfusion of the luminal and subluminal compartments at rates of 1–1.5 ml/min. Agents and solutions were added to the luminal and subluminal solutions.
Densitometry was done using Arcus II scanner (AGFA, New York, NY) and Un-Scan-It gel automated digital software (Silk Scientific, Orem, OR).
Antibodies
Mouse monoclonal antihuman caspase-3, -8, and -9 antibodies were from Chemicon (Temecula, CA). Mouse monoclonal antihuman ß-actin antibody was from Zymed Laboratory Inc. (San Francisco, CA). Mouse monoclonal antihuman Bcl-2 and Bax were from Santa Cruz Biotechnology (Santa Cruz, CA). The antibodies were used according to the manufacturers’ instructions.
Statistical analysis of the data
Data are presented as means (± SD), and significance of differences among means was estimated by Student’s t test. Trends were calculated using GB-STAT V5.3 (Dynamic Microsystems Inc., Silver Spring, MD) and analyzed with two-way ANOVA.
Chemicals and supplies were obtained from Sigma unless specified otherwise.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). P2X7 receptor-dependent apoptosis was induced by treatments with the P2X7 receptor-specific ligand BzATP or the naturally occurring ligand ATP, using concentrations that produce near-maximal effect (100 and 250 µM, respectively) (3). Treatment with BzATP (Fig. 1A) or ATP (Fig. 1B) induced DNA fragmentation in normal hECEs as well as hECE immortalized with HPV-16 (ECE16–1) and the human cancer cervical epithelial cells HT3, CaSki, SiHa, and Hela.
|
). To determine the effect of estrogen on baseline apoptosis, cells were shifted to steroid-free medium to deprive cells of endogenous estrogens. Incubation in steroid-free medium had little additional effect on baseline apoptosis (not shown). Treatment of cells grown in steroid-free medium with 17ß-estradiol at the physiological concentration of 10 nM decreased significantly baseline DNA solubilization in hECEs from 2.5 to 1.5% (P < 0.02, Fig. 2). Estradiol had no effect on baseline apoptosis in ECE16–1, HT3, CaSki, SiHa, or Hela cells (Fig. 2).
|
and 2) or BzATP (see Fig. 4 below) increased apoptosis, but the effect was significantly greater in hECEs and ECE16–1 cells than the cancer cells (Fig. 2). In hECEs and ECE16–1 cells, ATP increased DNA solubilization by 4.7 and 4.8%, respectively, whereas in the cancer cells by only 1.8–3.3% (P < 0.01–0.05, Fig. 2). In all six types of cells, the apoptosis induced by ATP could be blocked by incubation in low extracellular calcium (Fig. 2), suggesting that the apoptosis induced by activation of the P2X7 receptor is mediated by calcium influx and by increases in cytosolic calcium.
|
), indicating that estrogen blocked the apoptosis induced by ATP. Treatment with estradiol also attenuated apoptosis in ECE16–1 HT3 and CaSki cells [decreases in DNA solubilization of 3.4, 2.4, and 1.6%, respectively (P < 0.01, compared with baseline)], but DNA solubilization rates remained higher than in baseline conditions (Fig. 2). Estradiol had no effect on the ATP-induced apoptosis in SiHa and Hela cells (Fig. 2).
Pharmacodynamics of the estrogen effect
The experiments in this section used hECEs, which are a well-characterized model of normal cervical epithelial cells (14), and the cancer CaSki cells, which retain phenotypic characteristics of the cervical epithelium including functional estrogen receptors-
and -ß (25). The relatively small magnitude of changes in baseline apoptosis precluded us from studying directly mechanisms involved in the estrogen modulation of baseline apoptosis. Instead, the experiments looked into estrogen modulation of P2X7 receptor-induced apoptosis because the latter is the main contributor to the baseline apoptosis (3).
