Estrogen Activates Mitogen-Activated Protein Kinase in Native, Nontransfected CHO-K1, COS-7, and RAT2 Fibroblast Cell Lines
http://www.100md.com
内分泌学杂志 2005年第1期
Departments of Anatomy & Cell Biology (C.D.T.-A.), Obstetrics & Gynecology (I.S.N., A.A.T., V.K., C.D.L., C.D.T.-A.), and Neurology (C.D.T.-A.), and the Centers for Neurobiology and Behavior and Reproductive Sciences (I.S.N., A.A.T., V.K., C.D.L., C.D.T-A.), Columbia University College of Physicians & Surgeons, New York, New York 10032
Address all correspondence and requests for reprints to: C. Dominique Toran-Allerand, Columbia University College of Physicians & Surgeons, Departments of Anatomy & Cell Biology and Neurology, 650 West 168th Street, Black Building 1615, New York, New York 10032. E-mail: cdt2@columbia.edu.
Abstract
CHO-K1, COS-7, and Rat2 fibroblast cell lines are generally believed to be devoid of estrogen receptors (ERs) and have been widely used to study the functions of ER- and ER-? after transfection of their cDNAs. Numerous studies have demonstrated that transfected ER- or ER-? mediates estradiol-induced activation of multiple signaling pathways, including the MAPK/ERK pathways. We report here for the first time that both 17-estradiol and 17?-estradiol elicit activation of MAPK/ERK in native, nontransfected CHO-K1, COS-7, and Rat2 fibroblast cell lines. We further report that, contrary to the generally held belief, these cell lines are not unresponsive to estradiol in their native, nontransfected state, and that this estrogen responsiveness is associated with estrogen binding. Using multiple ER antibodies, we failed to find ER- or ER-? isoforms or even ER-X. In view of these findings, researchers, using such cells as models to investigate mechanisms of estrogen action, must always take into account the existence of endogenous estrogen binding proteins other than ER-, ER-?, or ER-X.
Introduction
ESTRADIOL HAS BEEN shown to elicit rapid activation of multiple signaling pathways, including the Ras-Raf-MAPK/ERK signaling pathway (1, 2, 3, 4, 5, 6). The estrogen receptor (ER) that mediates MAPK/ERK activation has not been fully characterized. Many studies have concluded that both ER- and ER-? mediate estrogen activation of the MAPK/ERK signaling pathway, based on experiments involving transfection of these ER cDNAs into reportedly ER-deficient cell lines such as COS-7 (1), CHO-K1 (7), and Rat2 fibroblasts (8). We reported previously that 17- and 17?-estradiol elicit activation of the MAPK cascade in the developing neocortex of wild-type and ER- gene-disrupted (ERKO) mice (4, 9). This response appears to be mediated by an approximately 62- to 63-kDa plasma membrane-associated (caveolar-like-membrane-associated) and developmentally regulated putative ER, which is neither ER- nor ER-? and that we have designated ER-X (10). Selective activation of ER- was inhibitory for MAPK, whereas activation of ER-?, was without effect (9, 10). As part of an ongoing study to determine the phenotypic distribution of ER-X, we analyzed cell lines said to be ER- and ER-? deficient, such as CHO-K1, COS-7, and Rat2 fibroblasts, to determine whether or not they express ER-X and respond to both 17- and 17?-estradiol by activation of MAPK/ERK, a hallmark of ER-X expression. Here, we describe that we found to our surprise that not only did both 17-estradiol and 17?-estradiol elicit activation of the MAPK cascade in native (untransfected) cells but that these cells also express immunoreactive ER proteins and estrogen binding sites.
Materials and Methods
CHO-K1 (CCL-61), COS-7 (CRL-1651), and Rat2 fibroblasts (CRL-1764) were purchased from ATCC (Manassas, VA). 17-Estradiol and water-soluble 17?-estradiol were purchased from Sigma (St. Louis, MO). Water-soluble 17-estradiol was prepared by encapsulation of 17-estradiol with cyclodextrin to make it comparable with the caged 17?-estradiol available from Sigma.
Cell culture
The cells were cultured as per the supplier’s protocols. CHO-K1 cells were grown in Ham’s F12K medium with 2 mM L-glutamine adjusted to contain 1.5 g/liter sodium bicarbonate, (ATCC 30–2004) and 10% fetal bovine serum (Summit Biotechnology, Fort Collins, CO), at 37 C in a humidified 5% CO2 atmosphere. COS-7 and Rat2 cells were cultured in DMEM with 4 mM L-glutamine adjusted to contain 1.5 g/liter sodium bicarbonate and 4.5 g/liter glucose (ATCC 30–2002) and 10% fetal bovine serum, at 37 C in a humidified 5% CO2 atmosphere. Media were replaced twice a week. Antibiotics were not used.
MAPK (ERK) assay
For the MAPK assay, the cells were used at approximately 50% confluence and were then switched to media containing gelded (castrated) horse serum (4) (JRH Biosciences, Lenexa, KS), 24 h before pulsing with 17-estradiol (0.1 nM) or 17?-estradiol (10 nM) for various intervals of time from 2 min to 60 min. 17-Estradiol was used at a lower concentration (0.1 nM) than 17?-estradiol (10 nM) because 17-estradiol activated MAPK in wild-type neocortical explants at a concentration as low as 0.001 nM (1 pM), whereas 17?-estradiol required higher concentrations (10). We attribute this to the need of 17?-estradiol having to overcome first the inhibitory effect of ER- that, unlike 17-estradiol, it activates as well (10). However, in the ER--deficient neocortical explants of ERKO, (11) mice, both 17-estradiol and 17?-estradiol elicited MAPK activation at 1 pM (Singh, M., and C. D. Toran-Allerand, unpublished observation). The cyclodextrin (54.4 ng/ml; Sigma), which was used to make 17-estradiol and 17?-estradiol water soluble, was added for 2 min and used alone as a control. A 2-min exposure to fibroblast growth factor (FGF)-2 (50 ng/ml, Peprotech, Rocky Hill, NJ) was used as a positive control for ERK phosphorylation. To determine whether estradiol- or FGF-induced activation of MAPK/ERK was specific, the cells were pretreated for 10 min with U0126 (10 μM), a selective inhibitor of MAPK kinase (MEK), the kinase immediately upstream of MAPK/ERK, or with the vehicle control (0.1% dimethylsulfoxide) 10 min before exposure with 17-estradiol or 17?-estradiol or FGF2. To determine whether the ER- and ER-? selective antagonist, ICI 182,780 blocks 17-estradiol- and 17?-estradiol-induced ERK phosphorylation, the cells were pretreated for 30 min with ICI 182,780 (1 μM) or with the vehicle control, ethanol (0.01%) before treating with the two estradiols.
Cell processing
The cultures were harvested into lysis buffer [50 mM Tris (pH 7.4), 150 mM NaCl, 10% glycerol, 1 mM EGTA, 1 mM Na3VO4, 5 mM ZnCl2, 100 mM NaF, and a tablet of protease inhibitor cocktail mini per 10 ml (Roche, Indianapolis, IN), containing 1% Triton X-100, and processed, as previously described (4)]. Protein content in sample lysates was estimated, based upon the method of Lowry (DC Protein Assay Kit, Bio-Rad, Hercules, CA).
Isolation of postnuclear supernatant (PNS) and plasma membrane
Isolation of PNS and plasma membrane was carried out, according to the procedure outlined by Smart et al. (12). Briefly, the cells were homogenized in buffer A [20 mM Tricine (pH 7.8), 1 mM EDTA, 250 mM sucrose, and antiprotease cocktail (Roche)], and the homogenate was centrifuged at 1000 x g for 10 min at 4 C. The supernatant was designated as PNS. Plasma membrane was obtained by layering PNS on a 30% percoll gradient in buffer A and ultracentrifugation at 100,000 x g for 30 min. The plasma membrane was washed twice with buffer A and was then solubilized in lysis buffer, containing the nonionic detergent, octylglucoside [n-octyl-?-D-glucopyranoside] (60 mM, Roche) for Western analysis.
