0013-7227/03/$15.00/0
Printed in U.S.A.
Endocrinology 144(1):121–128
Copyright © 2003 by The Endocrine Society
doi: 10.1210/en.2002-220625
Nongenomic Activity and Subsequent c-fos Induction by
Estrogen Receptor Ligands Are Not Sufficient to
Promote Deoxyribonucleic Acid Synthesis in Human
Endometrial Adenocarcinoma Cells
DAVID W. SINGLETON, YUXIN FENG, CRAIG J. BURD,
AND
SOHAIB A. KHAN
Department of Cell Biology, Neurobiology and Anatomy, University of Cincinnati College of Medicine, Cincinnati, Ohio
45219
Estrogen 17␤-estradiol (E2) rapidly modulates several signaling pathways related to cell growth, preservation, and differentiation. The physiological role of these nongenomic effects
with regard to downstream outcomes, and the relationship
with transcriptional estrogen activity are unclear. Furthermore, the ability of selective estrogen receptor modulators
(SERMs) to trigger nongenomic actions is largely unknown.
To determine whether estrogen receptor (ER) ligands exert
nongenomic activity in endometrial adenocarcinoma cells,
and whether this activity affects transcription and DNA synthesis, we challenged human Ishikawa cells with E2 or partial
ER agonists 4-hydroxytamoxifen (OHT) and raloxifene (ral).
Serum-starved Ishikawa cells exposed for 5 min to 0.1 nM E2
showed induced phosphorylation of MAPK (ERK1/2). Ral and
4-OHT each at 1 nM also stimulated ERK in a rapid transient
manner. E2 and 4-OHT induced proto-oncogene c-fos mRNA
expression in Ishikawa cells within 30 min, but ral had no
effect. In contrast to nongenomic action, only E2 stimulated
expression of an estrogen response element (ERE)-driven luciferase (LUC) reporter gene. To examine DNA synthesis, [3H]thymidine incorporation was measured in serum-starved cultures exposed to E2 or partial agonists for 2 d. E2 at 1 nM
stimulated thymidine uptake in an ERK-dependent manner,
but 1 nM 4-OHT, 1 nM ral, and 0.1-nM concentrations of E2 had
no significant effects. Taken together, these data indicate that
both nongenomic and direct transcriptional ER effects are
likely required to promote DNA synthesis. (Endocrinology 144:
121–128, 2003)
E
STROGEN 17␤-ESTRADIOL (E2) elicits a number of
rapid actions that do not directly involve transcriptional regulation by estrogen receptor (ER). These nongenotropic effects, which are thought to be ER dependent, include
stimulation of prolactin and nitric oxide production in pituitary and endothelial cells, respectively (1– 4), chemoprotection and preservation of neuronal cells (5– 8), and proliferative or antiapoptotic actions in breast cancer cells (9 –11).
Estrogen-dependent growth of ER-positive tumors could
have a significant nongenomic component, but the role of
these rapid actions in terms of estrogen’s mitogenic potential
has not been defined. Many nongenomic effects of E2 may be
mediated by p44/p42 MAPK, also known as ERK1/2. Several studies suggest involvement of plasma membrane associated ER␣ (for reviews see Refs. 12 and 13). The putative
membrane ER was shown in various systems to interact with
components of growth factor receptor pathways when liganded by E2. These include functional associations with
IGF-1 receptor (14), the adaptor protein Shc (15), tyrosine
kinase c-src (9), the p85 subunit of phosphatidylinositol3-OH kinase (PI3 kinase) (16), and caveolin-1 (17). Recently,
Kousteni et al. (18) illustrated that the ligand-binding (E)
domain of ER was necessary and sufficient for E2-mediated
antiapoptotic effects in transfected HeLa cells. These obser-
vations implicate several different pathways for nonnuclear
activity by ER, which likely affect cell growth.
Although tissue specific estrogen-like effects of selective
ER modulators (SERMs) are known, their ability to exert
nongenomic actions is not well characterized. In the present
study, we used ER-positive human endometrial Ishikawa
cells to compare classic transcriptional activity with potential
nongenomic actions for E2, 4-hydroxytamoxifen (OHT), and
raloxifene (ral). Specifically, we wished to determine if rapid
responses were present independently, or in combination
with estrogen response element (ERE) regulation that could
affect downstream events such as promotion of DNA synthesis. In the case of ral, which is not known to stimulate
uterine cell growth, nongenomic activity may lead to other,
currently unknown cellular outcomes. ERE-mediated transcriptional activity was assessed in Ishikawa cells exposed to
E2 or SERMs and compared with nongenomic effects on
ERK1/2 and PI3 kinase. These findings were followed up by
analyses of immediate early gene expression and DNA replication. Taken together, the data indicate that ERK activation
and c-fos induction were not sufficient but were required for
promotion of DNA synthesis by E2, suggesting that rapid
actions function in concert with nuclear ER-mediated
transcription.
Abbreviations: CMV, Cytomegalovirus; E2, 17␤-estradiol; ER, estrogen receptor; ERE, estrogen response element; GAPDH, glyceraldehydephosphate dehydrogenase; 4-OHT, 4-hydroxytamoxifen; LUC, luciferase; PI3 kinase, phosphatidylinositol-3-OH kinase; ral, raloxifene;
SERMs, selective ER modulators; SRE, serum response element.