Although the magnitude of the effects was different, the time course and modulation of estrogen effects were similar in hECEs and CaSki cells (Fig. 3, A and B). Pretreatment with 10 nM 17ß-estradiol abrogated the BzATP-induced apoptosis, but the hormone had to be administered at least together with the BzATP to block effectively the apoptosis. Earlier treatments had no additional effect to that of simultaneous administration of estradiol and BzATP, whereas treatments that began after BzATP were less effective (Fig. 3, A and B). Coadministration of actinomycin-D or cycloheximide attenuated the effect of estradiol (Fig. 3, A and B), indicating that the estrogen effect depends on transcription and on protein translation. The dose-response effects of estrogen were similar in hECEs and CaSki cells: the BzATP-modulating effects of 17ß-estradiol were observed already at 0.1 nM, and they reached maximum at about 10 nM (Fig. 3, insets).
|
-estradiol, the nonpermeable 17ß-estradiol-6-[o-carboxymethyl]oxime/BSA and aldosterone had no effect on the BzATP-induced apoptosis. Testosterone and hydrocortisone potentiated the apoptotic effect of BzATP (Fig. 4). The estrogen receptor modulators tamoxifen ICI-182780 and progesterone had no significant effect on baseline apoptosis when administered alone (not shown). Tamoxifen and progesterone potentiated the effect of BzATP, whereas cotreatment with 10 nM 17ß-estradiol attenuated the apoptosis but to a lesser degree than in cells treated with BzATP plus estradiol only (Fig. 5). ICI-182780 had no effect on the BzATP-induced apoptosis or on the responses to estradiol (Fig. 5).
|
) suggest genomic and protein-translational regulation. The specificity profile of estrogen action (Fig. 4) suggests involvement of estrogen-specific response elements. However, lack of appreciable estrogen antagonistic effect by tamoxifen ICI-182780 or progesterone (Fig. 5) does not support the involvement of the traditional nuclear estrogen receptors.
The responses to BzATP and estradiol do not depend on cell-cycle phase
Withdrawal of serum and prolonged quiescence (G0/G1 phase) can lead to apoptosis (26). Baseline apoptosis in hECEs depends on serum factors because shifting cervical cells to serum-free medium increases baseline apoptosis (3). In hECEs synchronization in the G0/G1 phase induced DNA solubilization of about 3.7% (Fig. 6), which was significantly higher than that observed in cells grown in regular medium (about 2%, P < 0.5) (3) and in S and G2/M phases (2.1%, P < 0.05, Fig. 6). Respective values in CaSki cells were 2.5 and 1.5% (3). Because estrogens stimulate proliferation and replication by inducing entry into cell cycle, we determined the degree by which the antiapoptotic effect of estrogen depends on exit from G0/G1 phase and entry into the G1, S, and G2/M phases. To this aim, preconfluent hECEs and CaSki cells on filters were grown in steroid-free medium and synchronized into different phases of the cell cycle as follows: for synchronization in the G0/G1 phase, cells were shifted to serum-free medium. To synchronize entry into cell cycle, cells were treated with epithelial growth factor (EGF), and for synchronization in G1 phase, lovastatin was added (27). For synchronization in S phase, hydroxyurea was added (28), whereas growth arrest in G2/M phase was obtained by treatment with EGF plus nocodazole (29). Cells were then treated with 100 µM BzATP plus 10 nM 17ß-estradiol for 9 h, and apoptosis was determined in terms of changes in DNA solubilization.
|
). Cotreatment with estradiol attenuated the effect of BzATP to the same degree, regardless of the cell-cycle phase (Fig. 6). Similar trends were obtained in CaSki cells (not shown). These results indicate that the antiapoptotic effect of estrogen does not depend on cell-cycle phase.