Western blotting
Cells were harvested into lysis buffer and subjected to SDS-PAGE, as previously described (4). After electrophoresis, polyacrylamide gels were transferred onto polyvinylidene difluoride (PVDF) membranes (Bio-Rad), blocked overnight with 5% Carnation nonfat milk in Tris-buffered saline containing 0.2% Tween 20 and probed with the following antibodies: for MAPK/ERK phosphorylation: rabbit anti-phosphoMAPK [dual phosphospecific (Thr202/Tyr204), 1:1000, New England Biolabs, Beverly, MA]; for ERK protein assessment: rabbit anti-ERK1 (1:1000); rabbit anti-ERK2 (1:1000, Santa Cruz Biotechnology, Inc., Santa Cruz, CA). To identify ER- and ER-X, the blots were probed with two polyclonal antibodies, MC20 (1:200, Santa Cruz Biotechnology, Inc.) and C1355 (1:1000, Upstate Biotechnology, Inc., Lake Placid, NY) and one monoclonal antibody, 6F11 (1:50, Novacastra, Newcastle upon Tyne, UK). For ER-?, the blots were probed with ER-? antibody (1:250, Zymed, South San Francisco, CA). To test the specificity of the MC20 or C1355 antibody binding, duplicates lanes were run on the same gel and the proteins in the gel were transferred onto the same PVDF membrane. The PVDF membrane was divided into equal halves, and the blots were exposed to the antibody in the presence or absence of 500 molar excess of the MC20 or C1355 peptides. The peptide sequence for the ER-? antibody could not be obtained from the manufacturer (Zymed). Adult mouse uterine and ovarian lysates were used as positive controls for ER- and ER-?, respectively. The PVDF membrane-bound antibody was detected, using a secondary antibody, conjugated to horseradish peroxidase (1:40,000, Pierce, Rockford, IL) and visualized on autoradiographic film, using enzyme-linked chemiluminescence (ECL; Amersham, Arlington Heights, IL). All blots probed for phosphoERK were stripped with stripping buffer (Chemicon, Temecula, CA) and reprobed with anti-ERK1 and anti-ERK2 antibodies to verify equal loading of protein across lanes. The molecular masses of the proteins on the blots were calculated, using magic markers (Invitrogen Life Technologies, Grand Island, NY). In Fig. 1, 2 and 5 μl of magic markers (Invitrogen Life Technologies) were loaded on the MC20 peptide and the MC20 antibody panels, respectively. We observed that by loading 5 μl of magic markers for the peptide blocking experiment, the marker lane had very big bands that overshadowed the neighboring lanes that resulted from the availability of excess unbound antibody.
FIG. 1. Western blots showing ER--like immunoreactivity. The plasma membranes of CHO-K1, COS-7, and Rat2 cells were isolated and transferred onto PVDF membranes and probed with MC20 antibody either alone or in the presence of MC20 peptide. Note that there are no bands corresponding to classical ER- [66/67 kDa]. However, the MC20 antibody revealed ER immunoreactive bands of approximately 32, 33, 74, and 76 kDa in the CHO-K1 plasma membrane, of 31-, 51-, 52-, 74-, and 76-kDa bands in COS-7 plasma membrane and of 32, 40, 53, 74, 76, 99, and 109 kDa in Rat2 plasma membranes. Two bands of molecular masses 74 and 76 kDa, identified by the MC20 antibody in CHO-K1 and COS-7 cells were too faint to be scanned. For specificity experiments, duplicate lanes were run on the same gel, the proteins in the gel were transferred onto the same PVDF membrane. The PVDF membranes were cut into equal halves (one half for the antibody and the other for antibody plus peptide incubation) and processed concurrently.
Densitometric analysis of ERK phosphorylation
Autoradiograms were scanned with an Epson ActionScanner II (Epson America, Torrance, CA) and analyzed by Kodak 1D Image Analysis software (Eastman Kodak, Rochester, NY). Net intensity values were calculated by subtracting the background within the area measured for each band from the total intensity within this same measured area to account for any variation in background intensity across the film.
Hormone binding studies
Measurements of estradiol binding by the cytosol were carried out, as described previously (13). To study estrogen binding in membranes, crude membranes were prepared from PNS, obtained according to Smart et al. (12), as described above. The PNS was ultracentrifuged at 100,000 x g for 15 min, and the pellet was resuspended and solubilized in PBS (pH 7.4), containing 1% Triton X-100 and 60 mM octylglucoside. Fifty micrograms of protein were added to microcentrifuge tubes containing 0.1 ml of 10% hydroxylapatite (HAP) (Bio-Rad) suspension in PBS diluted with 1 ml of cold PBS. Tubes were incubated on a rotating mixer for 30 min at room temperature to bind membrane proteins to HAP. The HAP pellet was washed twice with 1 ml cold PBS. Labeled 17?-estradiol ([2,4,6,7,16,17-3H(N)]), 110 Ci/mmol, NEN Life Science Products, PerkinElmer, Boston, MA), dissolved in PBS, was added without (total binding) or with (nonspecific binding) 1000-fold excess of cold ligand [17-estradiol or 17?-estradiol (Sigma)]. The tubes were gently vortexed and incubated with shaking at 4 C for 16–20 h. Unbound estradiol was removed by diluting the incubation mixture with 1.25 ml of cold PBS, pelleting the HAP and subsequently washing the HAP pellet with the same volume of PBS. The radioactivity associated with HAP was extracted with 1 ml of ethanol. Ethanol extract (0.8 ml) was taken for liquid scintillation counting (Tri Carb 1900 CA, Packard, Downers Grove, IL). Specific binding was determined as the difference between total binding of the radiolabeled ligand and nonspecific binding (binding of radiolabeled ligand in the presence of excess of cold ligand). Because the cell lines under study were reportedly ER negative, special efforts were taken to determine that we were dealing with specific binding. To characterize the binding sites, Hill analysis, a method of choice when cooperative binding to multiple sites (sigmoidal binding) is observed, was performed, as described elsewhere (14, 15). The calculation of dissociation constant (Kd) (Hill analysis) was based exclusively on the data for specific binding.
Results
17-Estradiol and 17?-estradiol elicit rapid phosphorylation of ERK1/2 in CHO-K1, COS-7, and Rat2 cell lines. To determine whether both 17-estradiol and 17?-estradiol elicit phosphorylation of ERK1/2 in the allegedly ER-deficient CHO-K1, COS-7, and Rat2 cell lines we analyzed the pattern and time course of ERK phosphorylation. Both 17- and 17?-estradiol elicited rapid phosphorylation of ERK1/2 within 2 min in all three cell lines (Figs. 2, A and B; 3, A and B; and 4, A and B). The pattern of ERK phosphorylation expression, after exposure to 17-estradiol, appeared to be biphasic in all three cell lines, whereas 17?-estradiol-induced ERK phosphorylation was biphasic only in COS-7 and Rat2 fibroblasts. The specificity of estradiol-induced ERK phosphorylation was confirmed by blocking ERK activation successfully with U0126 (Cell Signaling, Beverly, MA), the selective pharmacological inhibitor of MEK1/2, the kinase immediately upstream of ERK (Figs. 3, A and B; 4, A and B; and 5B). The cyclodextrin alone control did not raise baseline ERK phosphorylation (Fig. 5A). FGF2 phosphorylation of ERK was used as a positive control and was stronger than that elicited by 17-estradiol or 17?-estradiol in all the three cell lines, particularly in the Rat2 fibroblasts. (Fig. 6). In Rat2 fibroblasts, FGF2 phosphorylation of ERK2 was stronger than ERK1 (Fig. 6). The ER- and ER-? selective antagonist, ICI 182,780 failed to inhibit 17-estradiol- and 17?-estradiol-induced ERK phosphorylation, as shown in CHO-K1 cells (Fig. 5C), whereas the MEK inhibitor successfully blocked FGF2-induced ERK phosphorylation (data not shown).
FIG. 2. 17-Estradiol and 17?-estradiol elicit phosphorylation of ERK1 and ERK2 in CHO-K1 cells. The cells were pulsed with 17-estradiol (0.1 nM) or 17?-estradiol (10 nM) for various intervals of time. The cell lysates were analyzed by Western analysis for ERK phosphorylation. The upper two blots represent the time course for phosphorylation of ERK1/2 by 17-estradiol (A) and 17?-estradiol (B). The lower panel in each sub-figure represents total ERK to verify equal loading across lanes. C, The bar graph shows the densitometric analysis of the blots displayed. All the blots shown represent at least three individual experiments unless otherwise stated.
FIG. 3. 17-Estradiol and 17?-estradiol elicit phosphorylation of ERK1/2 in COS-7 cells. Western blots show the time course for phosphorylation of ERK1/2 and successful inhibition by U0126 of ERK phosphorylation with 17-estradiol (A) and 17?-estradiol (B). The lower panel in each sub-figure represents total ERK to verify equal loading. C, The bar graph shows the densitometric analysis of the blots.
FIG. 4. 17-Estradiol and 17?-estradiol also elicit phosphorylation of ERK1/2 in Rat2 cells. Western blots show the time course for activation of ERK phosphorylation and U0126-induced inhibition of ERK phosphorylation elicited by 17-estradiol (A) and 17?-estradiol (B). The lower panel in each sub-figure represents total ERK to verify equal loading. C, The bars represent the densitometric analysis of the blots.
FIG. 5. The effect of mock treatment (sham), cyclodextrin, U0126, and ICI 182,780 on estradiol-induced ERK phosphorylation in CHO-K1 cells. A, The blot shows that the cyclodextrin, which was used to prepare caged 17-estradiol and 17?-estradiol, did not raise baseline phosphorylation of ERK. B, The blot shows the successful inhibition of 17-estradiol and 17?-estradiol-induced ERK phosphorylation by the MEK1/2 selective inhibitor U0126. This is a representative blot of at least three individual experiments. C, The blot shows the inability of the ER-selective antagonist ICI 182,780 to block 17-estradiol and 17?-estradiol phosphorylation of ERK. The blots represent two independent experiments. The lower panel in each blot represents total ERK to verify equal loading. The bar graphs represent densitometric analysis of the blots.