Materials and Methods
Materials
Human endometrial Ishikawa cells were a generous gift of Dr. Masato
Nishida (Kasumigaura National Hospital, Ibaraki, Japan). E2 and 4-OHT
121
122
Endocrinology, January 2003, 144(1):121–128
Singleton et al. • ER Ligand Effects in Ishikawa Cells
were purchased from Sigma (St. Louis, MO). Ral was a gift from Dr. T.
Wilson, GlaxoSmithKline (Research Triangle Park, NC). ICI 182,780 was
obtained from Tocris Cookson Inc. (Bullwin, MO). Phospho-ERK1/2,
phospho-Akt kinase and total ERK1/2 polyclonal antibodies were from
Cell Signaling Technology (Beverly, MA). MEK1/2 inhibitor U0126 was
purchased from Promega Corp. (Madison, WI). Anti-ER␣ (H-184) and
ER␤ polyclonal antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) and Panvera Corp. (Madison, WI), respectively. Full-length recombinant ER␤ protein was also purchased
from Panvera. Adenoviral vector expressing RasN17, and control virus
were provided by Dr. Jeff Molkentin (Children’s Hospital Research
Foundation, Cincinnati, OH). pCMV5-ER␣ expressing human ER␣, and
p3xERE-LUC containing three copies of the vitellogenin A consensus
ERE, were gifts from Dr. Benita Katzenellenbogen (University of Illinois,
Urbana, IL) and Dr. Craig V. Jordan (Northwestern University, Chicago,
IL), respectively.
Dried membranes were hybridized in ExpressHyb solution (CLONTECH Laboratories, Inc., Palo Alto, CA) with 32P-labeled probes made
by random primer labeling of c-fos or glyceraldehyde-phosphate dehydrogenase (GAPDH) cDNAs. Signals were detected by exposure of
hybridized membranes to x-ray film. Quantitation of signal intensities
was performed with Metamorph software. For RT-PCR detection of ER
transcripts, Superscript reagents from Invitrogen (Carlsbad, CA) were used.
Oligonucleotide primer sequences were: 5⬘-ggccacgcgtcgactagtacAGACATGAGAGCTGCCAACC-3⬘ (ER␣ sense), 5⬘-ggccacgcgtcgactagtacGCCAGGCACATTGTAGAAGG-3⬘ (ER␣ antisense), 5⬘-ggccacgcgtcgactagtacTCATCACATCTGTATGCGGAACC-3⬘ (ER␤ sense), 5⬘-ggccacgcgtcgactagtacCGTAACACTTCCGAAGTCGG-3⬘ (ER␤ antisense). Uppercase letters
correspond to specific ER sequences, and lowercase letters denote a universal
amplification primer sequence (19).
Cell culture and transfections
To examine promotion of DNA synthesis in Ishikawa cells following
treatment with ER ligands, 24-well plates were seeded at 5 ⫻ 104 cells
per well in phenol red-free medium and maintained overnight at 37 C.
Cultures were then rinsed with PBS and the medium changed to MEM
containing 0.5% serum. The following day, test agents were diluted in
low-serum medium and added to the cultures for 48 h. During the final
6 h of incubation, [3H]-thymidine (1 ␮Ci/ml) was added to each well.
Cultures were rinsed with cold PBS, followed by addition of 5% TCA
and incubation for 20 min at 4 C. Wells were then rinsed in PBS before
extraction with 400 ␮l of 0.5 n NaOH, 0.5% sodium dodecyl sulfate, and
subsequent quantitation by scintillation counting.
Ishikawa cells were maintained at 37 C, constant 5% CO2 in MEM
(Life Technologies, Inc., Rockville, MD) supplemented with 10% fetal
bovine serum and 1⫻ penicillin-streptomycin. Cultures at 70 – 80% confluence were placed in serum-free MEM for 48 h before treatment with
various agents and harvesting of total cellular RNA or proteins. Cells
were washed several times with PBS (pH 7.4) during the 48-h incubation,
and the media changed to minimize basal ERK1/2 phosphorylation
levels. For experiments using dominant negative RasN17, cells at 70%
confluence were washed in PBS and incubated at 37 C for 3 h in DMEM
(Life Technologies, Inc.) containing 2% FBS and adenovirus. Infected
cultures were then serum deprived for 48 h before harvesting. For
transfection of Ishikawa cells, 0.5 ␮g of pCMV5-ER␣ or pCDNA3-S118A
(ER␣ mutant), 0.5 ␮g of pCMV-␤ (CLONTECH Laboratories, Inc., Palo
Alto, CA) and 1.0 ␮g p3xERE-LUC were transiently introduced using
lipofectamine (Life Technologies, Inc.) according to manufacturer’s instructions. LUC and ␤-galactosidase activities were determined with
reagents from Promega Corp.
