| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
and ERß Expressed in Chinese Hamster Ovary Cells
Division of Endocrinology (E.R.L.) Long Beach Veterans Affairs
Medical Center Long Beach, California 90822
Departments
of Medicine (M.R., A.P., E.R.L.) and Pharmacology (E.R.L.)
University of California, Irvine Irvine, California 92717
The Ben May Institute (G.L.G.) University of Chicago
Chicago, Illinois 60637
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
or ERß into Chinese hamster ovary (CHO) cells.
Transfection of ER resulted in a single transcript by Northern blot,
specific binding of labeled 17ß-estradiol
(E2), and expression of ER in both nuclear and
membrane cell fractions. Competitive binding studies in both
compartments revealed near identical dissociation constants
(Kds) of 0.283 and 0.287
nM, respectively, but the membrane receptor
number was only 3% as great as the nuclear receptor density.
Transfection of ERß also yielded a single transcript and nuclear and
membrane receptors with respective Kd values of
1.23 and 1.14 nM; the membrane receptor number
was only 2% compared with expressed nuclear receptors. Estradiol
binding to CHO-ER or CHO-ERß activated Gq and Gs proteins in
the membrane and rapidly stimulated corresponding inositol phosphate
production and adenylate cyclase activity. Binding by
17-ß-E2 to either expressed receptor
comparably enhanced the nuclear incorporation of thymidine, critically
dependent upon the activation of the mitogen-activated protein
kinase, ERK (extracellular regulated kinase). In contrast, c-Jun
N-terminal kinase activity was stimulated by
17-ß-E2 in ERß-expressing CHO, but was
inhibited in CHO-ER cells. In summary, membrane and nuclear ER can
be derived from a single transcript and have near-identical affinities
for 17-ß-E2, but there are considerably more
nuclear than membrane receptors. This is also the first report that
cells can express a membrane ERß. Both membrane ERs activate G
proteins, ERK, and cell proliferation, but there is novel differential
regulation of c-Jun kinase activity by ERß and ER. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Recently, we showed that primary cultures of human vascular smooth
muscle cells express what appears to be membrane ER (19). The putative
membrane ER on vascular smooth muscle cells participates in the
inhibition of growth factor-induced cell proliferation. This effect
occurs when 17-ß-E2 interferes with the ability of growth
factors to enact signal transduction to the nuclear growth program
(20). Estrogen can also inhibit growth factor signaling, which enhances
gene transcription (21, 22). This novel mechanism by which ER
negatively modulates transcription is probably mediated through a
putative membrane ER (21). However, the isolation and structural
characterization of the membrane ER have not been described, and the
derivation, roles, and cell functions of this putative protein are
still largely unknown. To begin to address these issues, we transfected
Chinese hamster ovary (CHO) cells with the cDNAs for ER and ERß
and carried out characterization studies in these cells, which normally
do not express ER. We found several novel aspects of ER expression and
signal transduction and evidence that the membrane ER importantly
contributes to 17-ß-E2 action in this model.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
or ERß Result in
the Expression of Both Cell Membrane and Nuclear Receptors or ERß. In both
situations, a single transcript was detected in either ER- or
ERß-expressing cells, determined by Northern analysis; this was
expected from the expression of single cDNAs for each receptor (Fig. 1A
). Empty vector-transfected (lanes 1)
or non transfected CHO (data not shown) did not express ER mRNA.
|
and 60 kDa for ERß (Fig. 1B
). This is comparable to what is
reported for the size of the native ER (22) or for ERß expressed
after transfection (23). Thus, both cell membrane and nuclear receptor
proteins result from expression of a single cDNA for either ER
subtype.
Competition binding studies on transiently transfected CHO cells were
then carried out and analyzed by Scatchard analysis. In initial
studies, high efficiency expression of ER or ERß after transient
transfection was revealed in that 60–70% of added
[3H]17-ß-E2 bound specifically to the CHO
cells. We estimated that the total number of ER expressed was
approximately 4 million/transfected CHO cell. In contrast,
nontransfected CHO failed to show specific binding of labeled
17-ß-E2. The ER nuclear and membrane (Fig. 1, inset) receptors had nearly identical affinities for
17-ß-E2, of 0.283 ± 0.017 and 0.287 ± 0.011
nM, respectively (Fig. 2A),
while the receptor densities in these two compartments were greatly
disparate, there being almost 40-fold more nuclear than membrane
receptors (nuclear Bmax is 362 ± 26 pM;
membrane Bmax is 9.7 ± 0.6 pM). For
ERß, nuclear and membrane receptors demonstrated near-equal estradiol
affinities of 1.23 ± 0.08 and 1.14 ± 0.06 nM,
respectively, and a receptor density ratio of 62:1 (nuclear
Bmax is 1.62 ± 0.07 nM; membrane
Bmax is 21 ± 0.8 pM) (Fig. 2B). The lower
affinity of ERß compared with ER for 17-ß-E2 has
been reported previously (24), confirming the utility of our model.