FIG. 6. Comparison of ERK phosphorylation elicited by 17-estradiol, 17?-estradiol, and FGF2. FGF2-induced ERK phosphorylation was stronger than that induced by 17-estradiol and 17?-estradiol in CHO-K1, COS-7, and Rat2 fibroblasts. FGF2-induced phosphorylation of ERK was strongest in Rat2 fibroblasts.
CHO-K1, COS-7, and Rat2 cells lack classical ER- and ER-? but express multiple ER-- and ER-?-like immunoreactive protein bands
Neither ER- (66/67 kDa) nor the classical ER-? isoforms (55 and 60 kDa) nor ER-X (62–63 kDa) (10) appeared to be expressed in these cell lines. However, Western analysis of CHO-K1, COS-7, and Rat2 plasma membranes (Fig. 1) with antibodies directed against the last 20 amino acids of the C terminus of mouse ER- (MC20, Santa Cruz Biotechnology, Inc.), revealed immunoreactive bands of approximately 32, 33, 74, and 76 kDa in the CHO-K1 plasma membrane, of 31-, 51-, 52-, 74-, and 76-kDa bands in COS-7 plasma membrane and of 32, 40, 53, 74, 76, 99, and 109 kDa in Rat2 plasma membranes. The specificity of the immunoreactive proteins was determined by blocking the antibody binding by pretreatment with the MC20 immunizing peptide, thus confirming the specificity of these bands (Fig. 1). All the bands recognized by the MC20 antibody were deemed specific. The protein bands of molecular masses 74 and 76 kDa, identified by the MC20 antibody in CHO-K1 and COS-7 cells were too faint to be scanned. The C1355 antibody identified two specific protein bands in CHO-K1 cells only, a band of molecular mass 24.5 kDa in both the plasma membrane and cell lysate and one of 46 kDa in the cell lysate only (data not shown). The 6F11 antibody did not recognize any immunoreactive protein in all three cell lines tested (data not shown). Although Western analysis with an ER-? antibody, raised against 18 amino acids in the C-terminal region of mouse ER-? (Zymed), revealed an approximately 113-kDa band in the CHO-K1 plasma membrane, a faint band of 21 kDa in the COS-7 plasma membrane and 21-, 40-, and 88-kDa bands in Rat2 plasma membrane their specificity could not be determined for lack of blocking peptides (data not shown).
Hormone binding studies
Repeated experiments failed to show evidence of specific estradiol binding in the cytosol of CHO-K1 cells (data not shown). This was in a good agreement with the generally held view that CHO-K1 cells are devoid of ERs (7, 16). However, because the mediator of rapid, nongenomic estrogen effects is thought to be located within the plasma membrane (5, 17), we also studied 17?-estradiol binding to crude membrane preparations of CHO-K1 cells. Unlike the cytosol, specific and saturable estrogen binding was seen in the total membranes (Fig. 7). The binding isotherm was sigmoidal (biphasic), suggestive of the presence of two binding sites, and the data fitted ideally into the Hill plot (14). Hill analysis showed two classes of estrogen-binding sites with Kds of 2.5 and 9.8 nM, respectively. A similar binding isotherm was reported in studies of cytosolic binding of [3H]raloxifene (14). We also measured estrogen binding in crude membranes of COS-7 and Rat2 cells at [3H]estradiol concentrations of 10 and 30 nM. At both concentrations and in both cell lines, specific estradiol binding, comparable with that in CHO-K1 cells, was observed (Fig. 8A). Because 17-estradiol also elicited MAPK activation in each cell line, we measured whether unlabeled 17-estradiol could displace [3H]17?-estradiol from the binding sites. Displacement was observed in CHO-K1, COS-7, and Rat2 cells at both 10 and 30 nM of [3H]estradiol (Fig. 8A), indicating the presence of specific membrane-associated 17-estradiol binding sites in all three cell lines.
FIG. 7. A, Binding isotherm; B, Hill analysis of [3H]estradiol binding to membranes of CHO-K1 cells. In the binding isotherm plot, results are presented as mean ± SEM. Each point represents five to eight independent measurements. In the Hill plot, horizontal (X) axis: log [free ligand], vertical (Y) axis: log [sites bound]/[sites free], crossover point at Y = 0 gives log Kd.
FIG. 8. Specific binding sites for 17?-estradiol and 17-estradiol in membranes of CHO-K1, COS-7, and Rat2 cells and 17?-estradiol dose-dependent ERK phosphorylation in CHO-K1 cells. A, Open and solid columns represent 10 nM 17?-[3H]estradiol and 30 nM 17?-[3H]estradiol concentrations, respectively. Specific binding was determined as the difference between total binding of labeled 17?-estradiol and nonspecific binding (labeled 17?-estradiol plus 1000-fold excess of either cold 17?-estradiol or 17-estradiol, correspondingly). Results are represented as mean ± SEM. B, The blots show ERK phosphorylation in response to 1–20 nM 17?-estradiol. The lower panel represents total ERK to verify equal loading. The bar graph represents the densitometric analysis of the blots.
Dose response analysis demonstrated that 1 nM 17?-estradiol was as efficient as 10 nM in eliciting ERK phosphorylation. A higher concentration of 17?-estradiol (20 nM) did not change the response substantially (Fig. 8B).
Discussion
Our results indicate clearly that both 17-estradiol and 17?-estradiol activate MAPK in native, nontransfected CHO-K1, COS-7, and Rat2 fibroblast cell lines and that transfection of the cDNA of either ER- or ER-? or of the 46-kDa truncated ER- variant (18) into these cells is not required for estradiol activation of MAPK. Estradiol-induced activation of MAPK in these cell lines takes place despite the absence of full-length ER- or classical ER-? isoforms. Thus, transfection with ER- or ER-? cDNA for the purpose of studying their role in nongenomic effects of estrogen in these cell lines has both serious conceptual and methodological limitations.
Many workers have concluded that ER- or ER-? mediates estradiol activation of MAPK as a result of transfecting ER- or ER-? cDNAs into allegedly ER-deficient cell lines, such as COS-7, CHO-K1, and Rat2 (1, 7, 8). Migliaccio et al. (1) determined that ER- is essential for estradiol-induced activation of ERK2 by transfecting ER- cDNA into COS-7 cells. Other studies have also reported that transiently transfected ER- or ER-? localizes to the CHO-K1 and COS-1 plasma membrane and mediates estradiol phosphorylation and activation of ERK (7, 19). In contrast, we demonstrate for the first time that transfection of ER- or ER-? cDNAs into these cells lines is not necessary for 17-estradiol or 17?-estradiol to elicit activation of MAPK, endogenous binding proteins suffice.
Previous studies apparently failed to detect estradiol-induced ERK phosphorylation in untransfected CHO-K1, COS-7, or Rat2 cell lines. This may be explained by the fact that ERK activation was either not analyzed in the untransfected cells (1) or that only one time point was measured (7, 8). Moreover, whereas Razandi et al. (7) failed to detect specific 17?-estradiol binding activity in untransfected CHO-K1 cells, Migliaccio et al. (1) and Wade et al. (8) apparently did not investigate estrogen binding at all; and Ince et al. (16) analyzed estrogen binding in whole cell extracts and not membranes.
Despite the absence of ER- and the classical ER-? isoforms, 17-estradiol and 17?-estradiol-induced activation of MAPK/ERK in CHO-K1, COS-7, and Rat2 cells is in agreement with our previous observations. Data obtained from experiments on wild-type and ERKO mice, using ER--selective [16-iodo-17?-estradiol (9) and propylpyrazole triol (10)] or ER-?-selective [genestein (9)] ligands, and the natural enantiomer 17-estradiol, demonstrated that neither ER- nor ER-? mediates MAPK/ERK activation (4, 10) in the developing neocortex. We proposed that a novel, plasma membrane-associated and developmentally regulated putative ER, ER-X, which has high affinity for both 17-estradiol and 17?-estradiol, mediates estradiol-induced activation of the MAPK cascade (10). The failure to detect ER-X in CHO-K1, COS-7, and Rat2 cells suggests 1) that ER-X may be restricted to the developing rodent (10) and primate brain (Nethrapalli, I. S., and C. D. Toran-Allerand, unpublished observations) and uterus (10), or 2) that transformed cell lines may not express ER-X, or 3) that ER-X levels in these cells may be too low for detection. However, the fact that both 17-estradiol and 17?-estradiol did elicit a response in all three cell lines is most unusual. Once we have cloned ER-X and have sequenced the protein, which are currently in progress, we will be in a better position to pursue this question more fully. The existence of estradiol-induced activation of the MAPK cascade in CHO-K1, COS-7, and Rat2 cell lines may be similar to that reported in HCC38 cells (20), which are also purported to be ER negative. In these cells, estradiol activation of protein kinase C was not blocked by the ER antagonist ICI 182,780 or by antibodies to ER- and ER-?. Similarly, as reported here, ICI 182,780 failed to inhibit ERK phosphorylation as shown in CHO-K1 cells (Fig. 5C). Various studies (10, 21, 22) have shown that rapid effects of estrogen that are ICI 182,780 insensitive may still be mediated by ERs, including ER-X. ICI 182, 80 insensitivity may merely reflect the absence of ER- and the classical isoforms of ER-? in these cell lines. Inhibition of estrogen-induced ERK phosphorylation by ICI 182,780, observed by other workers in COS-7 (1), CHO-K1 (7), and Rat2 (8) cells, may be misleading, because it is likely to result from the normal blocking of estrogen action on the transfected ER- and ER-? by the ICI compound. Another possibility, the G protein-coupled receptor homolog, GPR30, is also unlikely to be involved. Unlike what we report here, GPR30 is not immunoreactive for ER- or ER-?, has been shown to mediate only 17?-estradiol but not 17-estradiol activation of ERK in ER- and ER-? negative SKBR3 breast cancer cells (23) and is activated by ICI 182,780 and tamoxifen. Moreover, no immunoreactive ER band corresponding to the appropriate GPR30 38-kDa size was identified in any of the cell lines tested. 17-Estradiol- and 17?-estradiol-induced phosphorylation of ERK were comparable in intensity to that elicited by FGF2 in CHO-K1 and COS-7 cells. However, the hyperphosphorylation of ERK by FGF2 observed in Rat2 fibroblasts may well result from the fact that fibroblasts are more responsive to FGF2 (Fig. 6).