Immunoblot analysis
Ishikawa cultures, either untreated or exposed to ER ligands, were
lysed in boiling buffer consisting of 66 mm Tris-HCl (pH 7.5), 2% sodium
dodecyl sulfate, 2 mm sodium orthovanadate, and 2 mm sodium fluoride. Lysates were incubated at 95 C for 4 min and cleared by centrifugation. Protein concentrations were determined using bicinchoninic
acid assay reagents (Pierce Chemical Co., Rockford, IL). Fifteen micrograms of protein from each sample were resolved with 12% SDS-PAGE
followed by transfer to nitrocellulose membranes. The membranes were
blocked by incubation at room temperature for 1 h with 5% nonfat dry
milk in TBS-T (20 mm Tris, pH 7.6; 137 mm NaCl; 0.1% Tween-20). Dually
phosphorylated ERK1 and ERK2 or phospho-Akt were detected by
incubating membranes overnight at 4 C with phospho-specific antibodies in TBS-T/5% BSA. After washing with TBS-T, membranes were
incubated for 1 h at room temperature with horseradish peroxidase
coupled secondary antibody in TBS-T/5% dry milk followed by visualization using enhanced chemiluminescence (Amersham Pharmacia
Biotech, Arlington Heights, IL). For detection of total ERK1/2 levels,
membranes were stripped, reblocked, and incubated overnight at 4 C
with anti-ERK1/2 antibody in TBS-T/5% BSA followed by incubation
with secondary antibody and visualization as described above. ER␣ and
␤ expression were assessed by incubating membranes at room temperature for 1 h with specific primary antibodies diluted in TBS-T/5%
nonfat dry milk, followed by secondary antibody and enhanced chemiluminescence detection as above.
Northern analysis and RT-PCR
Serum-deprived Ishikawa cultures were briefly exposed to E2 or
partial agonists before harvesting total cellular RNA with the RNeasy
method (QIAGEN Inc., Valencia, CA). For experiments involving
MEK1/2 inhibition, cells were preincubated with U0126 for 10 min
followed by replacement with fresh media containing test compounds
and U0126. RNAs (20 ␮g) were separated on 1% agarose, 5% formaldehyde gels, and transferred onto nylon membranes by capillary action.
[3H]-Thymidine incorporation assay
Results
ER expression in Ishikawa cells
Human endometrial-derived Ishikawa cells are known to
express ER, but relative ␣ and ␤ isoform levels in various
clonal lines have not been examined (20). To determine ER
isoform expression in the Ishikawa line 3-H-12, total RNA or
protein was extracted, followed by RT-PCR and immunoblot
analyses. Figure 1A illustrates the results of RT-PCR using
specific ER␣ (lane 2), or ER␤ (lane 3), primers. A band corresponding to the expected size (350 bp) for the ER␣ product
was noted. A faint signal corresponding to the 400-bp ER␤
product was also seen (lane 3). Immunoblot analysis revealed
expression of both ER␣ and ␤ isoforms (Fig. 1B) of approximately 68 and 59 kDa, respectively. As might be expected,
brief exposure to E2 did not alter ER␣ levels (Fig. 1B, lane 2).
Activation of ERK1/2 by estradiol, ral, and 4-OHT
Antiestrogen 4-OHT is thought to possess uterotrophic
actions through partial ER agonism. To examine if antiestrogens 4-OHT and ral can trigger nongenomic activity,
serum-deprived Ishikawa cells were briefly incubated at 37
C with E2 or SERMs at physiologically relevant concentrations (0.1 or 1 nm, respectively). Marked phosphorylation of
ERK2 and to a lesser extent ERK1 was seen after 5 min
exposure to 0.1 nm estradiol (Fig. 2A). Activated ERK was at
control levels following 15 min of E2 exposure. Active (phosphorylated) Akt kinase was unaffected by E2 treatment. In
Fig. 2B, cultures were treated with 1 nm 4-OHT or ral. Both
agents stimulated ERK with a similar potency and time
course. Phosphorylation of ERK1/2 was induced to a maximum following 10 min exposure to each SERM. As with
estradiol, no effect on Akt phosphorylation was noted (data
not shown).
The ER requirement for nongenomic activity may depend
on the particular response being examined, as well as the
Singleton et al. • ER Ligand Effects in Ishikawa Cells
Endocrinology, January 2003, 144(1):121–128 123
FIG. 1. ER expression in Ishikawa cells. A,
RT-PCR indicates ER␣ mRNA expression
(lane 2) and a faint signal for ER␤ (lane 3).
Lane 1 is a reaction control using primers
for GAPDH with RNA from the same sample used for lanes 2 and 3. B, Western blot
analysis was performed with Ishikawa
whole-cell extract using ER␣- or ␤-specific
antibodies as indicated. Fifteen micrograms of total protein were loaded from
cultures incubated in the absence of E2
(lanes 1 and 3), or following a 5-min exposure to 0.1 nM E2 (lane 2). Lane 4 contained
25 ng of full-length recombinant ER␤ protein. Polyacrylamide gels (12% and 7.5%)
were used for ER␣ and ER␤ blots, respectively.
FIG. 2. ER-dependent activation of ERK1/2 by estrogen, 4-OHT, and ral. Ishikawa cultures were serum deprived for 48 h followed by exposure
to the indicated agents. Fifteen micrograms of protein from each sample were resolved by 12% SDS-PAGE, followed by transfer to nitrocellulose
and immunoblot analysis. Membranes were probed with antibodies recognizing active dually phosphorylated ERK (P-ERK), stripped, and
reprobed for total ERK1/2 or phospho-Akt. Similar results were obtained from three or more independent experiments. A, Vehicle control (0.001%
dimethylsulfoxide, harvested at 10 min), or 0.1 nM E2 for 1–30 min. B, Vehicle, 4-OHT (1 nM), or ral (1 nM) for 5, 10, and 15 min. C, Cells were
challenged with E2 (0.1 nM) and ICI 182,780 (10 or 100 nM) as indicated for 5 min. D, Ishikawa cultures infected with control virus or adenovirus
carrying RasN17 were serum starved and exposed to E2 (0.1 nM) for 5 min. Samples for control virus and Adeno-RasN17 were analyzed together
on one blot, but irrelevant intervening lanes were omitted for sake of clarity.
cellular model. To determine if ER is necessary for activation
of ERK1/2, Ishikawa cells were cotreated with E2 and the
pure ER antagonist ICI 182,780. Figure 2C indicates attenuation of E2 mediated ERK activation by ICI at 10 and 100 nm.