Estrogen blocks BzATP activation of caspase-9 and caspase-3
The execution of apoptosis involves activation of caspases: the key signaling caspase of the cell surface receptor pathway is caspase-8 and the mitochondrial pathway, caspase-9 (5). The cell surface receptor and mitochondrial-dependent pathways integrate at the level of the effector caspases, so that activation of caspases-9 and -8 triggers the nonreversible execution of apoptosis by caspases-6, -7 and -3 (5). In hECEs the apoptotic effect of P2X7 receptor activation is mediated mainly by the mitochondrial pathway (3). The next experiment tested the degree to which treatment with estrogen modulates BzATP-induced activation of caspases-8, -9, and -3. Caspases are synthesized as inactive zymogens and are converted to an active tetrameric complex composed of two heterodimeric subunits. On activation caspases are cleaved to smaller forms that can be detected by Western blots (30).
In hECEs treatment with BzATP-induced expression of the 10-kDa form of caspase-9, and it increased expression of the 17-kDa form of caspase-3 (Fig. 7), confirming that BzATP activates caspase-9 and the terminal caspase-3 (3). BzATP stimulated a mild increase in the expression of caspase-8, and it did not affect expression of the constitutive protein ß-actin (Fig. 7). Cotreatment with 17ß-estradiol blocked BzATP-induced expression of the caspase-9 10-kDa form and the caspase-3 17-kDa form without affecting the expression of caspase-8 or ß-actin (Fig. 7). Similar trends were also observed in CaSki cells (not shown). These results suggest that estrogen blocks the BzATP-induced apoptosis by modulation of BzATP activation of the apoptotic mitochondrial pathway and caspase-9.
|
) and protein assays (Fig. 8B). Using oligonucleotide primers complementary to cloned human Bcl-2 and Bax, single cDNA fragments of 300 and 455 bp, respectively, were amplified by RT-PCR from lysates of hECEs and CaSki cells (Fig. 8A). The cDNA fragments were isolated, amplified, and purified, and the products were sequenced by the dideoxy chain termination method. Sequence analysis of the cloned segments revealed homologies of 98–99% (sense and antisense) with the human Bcl-2 and Bax (the differences were sequence errors). These results indicate that cultured hECEs express mRNA for Bcl-2 and Bax.
|
). Treatment with estradiol also had no significant effect on the expression of mRNA of the ubiquitous enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in either cell type. In contrast, treatment with estradiol increased the expression of Bcl-2 mRNA in CaSki cells (Fig. 8A). To ascertain that the RT-PCR technique was sensitive in measuring changes in the expression of the Bcl-2 mRNA in CaSki cells, the experiment was repeated using different amounts of total RNA for the PCR amplification. The quantities of the amplified products of both the GAPDH and the Bcl-2 cDNA depended on the amount of total RNA used for the amplification. Treatment with estradiol had no effect on GAPDH RNA, but it increased Bcl-2 RNA (not shown). Subsequently and regardless of the amount of initial total RNA used for the PCR, in CaSki cells the ratio of Bcl-2 RNA relative to GAPDH RNA was greater in estradiol-treated cells than in estrogen-deprived cells. Densitometry analysis of three experiments in CaSki cells revealed that estradiol increased the ratio of Bcl-2 to GAPDH mRNA 5- to 7-fold. To further ascertain that the RT-PCR technique can yield interpretable semiquantitative results, the effect of number of PCR cycles on the expression of Bcl-2 and GAPDH mRNA was determined. The results (not shown) revealed that using 35 cycles resulted in synthesis reaction that did not reach plateau, indicating amplification conditions for log-phase synthesis. These data suggest that estrogen up-regulates Bcl-2 in the cancer CaSki cells but not the normal hECE cells.
This hypothesis was further supported by using Western blots to test the effects of estrogen on the expression of Bcl-2 and Bax proteins. Treatment with estradiol had no significant effect on the expression of Bcl-2 and Bax proteins in hECEs and Bax protein in CaSki cells (Fig. 8B). Estradiol also had no significant effect on the expression of the ubiquitous ß-actin in either cell type. In contrast, treatment with estradiol increased the expression of Bcl-2 protein in CaSki cells (Fig. 8B). Densitometry analysis of three experiments in CaSki cells revealed that estradiol increased the ratio of Bcl-2 to ß-actin 3- to 7-fold. Collectively, the data in Fig. 8 indicate that estrogen up-regulates Bcl-2 in CaSki cells and that estrogen up-regulation of Bcl-2 could be the mechanism by which estrogen blocks BzATP-induced apoptosis in that cell type.