The results of the binding studies support the data on estrogen-induced activation of ERK obtained in all three cell lines. We used 10 nM of 17?-estradiol in this study because it is the most commonly used concentration for studies of rapid estrogen effects. The finding of two estrogen-binding sites in the crude membranes of CHO-K1 cells prompted us to conduct the dose response experiments. We found no correlation between the patterns for dose-dependent ERK phosphorylation and for the saturation curve (binding isotherm) in the binding assays. If both estrogen-binding proteins were involved in the ERK activation then a biphasic dose response pattern (corresponding to a biphasic binding isotherm) would be expected. We found no such a correlation. On the contrary, 1 nM of 17?-estradiol was able to activate ERK phosphorylation as efficiently as 10 nM. This suggests that one high affinity receptor may be mediating the response. Thus, the higher-affinity estrogen binding site (Kd = 2.5 nM) found in CHO-K1 cells seems to mediate the effects of 17?-estradiol on ERK in these cells, although the role of the lower affinity binding site (Kd = 9.8) also may be important. Similar estrogen binding sites may be responsible for mediating estrogen signaling in COS-7 and Rat2 cells as well. However, although we found multiple ER immunoreactive proteins in all these cell lines, the estrogen binding sites cannot be associated with any particular proteins at this time.
Neocortical plasma membranes of postnatal d 7 ERKO mice also exhibited high-affinity estrogen binding, which was associated with ER-X (10). Although we did not find the approximately 62- to 63-kDa ER-X band in the cell lines under study, its characteristic affinity for 17-estradiol (10), a critical feature distinguishing ER-X from ER- or ER-?, was evident in all three cell lines studied.
Further research will throw more light on the mechanisms underlying the rapid, nongenomic actions of estradiol in light of the absence of full-length ER-, classical ER-? isoforms or ER-X. However, our results indicate clearly that, despite the absence of transfection of ERs into purported ER-deficient cell lines, such as CHO-K1, COS-7, and Rat2, both 17-estradiol and 17?-estradiol can elicit ERK activation and undoubtedly other signaling pathways as well. The presence of endogenous estrogen binding sites other than ER-, ER-?, or ER-X must always be taken into consideration, when using these cell lines as models for understanding the mechanisms of estrogen action.
References
Migliaccio A, Di Domenico M, Castoria G, de Falco A, Bontempo P, Nola E, Auricchio F 1996 Tyrosine kinase/p21ras/MAP-kinase pathway activation by estradiol-receptor complex in MCF-7 cells. EMBO J 15:1292–1300
Endoh H, Sasaki H, Maruyama K, Takeyama K, Waga I, Shimizu T, Kato S, Kawashima H 1997 Rapid activation of MAP kinase by estrogen in the bone cell line. Biochem Biophys Res Commun 235:99–102
Watters JJ, Campbell JS, Cunningham MJ, Krebs EG, Dorsa DM 1997 Rapid membrane effects of steroids in neuroblastoma cells: effects of estrogen on mitogen activated protein kinase signalling cascade and c-fos immediate early gene transcription. Endocrinology 138:4030–4033
Singh M, Setalo Jr G, Guan X, Warren M, Toran-Allerand CD 1999 Estrogen-induced activation of mitogen-activated protein kinase in cerebral cortical explants: convergence of estrogen and neurotrophin signaling pathways. J Neurosci 19:1179–1188
Kuroki Y, Fukushima K, Kanda Y, Mizuno K, Watanabe Y 2000 Putative membrane-bound estrogen receptors possibly stimulate mitogen-activated protein kinase in the rat hippocampus. Eur J Pharmacol 400:205–209
Nethrapalli IS, Singh M, Guan X, Guo Q, Lubahn DB, Korach KS, Toran-Allerand CD 2001 Estradiol (E2) elicits src phosphorylation in the mouse neocortex: the initial event in E2 activation of the MAPK cascade? Endocrinology 142:5145–5148
Razandi M, Pedram A, Greene GL, Levin ER 1999 Cell membrane and nuclear estrogen receptors (ERs) originate from a single transcript: studies of ER and ER? expressed in Chinese hamster ovary cells. Mol Endocrinol 13:307–319
Wade CB, Robinson S, Shapiro RA, Dorsa DM 2001 Estrogen receptor (ER) and ER? exhibit unique pharmacologic properties when coupled to activation of the mitogen-activated protein kinase pathway. Endocrinology 142:2336–2342
Singh M, Setalo Jr G, Guan X, Frail DE, Toran-Allerand CD 2000 Estrogen-induced activation of the mitogen-activated protein kinase cascade in the cerebral cortex of estrogen receptor- knock-out mice. J Neurosci 20:1694–1700
Toran-Allerand CD, Guan X, MacLusky NJ, Horvath TL, Diano S, Singh M, Connolly Jr ES, Nethrapalli IS, Tinnikov AA 2002 ER-X: a novel, plasma membrane-associated, putative estrogen receptor that is regulated during development and after ischemic brain injury. J Neurosci 22:8391–8401
Lubahn DB, Moyer JS, Golding TS, Couse JF, Korach KS, Smithies O 1993 Alteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor gene. Proc Natl Acad Sci USA 90:11162–11166
Smart EJ, Ying YS, Mineo C, Anderson RG 1995 A detergent-free method for purifying caveolae membrane from tissue culture cells. Proc Natl Acad Sci USA 92:10104–10108
Tinnikov AA, Bazhan NM, Yakovleva TV 1992 The study of properties of immunoreactive estradiol secreted by adrenals and ovaries of immature female rats. Steroids 57:174–177
Fiorelli G, Martineti V, Gori F, Benvenuti S, Freddiani U, Formigli L, Zecchi S, Brandi ML 1997 Heterogeneity of binding sites and bioeffects of raloxifene on the human leukemic cell line FLG 29.1. Biochem Biophys Res Commun 240:573–579
Dahlquist FW 1978 The meaning of Scatchard and Hill plots. Methods Enzymol 48:270–299
Ince BA, Zhuang Y, Wrenn CK, Shapiro DJ, Katzenellenbogen BS 1993 Powerful dominant negative mutants of the human estrogen receptor. J Biol Chem 268:14026–14032
Toran-Allerand CD 2000 Novel sites and mechanisms of oestrogen action in the brain. In: McEwen BS, ed. Neuronal and cognitive effects of oestrogens. Chichester, UK: Wiley (Novartis Foundation Symposium 230), pp 56–73
Li L, Haynes MP, Bender JR 2003 Plasma membrane localization and function of the estrogen receptor variant (ER46) in human endothelial cells. Proc Natl Acad Sci USA 100:4807–4812
Zhang Z, Maier B, Santen RJ, Song RX 2002 Membrane association of estrogen receptor mediates estrogen effect on MAPK activation. Biochem Biophys Res Commun 294:926–933
Boyan BD, Sylvia VL, Frambach T, Lohmann CH, Dietl J, Dean DD, Schwartz Z 2003 Estrogen-dependent rapid activation of protein kinase C in estrogen receptor-positive MCF-7 breast cancer cells and estrogen receptor-negative HCC38 cells is membrane-mediated and inhibited by tamoxifen. Endocrinology 144:1812–1824
Gu Q, Korach KS, Moss RL 1999 Rapid action of 17?-estradiol on kainate-induced currents in hippocampal neurons lacking intracellular estrogen receptors. Endocrinology 140:660–666
Qiu J, Bosch MA, Tobias SC, Grandy DK, Scanlan TS, Ronnekleiv OK, Kelly MJ 2003 Rapid signaling of estrogen in hypothalamic neurons involves a novel G-protein-coupled estrogen receptor that activates protein kinase C. J Neurosci 23:9529–9540
Filardo EJ, Quinn JA, Bland KI, Frackelton Jr AR 2000 Estrogen-induced activation of Erk-1 and Erk-2 requires the G protein-coupled receptor homolog, GPR30, and occurs via trans-activation of the epidermal growth factor receptor through release of HB-EGF. Mol Endocrinol 14:1649–1660(Imam S. Nethrapalli1, Ale)
Address all correspondence and requests for reprints to: C. Dominique Toran-Allerand, Columbia University College of Physicians & Surgeons, Departments of Anatomy & Cell Biology and Neurology, 650 West 168th Street, Black Building 1615, New York, New York 10032. E-mail: cdt2@columbia.edu.