The antagonist alone appeared to have a slight positive effect
on ERK2 phosphorylation. These observations implicate ER
involvement, with the modest positive effect by ICI alone
possibly reflecting differences in the mechanism for ERK
activation compared with classic ER-mediated transcriptional regulation. It is conceivable therefore, that ICI 182,780
124
Endocrinology, January 2003, 144(1):121–128
may not be a pure antagonist in Ishikawa cells with respect
to nongenomic activity. Because rapid activation of MAPK
typically requires the GTP-binding protein Ras, we examined
involvement of this membrane-associated component for E2mediated ERK1/2 activation. Ishikawa cells were infected
with adenoviral vector expressing the dominant negative
RasN17 mutant and 24 h later were briefly exposed to E2.
Figure 2D shows ERK activation from cultures infected with
control virus or adeno-RasN17. E2-mediated ERK1/2 phosphorylation was substantially reduced in cells infected with
dominant negative Ras, suggesting that Ras is a required
component.
Regulation of p3xERE-LUC reporter construct
As discussed earlier, the relationship between nongenomic and genomic estrogen mechanisms is not clear in
terms of overall cellular responses. Thus, we wanted to compare the nongenomic actions observed in Ishikawa cells with
classic ER-mediated transcriptional regulation. To this end,
cultures were transfected with p3xERE-LUC followed by
exposure to ER ligands for 16 h (Fig. 3). Surprisingly, less
than 3-fold stimulation of LUC activity by estradiol was seen
in the absence of cotransfected ER. An 8-fold induction by 10
nm E2 was found when ER␣ expression plasmid was included. E2 at 0.1 nm exhibited nearly the same potency.
Neither 4-OHT nor ral were able to induce LUC expression
regardless of ER␣ overexpression. The concentration of
SERMS (1 nm) was clearly sufficient for ERK phosphorylation, thus illustrating differential sensitivities and/or independence of nongenomic and transcriptional ER actions.
Estradiol and 4-OHT induce immediate early gene c-fos
MAPK (ERK) activation is associated with subsequent
phosphorylation of cytosolic and nuclear substrates, including transcription factors controlling proto-oncogene expression. To determine whether ERK activation leads to increased expression of c-fos mRNA in Ishikawa cells, cultures
were challenged with E2 or SERMs and examined by Northern analysis. Figure 4A shows induction of c-fos expression
FIG. 3. ERE-LUC expression in response to estrogen
and SERMs. Ishikawa cultures were transiently cotransfected with p3xERE-LUC and pCMV␤gal, with
and without wild-type ER␣ expression vector, followed by 16 h exposure to vehicle, E2 (10 and 0.1 nM),
ral (1 nM), or OHT (1 nM). LUC activity was determined as relative light units and normalized to ␤galactosidase activity. These data represent the results of three experiments, using duplicate samples.
Controls within each experiment were averaged and
assigned a value of 1. Error bars represent SEM.
Singleton et al. • ER Ligand Effects in Ishikawa Cells
by E2 (0.1 nm), following 30 min treatment. At 1 h, c-fos
mRNA was essentially returned to control levels. Phorbol
myristate 13-acetate (100 ng/ml) strongly induced c-fos at 30
min, and the signal was further increased at 1 h. Ral exhibited
no effect, whereas 4-OHT showed substantial induction,
which was roughly equal to that of E2 (Fig. 4, B–D). These
observations are in agreement with previous data from this
laboratory, in which tamoxifen and E2 were found to induce
c-fos mRNA in rat uterus (21, 22). To determine involvement
of ERK1/2, Ishikawa cells were briefly pretreated with the
MEK inhibitor U0126 before addition of E2, or 4-OHT. Inclusion of U0126 inhibited c-fos induction as well as basal
expression (Fig. 4C), indicating a requirement for activated
ERK. Densitometric analysis revealed 4-fold and 4.3-fold average increases in c-fos expression by E2 and 4-OHT, respectively (Fig. 4D).
MAPK (ERK) activation is not required for direct ER␣mediated transcription
During the course of this study, we observed that a 2-h
pulse treatment with E2 followed by overnight incubation in
normal growth medium was sufficient to induce LUC
in transfected Ishikawa cells. To examine if ERK plays a role
in nuclear ER activity, we exposed transfected Ishikawa cultures to E2 for 2 h in the absence and presence of U0126. This
was followed by removal of test agents, and overnight incubation before assessment of LUC induction. MEK inhibitor
U0126 had no effect on induction of LUC with 0.1 nm E2 (Fig.