Estrogen effects on Ca2+-induced mitochondrial swelling
Lack of estrogen up-regulation of Bcl-2 in hECEs prompted us to investigate involvement of other calcium-dependent mitochondrion-related mechanisms. In other types of cells, Ca2+-induced abrogation of the inner transmembrane mitochondrial potential involves formation of mitochondrial permeability transition pores (e.g. Ref.36). The next experiment studied to what degree estrogen modulates directly formation of mitochondrial permeability transition pores. Experiments used isolated mitochondria, and formation of mitochondrial permeability transition pores was determined in terms of mitochondrial swelling as a decrease in OD at 540 nm (see Materials and Methods). This method provides for the direct manipulation of mitochondrial status without influences of cell structures such as the nucleus or cytoskeleton.
Initial experiments used hECEs to determine the effective concentration of Ca2+ relative to the effect of the protonophore uncoupler carbonyl cyanide m-chlorophenylhydrazone (administered at 50 µM). In mitochondrial preparations of hECEs, the half-maximal effective concentration of Ca2+ was about 0.5 mM, and the effect could be blocked by 0.5 mM EGTA (Fig. 9A). The effect of Ca2+ could also be attenuated by coadministration of bongkrekic acid (Fig. 9A), which is a ligand of the adenine nucleotide translocator that stabilizes and prevents its degradation (37). These results suggest that in hECEs calcium induces mitochondrial swelling by activation of mitochondrial permeability transition pores.
|
, B and C). In contrast, in mitochondrial preparations obtained from CaSki cells, pretreatment with estrogen Ca2+-induced decrease in OD540 nm was smaller than in mitochondrial preparations obtained from CaSki cells grown in steroid-free medium in the absence of estrogen (–32 vs. –45%, P < 0.02, Fig. 9, B and C). These results support the hypothesis that in CaSki cells estrogen confers directly stabilizing effect on the mitochondria, and pretreatment with the hormone attenuates Ca2+-induced formation of mitochondrial permeability transition pores. The present results, however, refute a similar estrogen-dependent mechanism in the normal hECE cells.
Estrogen effects on P2X7 receptor-induced Ca2+ influx
The antiapoptotic effect of estrogen depended on Ca2+-dependent activation of the mitochondrial pathway (Figs. 2 and 7), but in hECEs estrogen did not affect directly mitochondrial function (Figs. 8 and 9). The next experiment tested the hypothesis that in the normal hECEs estrogen modulation of P2X7 receptor-dependent apoptosis involves modulation of calcium influx.
In hECEs the acute ATP-induced changes in cytosolic calcium were described (38) (Fig. 10A). They involve P2Y2 receptor-mediated transient calcium mobilization and a slower P2X4 receptor-mediated augmented calcium influx (3, 39, 40). In the continued presence of ATP or of BzATP, there was also a delayed sustained increase in cytosolic calcium that began 10–15 min after adding the agonist and persisted for the duration of the experiment (Fig. 10A and Table 1). Monitoring fura-2 fluorescence for 30 min without treatment revealed a relatively stable signal (not shown), ruling out extrusion of fura-2 as an explanation of the ATP-dependent late sustained increase in cytosolic calcium. Lowering extracellular calcium to less than 0.1 mM abolished the late calcium increase (Fig. 10B), indicating that the latter depends on calcium influx. This experimental design was used to investigate the effect of estrogen on P2X7 receptor-induced calcium influx. Changes in calcium influx were determined in terms of increases in cytosolic calcium.
|
|
, addition of CaCl2 to the bathing solutions (to increase extracellular calcium to 1.2 mM) resulted in rapid increase in cytosolic calcium, but the effect was smaller in estrogen-treated cells than estrogen-deprived cells. These data indicate that estrogen blocks P2X7 receptor-induced Ca2+ influx.