Abstract
CHO-K1, COS-7, and Rat2 fibroblast cell lines are generally believed to be devoid of estrogen receptors (ERs) and have been widely used to study the functions of ER- and ER-? after transfection of their cDNAs. Numerous studies have demonstrated that transfected ER- or ER-? mediates estradiol-induced activation of multiple signaling pathways, including the MAPK/ERK pathways. We report here for the first time that both 17-estradiol and 17?-estradiol elicit activation of MAPK/ERK in native, nontransfected CHO-K1, COS-7, and Rat2 fibroblast cell lines. We further report that, contrary to the generally held belief, these cell lines are not unresponsive to estradiol in their native, nontransfected state, and that this estrogen responsiveness is associated with estrogen binding. Using multiple ER antibodies, we failed to find ER- or ER-? isoforms or even ER-X. In view of these findings, researchers, using such cells as models to investigate mechanisms of estrogen action, must always take into account the existence of endogenous estrogen binding proteins other than ER-, ER-?, or ER-X.
Introduction
ESTRADIOL HAS BEEN shown to elicit rapid activation of multiple signaling pathways, including the Ras-Raf-MAPK/ERK signaling pathway (1, 2, 3, 4, 5, 6). The estrogen receptor (ER) that mediates MAPK/ERK activation has not been fully characterized. Many studies have concluded that both ER- and ER-? mediate estrogen activation of the MAPK/ERK signaling pathway, based on experiments involving transfection of these ER cDNAs into reportedly ER-deficient cell lines such as COS-7 (1), CHO-K1 (7), and Rat2 fibroblasts (8). We reported previously that 17- and 17?-estradiol elicit activation of the MAPK cascade in the developing neocortex of wild-type and ER- gene-disrupted (ERKO) mice (4, 9). This response appears to be mediated by an approximately 62- to 63-kDa plasma membrane-associated (caveolar-like-membrane-associated) and developmentally regulated putative ER, which is neither ER- nor ER-? and that we have designated ER-X (10). Selective activation of ER- was inhibitory for MAPK, whereas activation of ER-?, was without effect (9, 10). As part of an ongoing study to determine the phenotypic distribution of ER-X, we analyzed cell lines said to be ER- and ER-? deficient, such as CHO-K1, COS-7, and Rat2 fibroblasts, to determine whether or not they express ER-X and respond to both 17- and 17?-estradiol by activation of MAPK/ERK, a hallmark of ER-X expression. Here, we describe that we found to our surprise that not only did both 17-estradiol and 17?-estradiol elicit activation of the MAPK cascade in native (untransfected) cells but that these cells also express immunoreactive ER proteins and estrogen binding sites.
Materials and Methods
CHO-K1 (CCL-61), COS-7 (CRL-1651), and Rat2 fibroblasts (CRL-1764) were purchased from ATCC (Manassas, VA). 17-Estradiol and water-soluble 17?-estradiol were purchased from Sigma (St. Louis, MO). Water-soluble 17-estradiol was prepared by encapsulation of 17-estradiol with cyclodextrin to make it comparable with the caged 17?-estradiol available from Sigma.
Cell culture
The cells were cultured as per the supplier’s protocols. CHO-K1 cells were grown in Ham’s F12K medium with 2 mM L-glutamine adjusted to contain 1.5 g/liter sodium bicarbonate, (ATCC 30–2004) and 10% fetal bovine serum (Summit Biotechnology, Fort Collins, CO), at 37 C in a humidified 5% CO2 atmosphere. COS-7 and Rat2 cells were cultured in DMEM with 4 mM L-glutamine adjusted to contain 1.5 g/liter sodium bicarbonate and 4.5 g/liter glucose (ATCC 30–2002) and 10% fetal bovine serum, at 37 C in a humidified 5% CO2 atmosphere. Media were replaced twice a week. Antibiotics were not used.
MAPK (ERK) assay
For the MAPK assay, the cells were used at approximately 50% confluence and were then switched to media containing gelded (castrated) horse serum (4) (JRH Biosciences, Lenexa, KS), 24 h before pulsing with 17-estradiol (0.1 nM) or 17?-estradiol (10 nM) for various intervals of time from 2 min to 60 min. 17-Estradiol was used at a lower concentration (0.1 nM) than 17?-estradiol (10 nM) because 17-estradiol activated MAPK in wild-type neocortical explants at a concentration as low as 0.001 nM (1 pM), whereas 17?-estradiol required higher concentrations (10). We attribute this to the need of 17?-estradiol having to overcome first the inhibitory effect of ER- that, unlike 17-estradiol, it activates as well (10). However, in the ER--deficient neocortical explants of ERKO, (11) mice, both 17-estradiol and 17?-estradiol elicited MAPK activation at 1 pM (Singh, M., and C. D. Toran-Allerand, unpublished observation). The cyclodextrin (54.4 ng/ml; Sigma), which was used to make 17-estradiol and 17?-estradiol water soluble, was added for 2 min and used alone as a control. A 2-min exposure to fibroblast growth factor (FGF)-2 (50 ng/ml, Peprotech, Rocky Hill, NJ) was used as a positive control for ERK phosphorylation. To determine whether estradiol- or FGF-induced activation of MAPK/ERK was specific, the cells were pretreated for 10 min with U0126 (10 μM), a selective inhibitor of MAPK kinase (MEK), the kinase immediately upstream of MAPK/ERK, or with the vehicle control (0.1% dimethylsulfoxide) 10 min before exposure with 17-estradiol or 17?-estradiol or FGF2. To determine whether the ER- and ER-? selective antagonist, ICI 182,780 blocks 17-estradiol- and 17?-estradiol-induced ERK phosphorylation, the cells were pretreated for 30 min with ICI 182,780 (1 μM) or with the vehicle control, ethanol (0.01%) before treating with the two estradiols.
Cell processing
The cultures were harvested into lysis buffer [50 mM Tris (pH 7.4), 150 mM NaCl, 10% glycerol, 1 mM EGTA, 1 mM Na3VO4, 5 mM ZnCl2, 100 mM NaF, and a tablet of protease inhibitor cocktail mini per 10 ml (Roche, Indianapolis, IN), containing 1% Triton X-100, and processed, as previously described (4)]. Protein content in sample lysates was estimated, based upon the method of Lowry (DC Protein Assay Kit, Bio-Rad, Hercules, CA).
Isolation of postnuclear supernatant (PNS) and plasma membrane
Isolation of PNS and plasma membrane was carried out, according to the procedure outlined by Smart et al. (12). Briefly, the cells were homogenized in buffer A [20 mM Tricine (pH 7.8), 1 mM EDTA, 250 mM sucrose, and antiprotease cocktail (Roche)], and the homogenate was centrifuged at 1000 x g for 10 min at 4 C. The supernatant was designated as PNS. Plasma membrane was obtained by layering PNS on a 30% percoll gradient in buffer A and ultracentrifugation at 100,000 x g for 30 min. The plasma membrane was washed twice with buffer A and was then solubilized in lysis buffer, containing the nonionic detergent, octylglucoside [n-octyl-?-D-glucopyranoside] (60 mM, Roche) for Western analysis.
Western blotting
Cells were harvested into lysis buffer and subjected to SDS-PAGE, as previously described (4). After electrophoresis, polyacrylamide gels were transferred onto polyvinylidene difluoride (PVDF) membranes (Bio-Rad), blocked overnight with 5% Carnation nonfat milk in Tris-buffered saline containing 0.2% Tween 20 and probed with the following antibodies: for MAPK/ERK phosphorylation: rabbit anti-phosphoMAPK [dual phosphospecific (Thr202/Tyr204), 1:1000, New England Biolabs, Beverly, MA]; for ERK protein assessment: rabbit anti-ERK1 (1:1000); rabbit anti-ERK2 (1:1000, Santa Cruz Biotechnology, Inc., Santa Cruz, CA). To identify ER- and ER-X, the blots were probed with two polyclonal antibodies, MC20 (1:200, Santa Cruz Biotechnology, Inc.) and C1355 (1:1000, Upstate Biotechnology, Inc., Lake Placid, NY) and one monoclonal antibody, 6F11 (1:50, Novacastra, Newcastle upon Tyne, UK). For ER-?, the blots were probed with ER-? antibody (1:250, Zymed, South San Francisco, CA). To test the specificity of the MC20 or C1355 antibody binding, duplicates lanes were run on the same gel and the proteins in the gel were transferred onto the same PVDF membrane. The PVDF membrane was divided into equal halves, and the blots were exposed to the antibody in the presence or absence of 500 molar excess of the MC20 or C1355 peptides. The peptide sequence for the ER-? antibody could not be obtained from the manufacturer (Zymed). Adult mouse uterine and ovarian lysates were used as positive controls for ER- and ER-?, respectively. The PVDF membrane-bound antibody was detected, using a secondary antibody, conjugated to horseradish peroxidase (1:40,000, Pierce, Rockford, IL) and visualized on autoradiographic film, using enzyme-linked chemiluminescence (ECL; Amersham, Arlington Heights, IL). All blots probed for phosphoERK were stripped with stripping buffer (Chemicon, Temecula, CA) and reprobed with anti-ERK1 and anti-ERK2 antibodies to verify equal loading of protein across lanes. The molecular masses of the proteins on the blots were calculated, using magic markers (Invitrogen Life Technologies, Grand Island, NY). In Fig. 1, 2 and 5 μl of magic markers (Invitrogen Life Technologies) were loaded on the MC20 peptide and the MC20 antibody panels, respectively. We observed that by loading 5 μl of magic markers for the peptide blocking experiment, the marker lane had very big bands that overshadowed the neighboring lanes that resulted from the availability of excess unbound antibody.