5). To further assess the role of ERK, cells were transfected
with ER␣ mutated at serine residue 118 (S118A), a known
MAPK phosphorylation site in the activation function-1 region of ER, which may contribute to receptor function (23–
25). Cells expressing this ER mutant exhibited the same level
of LUC expression as those with wild-type receptor, suggesting that ERK1/2 played no role. However, it should be
emphasized that ERK activity likely has differential effects
on ER-regulated genes, depending on promoter context and
presence or absence of other elements such as activator
protein-1 sites.
Singleton et al. • ER Ligand Effects in Ishikawa Cells
Endocrinology, January 2003, 144(1):121–128 125
FIG. 4. Induction of c-fos mRNA by E2 and 4-OHT. Twenty micrograms per lane of total cellular RNA were resolved on 1% agarose-formaldehyde
gels and subjected to Northern analysis, as described in Materials and Methods. A, Serum-starved Ishikawa cultures were incubated for 30
and 60 min as indicated, with E2 (0.1 nM), or phorbol 12-myristate 13-acetate (100 ng/ml). B, Cultures were incubated for 20 or 30 min with
ral (1 nM), OHT (1 nM), or E2 (0.1 nM) followed by analysis of total RNA. C, Cells were exposed to 10% fetal bovine serum (S), E2, OHT (T) for
30 min, or preincubated with 2 ␮M U0126 (U) for 10 min before addition of test agents. D, Results from three experiments were quantified with
densitometry, normalized to GAPDH, and averaged. Values were expressed relative to vehicle controls set at 100%. Error bars represent SEM.
*, P ⬍ 0.05 relative to control (C).
FIG. 5. ERK inhibition and ER phosphorylation
mutant do not affect ERE-LUC regulation. Ishikawa
cells were cotransfected with p3xERE-LUC and either wild-type (wt) ER␣ or mutant (S118A), and
maintained overnight in normal growth medium.
Cultures were then challenged with E2 (0.1 nM) for
2 h in absence or presence of U0126 (U, 2 ␮M), followed by removal of test agents and incubation for
an additional 16 h. LUC activity was expressed relative to vehicle controls and normalized to ␤-galactosidase. Values were averaged for three experiments with error bars representing SEM.
Estrogen-induced DNA synthesis is ERK1/2 dependent
Because ERK activity influences cell growth and/or differentiation, we reasoned that E2 and SERMS might stimu-
late DNA replication in Ishikawa cells. To determine the
effects of E2, OHT, and ral on DNA synthesis rates, cultures
were incubated for 2 d in low-serum MEM with or without
126
Endocrinology, January 2003, 144(1):121–128
ER ligands as indicated in Fig. 6. Only E2 at 1 nm induced
significant increases in [3H]-thymidine incorporation (Fig.
6A). Lower levels of E2 (0.1 nm), also showed some increased
uptake, but the changes were not statistically significant in
three of four experiments. When U0126 was included (Fig.
6B), E2 did not produce any change in thymidine uptake,
suggesting that ERK is required for estrogen’s effect. Because
ERK was not required for ERE-LUC expression, direct ERmediated transcriptional activity alone may not be sufficient
to promote DNA synthesis.
Discussion
The Ishikawa 3-H-12 cells used in this study are a moderately differentiated line exhibiting an estradiol binding
capacity of 146 fmol/mg protein (20). Our findings indicated
that both known ER isoforms were expressed in these cells.
Bhat and Pezzuto (26) also noted expression of both ER␣ and
ER␤ in Ishikawa cells. Surprisingly, there was only modest
E2-mediated induction of the ERE-LUC reporter gene without cotransfection of ER␣. This could indicate limiting levels
FIG. 6. Estrogen promotes ERK-dependent
DNA synthesis in Ishikawa cells. A, Serum
deprived cultures were exposed to E2 (1 and
0.1 nM), OHT (1 nM), or ral (1 nM) for 48 h.
During the final 6 h of incubation, [3H]thymidine was added, and incorporation was
determined as described in Materials and
Methods. B, Cells were treated with E2 (1 nM)
with or without U0126 (5 ␮M) as indicated.
Similar results were attained in at least three
independent experiments. Error bars represent SEM (n ⫽ 4); *, P ⬍ 0.05 relative to vehicle
control (veh).
Singleton et al. • ER Ligand Effects in Ishikawa Cells
of endogenous ER in the context of our assay, or a potential
inhibitory effect by ER␤ through ␣:␤ heterodimer formation
or occupation of target genes by unliganded ␤ isoform, as
proposed by Hall and McDonnell (27). A number of clonal
Ishikawa lines were developed (20), differing in morphologic
characteristics and estrogen binding capacity, which may
also account for observed differences in estrogen responsiveness. Tamoxifen is an agonist in uterus, and the relative
ER␣ binding affinity of the 4-hydroxy metabolite (OHT) is
greater than that for E2 (1.75-fold) (28). However, a relatively
low concentration (1 nm) of OHT did not induce ERE-LUC
activity even when ER␣ was overexpressed. Ral possesses
high affinity for ER␣ as well (29), but also did not affect LUC
expression. These data present a clear divergence between
ERE activity, or lack thereof, and nongenomic action because
all three agents induced MAPK (ERK1/2) phosphorylation,
illustrating the independence of these effects.