These results indicate that the receptors in the two compartments have
near-identical affinities for 17-ß-E2, but many more
nuclear than membrane receptors are expressed.
|
cells were incubated with fluorescein isothiocyanate
(FITC)-conjugated E2-BSA (Fig. 2C). E2-BSA has
been shown to not enter the cell and has been found by several
laboratories, including our own, to bind putative membrane ER, although
with lower affinity than 17-ß-E2 (18, 19). The lower
affinity may be due to the BSA protein physically hindering
17-ß-E2 binding to its receptor. Labeling/binding by
E2-BSA-FITC of the outer cell membrane of most transfected
cells was evident (panel A), consistent with the expression of membrane
ER. No labeling of nontransfected CHO-K1 cells was seen with this
compound (data not shown). The binding was competed off by unlabeled
17-ß-E2 (panel B), the ER receptor antagonist, ICI
182,780 (C), or H222 (D), an antibody directed against the
ligand-binding domain (LBD) of ER (25). Inhibition by the H222 ER
antibody is especially important, since it is extremely unlikely that
this antibody can enter the cell and bind cytoplasmic/nuclear ER,
especially after a short incubation with the CHO-ER. The specificity
of this well characterized antibody also strongly supports the idea
that E2-BSA is specifically binding a membrane ER, one
whose LBD is similar to that of the nuclear ER. In contrast, ER-21, an
antibody directed against the N-terminal first 21 amino acids of ER
(26), did not compete off binding/labeling (panel E). Because antibody
specific to the LBD of ERß was not available to us for these studies,
we were not able to carry out comparable studies in CHO-ERß.
To demonstrate the nuclear pool of ER, we permeabilized the CHO-ER
cells by incubation with detergent; this resulted in a dense nuclear
labeling (panel F). This finding is consistent with the binding
studies, which showed ER in both membrane and nuclear compartments and
that the number of nuclear receptors greatly exceeds those localized to
the cell membrane. The results also support our previous ER labeling
studies of a membrane receptor expressed in primary cultures of
nontransfected human vascular smooth muscle cells (19).
Membrane ERs Are G Protein-Coupled Receptors
The ability of 17-ß-E2 to rapidly stimulate cAMP
generation or Ca++i mobilization has been
reported by others (9, 10). This led us to hypothesize that the
membrane ER could stimulate G protein-induced signal transduction,
qualifying this receptor as G protein linked (directly or indirectly).
To support this idea, we determined the ability of
17-ß-E2 to stimulate the production of IP3.
In ER or ERß expressing CHO cells, 10 nM
17-ß-E2 stimulated a respective 210% and 122% increase
in IP3 generation above control (ER or ERß expressed
in the absence of 17-ß-E2 incubation) (Fig. 3A). The respective stimulations were
prevented 56% and 53% by ICI 182,780, while results with ICI
alone were similar to control. These data suggest that the membrane ER
may couple to Gq activation, which enacts inositol phosphate
hydrolysis through the activation of phospholipase C in many cells
(27).
|
or ERß; this was determined by measuring cAMP
as a readout (Fig. 3B). Interestingly, neither ICI 182,780 nor
tamoxifen significantly inhibited 17-ß-E2-activation of
adenylate cyclase and actually significantly stimulated enzyme activity
in the absence of E2. These latter findings are very
similar to those reported by Aronica et al. (9), concerning
cAMP generation in MCF-7 cells. Collectively, these results could
indicate that the membrane ER is linked to Gs.
To directly show that the membrane ER activates G proteins, we again
used membrane preparations from ER- and ERß-expressing CHO cells.
Increased binding of guanosine 5'-3-O-(thio) triphosphate
(GTP
S) to Gs or Gq in response to 10 nM
17-ß-E2 in CHO-ER was seen at 5 min;
17-ß-E2 increased the respective binding to Gs and
Gq by 93% and 70% above control (Fig. 3, C and D, lanes 2
vs. 3). Preincubation with either ICI 182,780 (lane 5 in
both figures) or the ER H222 antibody (lane 6), reduced the
augmented binding by 70–85%. By contrast, incubation with
17--E2 (lane 4) had no effect on GTPS binding to the
G proteins in the CHO-ER membrane. Similarly, in CHO-ERß cells,
17-ß-E2 (but not 17--E2) stimulated
GTPs binding to Gs and Gq above basal levels by 83% and 61%
(Fig. 3, C and D, lanes 10 vs. 11), only slightly less than
in cells expressing ER. ICI 182,780 inhibited G protein activation
by 50–60% in each of the ERß studies (lane 13). These data indicate
that both membrane ERs have the capacity to stimulate signal
transduction through the activation of Gs and Gq proteins, accounting
for the generation of cAMP (Gs) and IP3 (Gq).