Table 1 summarizes three independent experiments in hECEs and CaSki cells. Resting levels of cytosolic calcium did not differ significantly between hECEs and CaSki cells. Shifting cells to normal calcium 25 min after treatment with BzATP resulted in greater increases in cytosolic calcium in hECEs than CaSki cells. In hECEs pretreatment with estrogen blocked significantly the increases in cytosolic calcium in response to BzATP (
Cai). In contrast, in CaSki cells the effect of estrogen was small and not significant (Table 1). These data suggest that the antiapoptotic effect of estrogen in hECEs depends mainly on blocking P2X7 receptor-induced Ca2+ influx.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The present results also suggest two mechanisms by which cancer cervical cells evade P2X7 receptor-induced apoptosis. The late sustained increases in cytosolic calcium in response to activation of the P2X7 receptor were smaller in cancer cells than in normal cells. In addition, in the cancer cervical cells, CaSki, the antiapoptotic Bcl-2, was up-regulated in response to estrogen. The former would entail lesser proapoptotic signal, whereas the latter an increased antiapoptotic mechanism, and both could act in concert to block P2X7 receptor/calcium-induced apoptosis. Whether those mechanisms contribute to the neoplastic transformation of cervical epithelial cells in vivo is at present unknown.
The main objective of the present study was to understand estrogen modulation of apoptosis. Estrogen blocked baseline and P2X7 receptor-induced apoptosis in hECEs, ECE16–1, HT3, and CaSki cells. The effect of estrogen involved transcription and protein translation, but it did not depend on cell-cycle phase or the state of cell differentiation, indicating that the antiapoptotic effect of estrogen is unrelated to its mitogenic function. The specificity profile of estrogen action supports interaction with specific response elements but lack of inhibition by tamoxifen ICI-182780, and progesterone does not support activation of the traditional nuclear estrogen receptor(s). Because the time course of apoptosis assays required 9 h of treatment with ATP or BzATP, it was difficult to pinpoint the direct time-related perturbation by estrogen. Nevertheless, the present data suggest effects as early as 1–3 h after adding the hormone. This short time course is theoretically compatible with activation of the classic nuclear receptors, as was previously reported for estrogen-regulation of endothelial nitric oxide synthase and cGMP activity in human cervical epithelial cells (19). Another possibility is activation of nonnuclear receptors (44). The latter family of nontraditional estrogen receptors involves response mechanisms that use kinases and phosphatases as their downstream signaling pathways, such as phosphatidylinositol-3-OH kinase and protein kinase B/Akt pathways (45, 46), which are involved in apoptosis regulation.
In contrast to HT3 and CaSki cells, treatment of SiHa and Hela cells with 17ß-estradiol had no effect on the P2X7 receptor-induced apoptosis, despite the fact that both types of cells express specific binding sites for the hormone (not shown). Previous studies have shown that both types of cells can undergo apoptosis in response to other stimuli (47, 48), and in those cases estrogen could modulate their effects (49, 50). It is therefore suggested that SiHa and Hela cells evolved mechanisms that resist estrogen modulation of P2X7 receptor-induced apoptosis. One of the mitochondrion-related antiapoptotic mechanisms in cancer cells is Bcl-2. Cervical cancer cells such as CaSki (present study), Siha (49), and Hela (50) express Bcl-2, and the protein plays a role in the regulation of apoptosis (e.g. Ref.51). In CaSki cells estrogen up-regulated Bcl-2 (present study), suggesting it is an antiapoptotic mechanism of estrogen. On the other hand in SiHa and Hela cells, we did not find estrogen regulation of Bcl-2 (not shown), suggesting that SiHa and Hela cells have lost the capacity for estrogen up-regulation of Bcl-2. One possible explanation is the loss of apoptosis regulation. An alternative explanation is of already submaximal Bcl-2 activity, with no additional effect by estrogen.