FIG. 1. Western blots showing ER--like immunoreactivity. The plasma membranes of CHO-K1, COS-7, and Rat2 cells were isolated and transferred onto PVDF membranes and probed with MC20 antibody either alone or in the presence of MC20 peptide. Note that there are no bands corresponding to classical ER- [66/67 kDa]. However, the MC20 antibody revealed ER immunoreactive bands of approximately 32, 33, 74, and 76 kDa in the CHO-K1 plasma membrane, of 31-, 51-, 52-, 74-, and 76-kDa bands in COS-7 plasma membrane and of 32, 40, 53, 74, 76, 99, and 109 kDa in Rat2 plasma membranes. Two bands of molecular masses 74 and 76 kDa, identified by the MC20 antibody in CHO-K1 and COS-7 cells were too faint to be scanned. For specificity experiments, duplicate lanes were run on the same gel, the proteins in the gel were transferred onto the same PVDF membrane. The PVDF membranes were cut into equal halves (one half for the antibody and the other for antibody plus peptide incubation) and processed concurrently.
Densitometric analysis of ERK phosphorylation
Autoradiograms were scanned with an Epson ActionScanner II (Epson America, Torrance, CA) and analyzed by Kodak 1D Image Analysis software (Eastman Kodak, Rochester, NY). Net intensity values were calculated by subtracting the background within the area measured for each band from the total intensity within this same measured area to account for any variation in background intensity across the film.
Hormone binding studies
Measurements of estradiol binding by the cytosol were carried out, as described previously (13). To study estrogen binding in membranes, crude membranes were prepared from PNS, obtained according to Smart et al. (12), as described above. The PNS was ultracentrifuged at 100,000 x g for 15 min, and the pellet was resuspended and solubilized in PBS (pH 7.4), containing 1% Triton X-100 and 60 mM octylglucoside. Fifty micrograms of protein were added to microcentrifuge tubes containing 0.1 ml of 10% hydroxylapatite (HAP) (Bio-Rad) suspension in PBS diluted with 1 ml of cold PBS. Tubes were incubated on a rotating mixer for 30 min at room temperature to bind membrane proteins to HAP. The HAP pellet was washed twice with 1 ml cold PBS. Labeled 17?-estradiol ([2,4,6,7,16,17-3H(N)]), 110 Ci/mmol, NEN Life Science Products, PerkinElmer, Boston, MA), dissolved in PBS, was added without (total binding) or with (nonspecific binding) 1000-fold excess of cold ligand [17-estradiol or 17?-estradiol (Sigma)]. The tubes were gently vortexed and incubated with shaking at 4 C for 16–20 h. Unbound estradiol was removed by diluting the incubation mixture with 1.25 ml of cold PBS, pelleting the HAP and subsequently washing the HAP pellet with the same volume of PBS. The radioactivity associated with HAP was extracted with 1 ml of ethanol. Ethanol extract (0.8 ml) was taken for liquid scintillation counting (Tri Carb 1900 CA, Packard, Downers Grove, IL). Specific binding was determined as the difference between total binding of the radiolabeled ligand and nonspecific binding (binding of radiolabeled ligand in the presence of excess of cold ligand). Because the cell lines under study were reportedly ER negative, special efforts were taken to determine that we were dealing with specific binding. To characterize the binding sites, Hill analysis, a method of choice when cooperative binding to multiple sites (sigmoidal binding) is observed, was performed, as described elsewhere (14, 15). The calculation of dissociation constant (Kd) (Hill analysis) was based exclusively on the data for specific binding.
Results
17-Estradiol and 17?-estradiol elicit rapid phosphorylation of ERK1/2 in CHO-K1, COS-7, and Rat2 cell lines. To determine whether both 17-estradiol and 17?-estradiol elicit phosphorylation of ERK1/2 in the allegedly ER-deficient CHO-K1, COS-7, and Rat2 cell lines we analyzed the pattern and time course of ERK phosphorylation. Both 17- and 17?-estradiol elicited rapid phosphorylation of ERK1/2 within 2 min in all three cell lines (Figs. 2, A and B; 3, A and B; and 4, A and B). The pattern of ERK phosphorylation expression, after exposure to 17-estradiol, appeared to be biphasic in all three cell lines, whereas 17?-estradiol-induced ERK phosphorylation was biphasic only in COS-7 and Rat2 fibroblasts. The specificity of estradiol-induced ERK phosphorylation was confirmed by blocking ERK activation successfully with U0126 (Cell Signaling, Beverly, MA), the selective pharmacological inhibitor of MEK1/2, the kinase immediately upstream of ERK (Figs. 3, A and B; 4, A and B; and 5B). The cyclodextrin alone control did not raise baseline ERK phosphorylation (Fig. 5A). FGF2 phosphorylation of ERK was used as a positive control and was stronger than that elicited by 17-estradiol or 17?-estradiol in all the three cell lines, particularly in the Rat2 fibroblasts. (Fig. 6). In Rat2 fibroblasts, FGF2 phosphorylation of ERK2 was stronger than ERK1 (Fig. 6). The ER- and ER-? selective antagonist, ICI 182,780 failed to inhibit 17-estradiol- and 17?-estradiol-induced ERK phosphorylation, as shown in CHO-K1 cells (Fig. 5C), whereas the MEK inhibitor successfully blocked FGF2-induced ERK phosphorylation (data not shown).
FIG. 2. 17-Estradiol and 17?-estradiol elicit phosphorylation of ERK1 and ERK2 in CHO-K1 cells. The cells were pulsed with 17-estradiol (0.1 nM) or 17?-estradiol (10 nM) for various intervals of time. The cell lysates were analyzed by Western analysis for ERK phosphorylation. The upper two blots represent the time course for phosphorylation of ERK1/2 by 17-estradiol (A) and 17?-estradiol (B). The lower panel in each sub-figure represents total ERK to verify equal loading across lanes. C, The bar graph shows the densitometric analysis of the blots displayed. All the blots shown represent at least three individual experiments unless otherwise stated.
FIG. 3. 17-Estradiol and 17?-estradiol elicit phosphorylation of ERK1/2 in COS-7 cells. Western blots show the time course for phosphorylation of ERK1/2 and successful inhibition by U0126 of ERK phosphorylation with 17-estradiol (A) and 17?-estradiol (B). The lower panel in each sub-figure represents total ERK to verify equal loading. C, The bar graph shows the densitometric analysis of the blots.
FIG. 4. 17-Estradiol and 17?-estradiol also elicit phosphorylation of ERK1/2 in Rat2 cells. Western blots show the time course for activation of ERK phosphorylation and U0126-induced inhibition of ERK phosphorylation elicited by 17-estradiol (A) and 17?-estradiol (B). The lower panel in each sub-figure represents total ERK to verify equal loading. C, The bars represent the densitometric analysis of the blots.
FIG. 5. The effect of mock treatment (sham), cyclodextrin, U0126, and ICI 182,780 on estradiol-induced ERK phosphorylation in CHO-K1 cells. A, The blot shows that the cyclodextrin, which was used to prepare caged 17-estradiol and 17?-estradiol, did not raise baseline phosphorylation of ERK. B, The blot shows the successful inhibition of 17-estradiol and 17?-estradiol-induced ERK phosphorylation by the MEK1/2 selective inhibitor U0126. This is a representative blot of at least three individual experiments. C, The blot shows the inability of the ER-selective antagonist ICI 182,780 to block 17-estradiol and 17?-estradiol phosphorylation of ERK. The blots represent two independent experiments. The lower panel in each blot represents total ERK to verify equal loading. The bar graphs represent densitometric analysis of the blots.
FIG. 6. Comparison of ERK phosphorylation elicited by 17-estradiol, 17?-estradiol, and FGF2. FGF2-induced ERK phosphorylation was stronger than that induced by 17-estradiol and 17?-estradiol in CHO-K1, COS-7, and Rat2 fibroblasts. FGF2-induced phosphorylation of ERK was strongest in Rat2 fibroblasts.