To our knowledge, it has not been previously clarified as
to whether in vivo estrogen-mediated c-fos expression results
from ERE regulation, or rapid indirect activity. Data pre-
Singleton et al. • ER Ligand Effects in Ishikawa Cells
sented here indicate that c-fos induction is an ERK1/2 dependent event not involving direct genomic ER action. Although increased c-fos mRNA levels are clearly indicated, we
do not presently know if this is a result of induced transcription or a transient increase in mRNA stability. Nonetheless, the relatively low concentrations of E2 and OHT that
modulate c-fos expression could have significance in terms of
in vivo mitogenic effects. Alternatively, nongenomic and
genomic estrogen actions may be required to work in conjunction to promote cellular outcomes, as was recently proposed by Vasudevan et al. (30). Immediate early gene induction through ERK typically involves a serum response
element (SRE) (31, 32), in regulatory regions. Duan et al. (33)
recently demonstrated E2-induced expression of a c-fos promoter construct, which required an SRE within the human
c-fos gene fragment and was MAPK dependent. Our data
showing induction of c-fos mRNA by E2 and OHT in Ishikawa cells is consistent with ERK1/2 activation and SRE
involvement. Interestingly, ral exhibited no effect on c-fos
expression even though it activated ERK1/2 to a similar
extent and duration as OHT. Our results indicated a dependence on ERK activity for c-fos induction because the
MEK1/2 inhibitor U0126 prevented this action. The reason
for lack of immediate early gene regulation by ral is presently
unclear, but the finding is consistent with reports of no uterotrophic activity (34 –36). Possibly, other signaling events in
addition to ERK1/2 are required for c-fos induction by ER
agonists. Alternatively, ral may activate an unknown pathway that is inhibitory toward c-fos expression, or stability.
Like other SERMs, ral stimulates or antagonizes ER depending on cell/tissue type and coregulatory factor availability.
Nongenomic activity, such as ERK activation may contribute
to the tissue specific estrogenic effects of ral. Another recently
described estrogen-mediated action is cell preservation
through inhibition of apoptosis (10, 37, 38). This appears to
involve PI3 kinase and the downstream kinase B (Akt), or
ERK1/2, depending on cell type. In light of these reports, we
tested for rapid Akt activation by ER ligands in Ishikawa
cells. No effect on Akt phosphorylation was noted, indicating
no involvement of PI3 kinase.
Although our data indicate a requirement for active Ras,
the mechanism for ER-mediated MAPK activation is not well
understood. We observed that the pure ER antagonist ICI
182,780 did not completely inhibit ERK phosphorylation, and
alone had a slight positive effect. This may reflect the mechanism for nongenomic activity, in which receptor dimerization, DNA binding, and coactivator recruitment are presumably not involved. Song et al. (15) demonstrated that the
amino terminal activation function-1 of ER␣ binds to Shc,
leading to ERK activation in response to E2. Thus, ligands
such as ICI may promote an ER␣ conformation that is transcriptionally inactive, yet able to interact with signaling components including Shc. A number of investigators have proposed a subpopulation of ERs associated with the plasma
membrane, which are required for rapid actions by E2 (10,
39 – 42). The present study suggests involvement of ER␣
functioning in proximity to the Ishikawa cell membrane.
However, further experiments must be performed to confirm
the presence and precise location of membrane bound ER.
Much work is currently being done to elucidate the po-
Endocrinology, January 2003, 144(1):121–128 127
tential importance of estrogen’s nongenomic actions in mitogenic signaling. This study found that ER-mediated ERK
activation and subsequent c-fos induction were not sufficient
for promotion of DNA synthesis because relatively low concentrations of E2 and OHT did not induce thymidine uptake.
However, ERK activity was required for E2-induced DNA
synthesis because U0126 inhibited this effect. In contrast,
ERK was not involved in ERE-mediated transcriptional regulation, because neither U0126, nor the (S118A) ER␣ mutant
had any effect on reporter gene expression. Taken together,
these data suggest that nongenomic and classic ER␣ transcriptional activities are independent events, but each is required for estrogen induced DNA replication in uterine Ishikawa cells. As mentioned earlier, rapid effects by estrogen
have been observed only in a limited number of tissue and
cell types. In addition to Ishikawa cells, we also noted E2mediated ERK activation in ER␣-transfected HeLa cells and
MCF7 breast cancer cells (data not shown). It should be
emphasized, however, that these are cancer cell lines whose
responses may differ considerably from nonneoplastic tissues. Therefore, conclusions such as those described here
should be confirmed in other models and in vivo as well.
In summary, we have characterized a rapid stimulation of
the Ras/ERK1/2 pathway in human endometrial cells by
estradiol and partial ER agonists OHT and ral. This effect is
linked to induction of proto-oncogene c-fos expression and is
likely independent of direct ER-mediated transcription. Active Ras and ER␣ were required for stimulation of ERK
indicating the possibility of interaction between liganded ER
and growth factor/cytokine signaling components. These
events were not sufficient to stimulate DNA synthesis in
Ishikawa cells, but do appear necessary. Therefore, our
model suggests that nongenomic effects are key components
in estrogen action, but not the predominant means in which
cells are induced to replicate DNA. Future studies will continue to examine rapid estrogen activity, with respect to
mechanism, existence of nonnuclear ER, and physiological
significance.