In both hECEs and CaSki cells, treatment with estrogen blocked P2X7 receptor-induced activation of caspase-9. However, the regulation of upstream signaling proximal to the caspase-9 differed significantly between the two types of cells. In the normal hECEs, treatment with estrogen did not affect directly mitochondrial swelling in response to Ca2+ or the expression of Bcl-2, suggesting that estrogen regulation of apoptosis involves steps proximal to the mitochondrion. Estrogen did block BzATP-induced calcium influx, suggesting that the latter is the main cellular mechanism of the antiapoptotic effect of estrogen in normal cervical cells. Previous studies in hECEs (52) and other types of cells (e.g. Refs.53, 54, 55) have shown that estrogen can lower cytosolic calcium, but the present study is the first to suggest estrogen regulation of P2X7 receptor-induced calcium-influx. The molecular mechanism by which estrogen modulates P2X7 receptor-induced Ca2+ influx is unknown.
In contrast to hECEs, in the cancer CaSki cells, estrogen had only a mild effect on BzATP-induced calcium influx, but it up-regulated expression of Bcl-2. The Bcl-2/Bcl-xL and Bax/Bak family proteins control directly mitochondrial apoptotic pathways, and up-regulation of Bcl-2/Bcl-xL activity or down-regulation of Bax/Bak activity blocks release of cytochrome C and the downstream mitochondrial-dependent activation of caspases (31). In this regard the up-regulation of Bcl-2 is most likely the mechanism that explains the attenuation of Ca2+-induced mitochondrial swelling in CaSki mitochondrial preparations.
Interestingly, infection with HPV and the expression of HPV copies by cervical cells played relatively little role in the P2X7 receptor-induced apoptosis and the responses to estrogen. Effects in ECE16–1 cells resembled those in hECEs, whereas effects in CaSki cells were similar to those in HT3 cells. It is therefore possible that in the cervix HPV plays a more prominent role in the process of carcinogenesis (56) than in apoptosis regulation.
Based on the present results, we propose that in the normal cervix, estrogen functions as an antiapoptotic agent and that the effect involves blocking P2X7 receptor-induced calcium influx. Lack of estrogen, such as after menopause, can stimulate thinning of the cervical-vaginal epithelia for two main reasons: the loss of estrogen mitogenic effect and the loss of estrogenic antiapoptotic influence. The other facet of this hypothesis is that as antiapoptotic agents estrogens could promote the growth of cervical cancer. This could potentially be the mechanism by which chronic exogenously administered estrogens induce cervical and vaginal squamous carcinogenesis in HPV-16 transgenic mice (57). Support for this speculation is the fact that only a minimal degree of apoptosis could be demonstrated in those tissues after estrogen treatment (Arbeit, J. M., personal communication). Furthermore, epidemiological studies have shown an increased risk for cervical cancer in premenopausal women on high-dose birth control pills (58, 59) [although not in postmenopausal women on hormone replacement (60)]. These data would suggest that estrogens could have a permissive effect for the development and growth of cervical cancer. However, the conclusions from those studies should be interpreted with caution because none of those studies or the present study has shown estrogens to act directly as carcinogens in the cervix. Rather, estrogen could promote the growth of cancer cells secondarily by acting as antiapoptotic agents.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: BzATP, 2',3'-0-(4-Benzoylbenzoyl)-ATP; EGF, epithelial growth factor; FCS, fetal calf serum; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; hECE, human cervical epithelial cell; HPV, human papillomavirus; SDS, sodium dodecyl sulfate.
Received June 28, 2004.
Accepted for publication August 12, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ß in the regulation of cervical permeability. Am J Physiol 278:C689–C696