CHO-K1, COS-7, and Rat2 cells lack classical ER- and ER-? but express multiple ER-- and ER-?-like immunoreactive protein bands
Neither ER- (66/67 kDa) nor the classical ER-? isoforms (55 and 60 kDa) nor ER-X (62–63 kDa) (10) appeared to be expressed in these cell lines. However, Western analysis of CHO-K1, COS-7, and Rat2 plasma membranes (Fig. 1) with antibodies directed against the last 20 amino acids of the C terminus of mouse ER- (MC20, Santa Cruz Biotechnology, Inc.), revealed immunoreactive bands of approximately 32, 33, 74, and 76 kDa in the CHO-K1 plasma membrane, of 31-, 51-, 52-, 74-, and 76-kDa bands in COS-7 plasma membrane and of 32, 40, 53, 74, 76, 99, and 109 kDa in Rat2 plasma membranes. The specificity of the immunoreactive proteins was determined by blocking the antibody binding by pretreatment with the MC20 immunizing peptide, thus confirming the specificity of these bands (Fig. 1). All the bands recognized by the MC20 antibody were deemed specific. The protein bands of molecular masses 74 and 76 kDa, identified by the MC20 antibody in CHO-K1 and COS-7 cells were too faint to be scanned. The C1355 antibody identified two specific protein bands in CHO-K1 cells only, a band of molecular mass 24.5 kDa in both the plasma membrane and cell lysate and one of 46 kDa in the cell lysate only (data not shown). The 6F11 antibody did not recognize any immunoreactive protein in all three cell lines tested (data not shown). Although Western analysis with an ER-? antibody, raised against 18 amino acids in the C-terminal region of mouse ER-? (Zymed), revealed an approximately 113-kDa band in the CHO-K1 plasma membrane, a faint band of 21 kDa in the COS-7 plasma membrane and 21-, 40-, and 88-kDa bands in Rat2 plasma membrane their specificity could not be determined for lack of blocking peptides (data not shown).
Hormone binding studies
Repeated experiments failed to show evidence of specific estradiol binding in the cytosol of CHO-K1 cells (data not shown). This was in a good agreement with the generally held view that CHO-K1 cells are devoid of ERs (7, 16). However, because the mediator of rapid, nongenomic estrogen effects is thought to be located within the plasma membrane (5, 17), we also studied 17?-estradiol binding to crude membrane preparations of CHO-K1 cells. Unlike the cytosol, specific and saturable estrogen binding was seen in the total membranes (Fig. 7). The binding isotherm was sigmoidal (biphasic), suggestive of the presence of two binding sites, and the data fitted ideally into the Hill plot (14). Hill analysis showed two classes of estrogen-binding sites with Kds of 2.5 and 9.8 nM, respectively. A similar binding isotherm was reported in studies of cytosolic binding of [3H]raloxifene (14). We also measured estrogen binding in crude membranes of COS-7 and Rat2 cells at [3H]estradiol concentrations of 10 and 30 nM. At both concentrations and in both cell lines, specific estradiol binding, comparable with that in CHO-K1 cells, was observed (Fig. 8A). Because 17-estradiol also elicited MAPK activation in each cell line, we measured whether unlabeled 17-estradiol could displace [3H]17?-estradiol from the binding sites. Displacement was observed in CHO-K1, COS-7, and Rat2 cells at both 10 and 30 nM of [3H]estradiol (Fig. 8A), indicating the presence of specific membrane-associated 17-estradiol binding sites in all three cell lines.
FIG. 7. A, Binding isotherm; B, Hill analysis of [3H]estradiol binding to membranes of CHO-K1 cells. In the binding isotherm plot, results are presented as mean ± SEM. Each point represents five to eight independent measurements. In the Hill plot, horizontal (X) axis: log [free ligand], vertical (Y) axis: log [sites bound]/[sites free], crossover point at Y = 0 gives log Kd.
FIG. 8. Specific binding sites for 17?-estradiol and 17-estradiol in membranes of CHO-K1, COS-7, and Rat2 cells and 17?-estradiol dose-dependent ERK phosphorylation in CHO-K1 cells. A, Open and solid columns represent 10 nM 17?-[3H]estradiol and 30 nM 17?-[3H]estradiol concentrations, respectively. Specific binding was determined as the difference between total binding of labeled 17?-estradiol and nonspecific binding (labeled 17?-estradiol plus 1000-fold excess of either cold 17?-estradiol or 17-estradiol, correspondingly). Results are represented as mean ± SEM. B, The blots show ERK phosphorylation in response to 1–20 nM 17?-estradiol. The lower panel represents total ERK to verify equal loading. The bar graph represents the densitometric analysis of the blots.
Dose response analysis demonstrated that 1 nM 17?-estradiol was as efficient as 10 nM in eliciting ERK phosphorylation. A higher concentration of 17?-estradiol (20 nM) did not change the response substantially (Fig. 8B).
Discussion
Our results indicate clearly that both 17-estradiol and 17?-estradiol activate MAPK in native, nontransfected CHO-K1, COS-7, and Rat2 fibroblast cell lines and that transfection of the cDNA of either ER- or ER-? or of the 46-kDa truncated ER- variant (18) into these cells is not required for estradiol activation of MAPK. Estradiol-induced activation of MAPK in these cell lines takes place despite the absence of full-length ER- or classical ER-? isoforms. Thus, transfection with ER- or ER-? cDNA for the purpose of studying their role in nongenomic effects of estrogen in these cell lines has both serious conceptual and methodological limitations.
Many workers have concluded that ER- or ER-? mediates estradiol activation of MAPK as a result of transfecting ER- or ER-? cDNAs into allegedly ER-deficient cell lines, such as COS-7, CHO-K1, and Rat2 (1, 7, 8). Migliaccio et al. (1) determined that ER- is essential for estradiol-induced activation of ERK2 by transfecting ER- cDNA into COS-7 cells. Other studies have also reported that transiently transfected ER- or ER-? localizes to the CHO-K1 and COS-1 plasma membrane and mediates estradiol phosphorylation and activation of ERK (7, 19). In contrast, we demonstrate for the first time that transfection of ER- or ER-? cDNAs into these cells lines is not necessary for 17-estradiol or 17?-estradiol to elicit activation of MAPK, endogenous binding proteins suffice.
Previous studies apparently failed to detect estradiol-induced ERK phosphorylation in untransfected CHO-K1, COS-7, or Rat2 cell lines. This may be explained by the fact that ERK activation was either not analyzed in the untransfected cells (1) or that only one time point was measured (7, 8). Moreover, whereas Razandi et al. (7) failed to detect specific 17?-estradiol binding activity in untransfected CHO-K1 cells, Migliaccio et al. (1) and Wade et al. (8) apparently did not investigate estrogen binding at all; and Ince et al. (16) analyzed estrogen binding in whole cell extracts and not membranes.
Despite the absence of ER- and the classical ER-? isoforms, 17-estradiol and 17?-estradiol-induced activation of MAPK/ERK in CHO-K1, COS-7, and Rat2 cells is in agreement with our previous observations. Data obtained from experiments on wild-type and ERKO mice, using ER--selective [16-iodo-17?-estradiol (9) and propylpyrazole triol (10)] or ER-?-selective [genestein (9)] ligands, and the natural enantiomer 17-estradiol, demonstrated that neither ER- nor ER-? mediates MAPK/ERK activation (4, 10) in the developing neocortex. We proposed that a novel, plasma membrane-associated and developmentally regulated putative ER, ER-X, which has high affinity for both 17-estradiol and 17?-estradiol, mediates estradiol-induced activation of the MAPK cascade (10). The failure to detect ER-X in CHO-K1, COS-7, and Rat2 cells suggests 1) that ER-X may be restricted to the developing rodent (10) and primate brain (Nethrapalli, I. S., and C. D. Toran-Allerand, unpublished observations) and uterus (10), or 2) that transformed cell lines may not express ER-X, or 3) that ER-X levels in these cells may be too low for detection. However, the fact that both 17-estradiol and 17?-estradiol did elicit a response in all three cell lines is most unusual. Once we have cloned ER-X and have sequenced the protein, which are currently in progress, we will be in a better position to pursue this question more fully. The existence of estradiol-induced activation of the MAPK cascade in CHO-K1, COS-7, and Rat2 cell lines may be similar to that reported in HCC38 cells (20), which are also purported to be ER negative. In these cells, estradiol activation of protein kinase C was not blocked by the ER antagonist ICI 182,780 or by antibodies to ER- and ER-?. Similarly, as reported here, ICI 182,780 failed to inhibit ERK phosphorylation as shown in CHO-K1 cells (Fig. 5C). Various studies (10, 21, 22) have shown that rapid effects of estrogen that are ICI 182,780 insensitive may still be mediated by ERs, including ER-X. ICI 182, 80 insensitivity may merely reflect the absence of ER- and the classical isoforms of ER-? in these cell lines. Inhibition of estrogen-induced ERK phosphorylation by ICI 182,780, observed by other workers in COS-7 (1), CHO-K1 (7), and Rat2 (8) cells, may be misleading, because it is likely to result from the normal blocking of estrogen action on the transfected ER- and ER-? by the ICI compound. Another possibility, the G protein-coupled receptor homolog, GPR30, is also unlikely to be involved. Unlike what we report here, GPR30 is not immunoreactive for ER- or ER-?, has been shown to mediate only 17?-estradiol but not 17-estradiol activation of ERK in ER- and ER-? negative SKBR3 breast cancer cells (23) and is activated by ICI 182,780 and tamoxifen. Moreover, no immunoreactive ER band corresponding to the appropriate GPR30 38-kDa size was identified in any of the cell lines tested. 17-Estradiol- and 17?-estradiol-induced phosphorylation of ERK were comparable in intensity to that elicited by FGF2 in CHO-K1 and COS-7 cells. However, the hyperphosphorylation of ERK by FGF2 observed in Rat2 fibroblasts may well result from the fact that fibroblasts are more responsive to FGF2 (Fig. 6).