Acknowledgments
The authors wish to thank Dr. Masato Nishida (Kasumigaura National Hospital, Japan) for providing the Ishikawa cells, Dr. Jeff Molkentin (Children’s Hospital Research foundation, Cincinnati, OH) for
giving us the RasN17 adenovirus, and Dr. Tim Wilson (Glaxo Wellcome,
Research Triangle Park, NC) for generously providing ral. We also thank
Robin Tharakan for critical review of this manuscript.
Received June 17, 2002. Accepted October 8, 2002.
Address all correspondence and requests for reprints to: Dr. Sohaib
Khan, University of Cincinnati, Department of Cell Biology, Neurobiology and Anatomy, Vontz Center for Molecular Studies, 3125 Eden
Avenue, Cincinnati, Ohio 45267-0521. E-mail: [email protected].
NIH Training Grant T-32-CA-59268, Department of Defense DAMD17-00-1-0057, and NIH Grant R01-CA-72039 supported this work.
References
1. Watson CS, Norfleet AM, Pappas TC, Gametchu B 1999 Rapid actions of
estrogens in GH3/B6 pituitary tumor cells via a plasma membrane version of
estrogen receptor-␣. Steroids 64:5–13
2. Pappas T, Gametchu B, Yannariello-Brown J, Collins TJ, Watson C 1994
Membrane estrogen receptors in GH3/B6 cells are associated with rapid estrogen-induced release of prolactin. Endocrine 2:813– 822
3. Chen Z, Yuhanna IS, Galcheva-Gargova Z, Karas RH, Mendelsohn ME,
128
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
Endocrinology, January 2003, 144(1):121–128
Shaul PW 1999 Estrogen receptor ␣ mediates the nongenomic activation of
endothelial nitric oxide synthase by estrogen. J Clin Invest 103:401– 406
Hayashi T, Yamada K, Esaki T, Kuzuya M, Satake S, Ishikawa T, Hidaka H,
Iguchi A 1995 Estrogen increases endothelial nitric oxide by a receptormediated system. Biochem Biophys Res Commun 214:847– 855
Panickar KS, Purushotham K, King MA, Rajakumar G, Simpkins JW 1998
Hypoglycemia-induced seizures reduce cyclic AMP response element binding
protein levels in the rat hippocampus. Neuroscience 83:1155–1160
Gridley KE, Green PS, Simpkins JW 1998 A novel, synergistic interaction
between 17 ␤-estradiol and glutathione in the protection of neurons against
␤-amyloid 25–35-induced toxicity in vitro. Mol Pharmacol 54:874 – 880
Bi R, Broutman G, Foy MR, Thompson RF, Baudry M 2000 The tyrosine
kinase and mitogen-activated protein kinase pathways mediate multiple effects of estrogen in hippocampus. Proc Natl Acad Sci USA 97:3602–3607
Singer CA, Figueroa-Masot XA, Batchelor RH, Dorsa DM 1999 The mitogenactivated protein kinase pathway mediates estrogen neuroprotection after
glutamate toxicity in primary cortical neurons. J Neurosci 19:2455–2463
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
Razandi M, Pedram A, Levin ER 2000 Plasma membrane estrogen receptors
signal to antiapoptosis in breast cancer. Mol Endocrinol 14:1434 –1447
Castoria G, Barone MV, Di Domenico M, Bilancio A, Ametrano D, Migliaccio A, Auricchio F 1999 Non-transcriptional action of oestradiol and progestin triggers DNA synthesis. EMBO J 18:2500 –2510
Collins P, Webb C 1999 Estrogen hits the surface. Nat Med 5:1130 –1131
Levin ER 1999 Cellular functions of the plasma membrane estrogen receptor.
Trends Endocrinol Metab 10:374 –377
Kahlert S, Nuedling S, van Eickels M, Vetter H, Meyer R, Grohe C 2000
Estrogen receptor ␣ rapidly activates the IGF-1 receptor pathway. J Biol Chem
275:18447–18453
Song RX, McPherson RA, Adam L, Bao Y, Shupnik M, Kumar R, Santen RJ
2002 Linkage of rapid estrogen action to MAPK activation by ER␣-Shc association and Shc pathway activation. Mol Endocrinol 16:116 –127
Simoncini T, Hafezi-Moghadam A, Brazil DP, Ley K, Chin WW, Liao JK 2000
Interaction of oestrogen receptor with the regulatory subunit of phosphatidylinositol-3-OH kinase. Nature 407:538 –541
Schlegel A, Wang C, Katzenellenbogen BS, Pestell RG, Lisanti MP 1999
Caveolin-1 potentiates estrogen receptor ␣ (ER␣) signaling. caveolin-1 drives
ligand-independent nuclear translocation and activation of ER␣. J Biol Chem
274:33551–33556
Kousteni S, Bellido T, Plotkin LI, O’Brien CA, Bodenner DL, Han L, Han K,
DiGregorio GB, Katzenellenbogen JA, Katzenellenbogen BS, Roberson PK,
Weinstein RS, Jilka RL, Manolagas SC 2001 Nongenotropic, sex-nonspecific
signaling through the estrogen or androgen receptors: dissociation from transcriptional activity. Cell 104:719 –730
Shepard SB, Cooper AG 2000 Assessing the expression of two genes simultaneously in surgical specimens using polymerase chain reaction. Mod Pathol
13:401– 406
Nishida M, Kasahara K, Oki A, Satoh T, Arai Y, Kubo T 1996 Establishment
of eighteen clones of Ishikawa cells. Hum Cell 9:109 –116
Nephew KP, Peters GA, Khan SA 1995 Cellular localization of estradiolinduced c-fos messenger ribonucleic acid in the rat uterus: c-fos expression and
uterine cell proliferation do not correlate strictly. Endocrinology 136:3007–3015
Nephew KP, Polek TC, Khan SA 1996 Tamoxifen-induced proto-oncogene
expression persists in uterine endometrial epithelium. Endocrinology 137:
219 –224
Kato S, Endoh H, Masuhiro Y, Kitamoto T, Uchiyama S, Sasaki H, Masushige
S, Gotoh Y, Nishida E, Kawashima H, Metzger D, Chambon P 1995 Acti-
Singleton et al. • ER Ligand Effects in Ishikawa Cells
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
vation of the estrogen receptor through phosphorylation by mitogen-activated
protein kinase. Science 270:1491–1494
Joel PB, Traish AM, Lannigan DA 1995 Estradiol and phorbol ester cause
phosphorylation of serine 118 in the human estrogen receptor. Mol Endocrinol
9:1041–1052
Castano E, Chen CW, Vorojeikina DP, Notides AC 1998 The role of phosphorylation in human estrogen receptor function. J Steroid Biochem Mol Biol
65:101–110
Bhat KP, Pezzuto JM 2001 Resveratrol exhibits cytostatic and antiestrogenic
properties with human endometrial adenocarcinoma (Ishikawa) cells. Cancer
Res 61:6137– 6144
Hall JM, McDonnell DP 1999 The estrogen receptor ␤-isoform (ER␤) of the
human estrogen receptor modulates ER␣ transcriptional activity and is a key
regulator of the cellular response to estrogens and antiestrogens. Endocrinology 140:5566 –5578
Fang H, Tong W, Shi LM, Blair R, Perkins R, Branham W, Hass BS, Xie Q,
Dial SL, Moland CL, Sheehan DM 2001 Structure-activity relationships for a
large diverse set of natural, synthetic, and environmental estrogens. Chem Res
Toxicol 14:280 –294
Bryant HU 2001 Mechanism of action and preclinical profile of raloxifene, a
selective estrogen receptor modulation. Rev Endocr Metab Disord 2:129 –138
Vasudevan N, Kow LM, Pfaff DW 2001 Early membrane estrogenic effects
required for full expression of slower genomic actions in a nerve cell line. Proc
Natl Acad Sci USA 98:12267–12271
Karin M 1994 Signal transduction from the cell surface to the nucleus through
the phosphorylation of transcription factors. Curr Opin Cell Biol 6:415– 424
Treisman R 1995 Journey to the surface of the cell: Fos regulation and the SRE.
EMBO J 14:4905– 4913
Duan R, Xie W, Burghardt RC, Safe S 2001 Estrogen receptor-mediated
activation of the serum response element in MCF-7 cells through MAPKdependent phosphorylation of Elk-1. J Biol Chem 276:11590 –11598
Seeman E 2001 Raloxifene. J Bone Miner Metab 19:65–75
Sexton MJ, Gherman RB 2001 Selective estrogen receptor modulators: the
ideal estrogen replacement? Prim Care Update Ob Gyns 8:25–30
Al-Jamal JH, Dubin NH 2000 The effect of raloxifene on the uterine weight
response in immature mice exposed to 17␤-estradiol, 1,1,1-trichloro-2, 2-bis(pchlorophenyl)ethane, and methoxychlor. Am J Obstet Gynecol 182:1099 –1102
Honda K, Sawada H, Kihara T, Urushitani M, Nakamizo T, Akaike A,
Shimohama S 2000 Phosphatidylinositol 3-kinase mediates neuroprotection
by estrogen in cultured cortical neurons. J Neurosci Res 60:321–327
Harms C, Lautenschlager M, Bergk A, Katchanov J, Freyer D, Kapinya K,
Herwig U, Megow D, Dirnagl U, Weber JR, Hortnagl H 2001 Differential
mechanisms of neuroprotection by 17␤-estradiol in apoptotic versus necrotic
neurodegeneration. J Neurosci 21:2600 –2609
Chambliss KL, Yuhanna IS, Mineo C, Liu P, German Z, Sherman TS, Mendelsohn ME, Anderson RG, Shaul PW 2000 Estrogen receptor ␣ and endothelial nitric oxide synthase are organized into a functional signaling module
in caveolae. Circ Res 87:E44 –E52
Stefano GB, Prevot V, Beauvillain JC, Cadet P, Fimiani C, Welters I, Fricchione GL, Breton C, Lassalle P, Salzet M, Bilfinger TV 2000 Cell-surface
estrogen receptors mediate calcium-dependent nitric oxide release in human
endothelia. Circulation 101:1594 –1597
Donahue JE, Stopa EG, Chorsky RL, King JC, Schipper HM, Tobet SA,
Blaustein JD, Reichlin S 2000 Cells containing immunoreactive estrogen
receptor-␣ in the human basal forebrain. Brain Res 856:142–151
Norfleet AM, Thomas ML, Gametchu B, Watson CS 1999 Estrogen receptor-␣
detected on the plasma membrane of aldehyde-fixed GH3/B6/F10 rat pituitary tumor cells by enzyme-linked immunocytochemistry. Endocrinology
140:3805–3814