The results of the binding studies support the data on estrogen-induced activation of ERK obtained in all three cell lines. We used 10 nM of 17?-estradiol in this study because it is the most commonly used concentration for studies of rapid estrogen effects. The finding of two estrogen-binding sites in the crude membranes of CHO-K1 cells prompted us to conduct the dose response experiments. We found no correlation between the patterns for dose-dependent ERK phosphorylation and for the saturation curve (binding isotherm) in the binding assays. If both estrogen-binding proteins were involved in the ERK activation then a biphasic dose response pattern (corresponding to a biphasic binding isotherm) would be expected. We found no such a correlation. On the contrary, 1 nM of 17?-estradiol was able to activate ERK phosphorylation as efficiently as 10 nM. This suggests that one high affinity receptor may be mediating the response. Thus, the higher-affinity estrogen binding site (Kd = 2.5 nM) found in CHO-K1 cells seems to mediate the effects of 17?-estradiol on ERK in these cells, although the role of the lower affinity binding site (Kd = 9.8) also may be important. Similar estrogen binding sites may be responsible for mediating estrogen signaling in COS-7 and Rat2 cells as well. However, although we found multiple ER immunoreactive proteins in all these cell lines, the estrogen binding sites cannot be associated with any particular proteins at this time.
Neocortical plasma membranes of postnatal d 7 ERKO mice also exhibited high-affinity estrogen binding, which was associated with ER-X (10). Although we did not find the approximately 62- to 63-kDa ER-X band in the cell lines under study, its characteristic affinity for 17-estradiol (10), a critical feature distinguishing ER-X from ER- or ER-?, was evident in all three cell lines studied.
Further research will throw more light on the mechanisms underlying the rapid, nongenomic actions of estradiol in light of the absence of full-length ER-, classical ER-? isoforms or ER-X. However, our results indicate clearly that, despite the absence of transfection of ERs into purported ER-deficient cell lines, such as CHO-K1, COS-7, and Rat2, both 17-estradiol and 17?-estradiol can elicit ERK activation and undoubtedly other signaling pathways as well. The presence of endogenous estrogen binding sites other than ER-, ER-?, or ER-X must always be taken into consideration, when using these cell lines as models for understanding the mechanisms of estrogen action.
References
Migliaccio A, Di Domenico M, Castoria G, de Falco A, Bontempo P, Nola E, Auricchio F 1996 Tyrosine kinase/p21ras/MAP-kinase pathway activation by estradiol-receptor complex in MCF-7 cells. EMBO J 15:1292–1300
Endoh H, Sasaki H, Maruyama K, Takeyama K, Waga I, Shimizu T, Kato S, Kawashima H 1997 Rapid activation of MAP kinase by estrogen in the bone cell line. Biochem Biophys Res Commun 235:99–102
Watters JJ, Campbell JS, Cunningham MJ, Krebs EG, Dorsa DM 1997 Rapid membrane effects of steroids in neuroblastoma cells: effects of estrogen on mitogen activated protein kinase signalling cascade and c-fos immediate early gene transcription. Endocrinology 138:4030–4033
Singh M, Setalo Jr G, Guan X, Warren M, Toran-Allerand CD 1999 Estrogen-induced activation of mitogen-activated protein kinase in cerebral cortical explants: convergence of estrogen and neurotrophin signaling pathways. J Neurosci 19:1179–1188
Kuroki Y, Fukushima K, Kanda Y, Mizuno K, Watanabe Y 2000 Putative membrane-bound estrogen receptors possibly stimulate mitogen-activated protein kinase in the rat hippocampus. Eur J Pharmacol 400:205–209
Nethrapalli IS, Singh M, Guan X, Guo Q, Lubahn DB, Korach KS, Toran-Allerand CD 2001 Estradiol (E2) elicits src phosphorylation in the mouse neocortex: the initial event in E2 activation of the MAPK cascade? Endocrinology 142:5145–5148
Razandi M, Pedram A, Greene GL, Levin ER 1999 Cell membrane and nuclear estrogen receptors (ERs) originate from a single transcript: studies of ER and ER? expressed in Chinese hamster ovary cells. Mol Endocrinol 13:307–319
Wade CB, Robinson S, Shapiro RA, Dorsa DM 2001 Estrogen receptor (ER) and ER? exhibit unique pharmacologic properties when coupled to activation of the mitogen-activated protein kinase pathway. Endocrinology 142:2336–2342
Singh M, Setalo Jr G, Guan X, Frail DE, Toran-Allerand CD 2000 Estrogen-induced activation of the mitogen-activated protein kinase cascade in the cerebral cortex of estrogen receptor- knock-out mice. J Neurosci 20:1694–1700
Toran-Allerand CD, Guan X, MacLusky NJ, Horvath TL, Diano S, Singh M, Connolly Jr ES, Nethrapalli IS, Tinnikov AA 2002 ER-X: a novel, plasma membrane-associated, putative estrogen receptor that is regulated during development and after ischemic brain injury. J Neurosci 22:8391–8401
Lubahn DB, Moyer JS, Golding TS, Couse JF, Korach KS, Smithies O 1993 Alteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor gene. Proc Natl Acad Sci USA 90:11162–11166
Smart EJ, Ying YS, Mineo C, Anderson RG 1995 A detergent-free method for purifying caveolae membrane from tissue culture cells. Proc Natl Acad Sci USA 92:10104–10108
Tinnikov AA, Bazhan NM, Yakovleva TV 1992 The study of properties of immunoreactive estradiol secreted by adrenals and ovaries of immature female rats. Steroids 57:174–177
Fiorelli G, Martineti V, Gori F, Benvenuti S, Freddiani U, Formigli L, Zecchi S, Brandi ML 1997 Heterogeneity of binding sites and bioeffects of raloxifene on the human leukemic cell line FLG 29.1. Biochem Biophys Res Commun 240:573–579
Dahlquist FW 1978 The meaning of Scatchard and Hill plots. Methods Enzymol 48:270–299
Ince BA, Zhuang Y, Wrenn CK, Shapiro DJ, Katzenellenbogen BS 1993 Powerful dominant negative mutants of the human estrogen receptor. J Biol Chem 268:14026–14032
Toran-Allerand CD 2000 Novel sites and mechanisms of oestrogen action in the brain. In: McEwen BS, ed. Neuronal and cognitive effects of oestrogens. Chichester, UK: Wiley (Novartis Foundation Symposium 230), pp 56–73
Li L, Haynes MP, Bender JR 2003 Plasma membrane localization and function of the estrogen receptor variant (ER46) in human endothelial cells. Proc Natl Acad Sci USA 100:4807–4812
Zhang Z, Maier B, Santen RJ, Song RX 2002 Membrane association of estrogen receptor mediates estrogen effect on MAPK activation. Biochem Biophys Res Commun 294:926–933
Boyan BD, Sylvia VL, Frambach T, Lohmann CH, Dietl J, Dean DD, Schwartz Z 2003 Estrogen-dependent rapid activation of protein kinase C in estrogen receptor-positive MCF-7 breast cancer cells and estrogen receptor-negative HCC38 cells is membrane-mediated and inhibited by tamoxifen. Endocrinology 144:1812–1824
Gu Q, Korach KS, Moss RL 1999 Rapid action of 17?-estradiol on kainate-induced currents in hippocampal neurons lacking intracellular estrogen receptors. Endocrinology 140:660–666
Qiu J, Bosch MA, Tobias SC, Grandy DK, Scanlan TS, Ronnekleiv OK, Kelly MJ 2003 Rapid signaling of estrogen in hypothalamic neurons involves a novel G-protein-coupled estrogen receptor that activates protein kinase C. J Neurosci 23:9529–9540
Filardo EJ, Quinn JA, Bland KI, Frackelton Jr AR 2000 Estrogen-induced activation of Erk-1 and Erk-2 requires the G protein-coupled receptor homolog, GPR30, and occurs via trans-activation of the epidermal growth factor receptor through release of HB-EGF. Mol Endocrinol 14:1649–1660(Imam S. Nethrapalli1, Ale)