Stimulation of Extracellular Regulated Kinase (ERK) Activity
through the Membrane Receptor Is Necessary for the Stimulation of Cell
Proliferation
Expression of either ER or ERß in the CHO cells resulted in
the significant stimulation of ERK activity by 17-ß-E2
(Fig. 4A). Stimulation of ERK activity
occurred after 10 min incubation with 17-ß-E2 in the ER-
transfected CHO (lanes 3 and 9), compared with the control
vector-transfected (lane 1) or ER-transfected CHO incubated in the
absence of 17-ß-E2 (lane 2). Nontransfected CHO also did
not respond to 17-ß-E2 (data not shown).
|
- or ERß-expressing CHO cells (Fig. 4A).
ERK activation by either 17-ß-E2 or E2-BSA
was approximately 3- to 4-fold above basal activity (lane 2
vs. lanes 3 and 4, or lane 8 vs. lanes 9 and 10)
and was not appreciably different for ER or ERß. In
ER-expressing CHO cells, approximately 70% of 17-ß-E2
or E2-BSA-enhanced ERK activity was inhibited by
coincubation of the cells with ER antibody (lanes 5 and 6). Since
the antibody is highly unlikely to enter the cell, these results
support the idea that 17-ß-E2 (and E2-BSA)
activate ERK through the membrane receptor. In the ERß expressing
CHO, we found that ICI 182,780 inhibited 17-ß-E2 or
E2-BSA activation of ERK by 60 and 70%, respectively
(lanes 9 and 10 vs. lanes 11 and 12). BSA itself was found
not to affect ERK activity in our previous studies (19). It is possible
that the E2-BSA compound could have dissociated, allowing
E2 to enter the cell, or that E2-BSA was
internalized. In either situation, activation of the nuclear ER could
contribute to the results shown here. To validate the use of this
compound to activate only the membrane ER, CHO-ER cells were
cotransfected with an ERE-luciferase reporter, and then incubated with
either 17-ß-E2, 10 nM, or E2-BSA,
100 nM, for 10 min or 24 h. Luciferase activity at 10
min was not affected by either of the treatments. At 24 h,
17-ß-E2 stimulated a 2.5-fold increase above baseline in
luciferase activity, but E2-BSA had no effect on reporter
function (data not shown). These results corroborate similar findings
from Watters et al. (20), and indicate that
E2-BSA probably does not bind the nuclear ER.
It is widely accepted that growth factors stimulate cell proliferation
in part through the activation and subsequent translocation of ERK to
the nucleus (28). In target cells, ER has been shown to activate ERK
(29), presumably contributing to induced cell division in which
17-ß-E2 is mitogenic. As an index of cell proliferation,
we determined that 17-ß-E2 or E2-BSA could
induce a modest but consistent 42% and 27% increase in nuclear
thymidine incorporation in CHO-ER, respectively (Fig. 4B, left). Comparable stimulation was also seen in CHO-ERß
(Fig. 4B, right). It must be appreciated that CHO cells do
not normally respond to 17-ß-E2, and hence the endogenous
systems (e.g. signal protein scaffold mechanisms) that
mediate growth effects of the sex steroid in target cells, such as
breast or uterus, are probably not comparable in CHO. Nevertheless,
CHO-ER respond to 17-ß-E2 in a qualitatively comparable
way to MCF-7 cells (29), and the effects shown here are mediated
through ER since ICI 182,780 significantly inhibits this stimulation.
The ability of 17-ß-E2 or E2-BSA to enhance
DNA synthesis was reversed 77% and 85%, respectively, in CHO-ER,
and 64% and 75% in CHO-ERß by PD98059, a soluble MAP kinase kinase
(MEK) inhibitor (30). This indicates that the ability of
17-ß-E2 to stimulate ERK activity is required for the
stimulation of thymidine incorporation, and that E2-induced
MEK activation of ERK is involved. The substantial reversal of
E2-BSA-stimulated thymidine incorporation by PD98059 is of
additional interest. This finding particularly supports the idea that
stimulation of ERK activity through the cell membrane ER is
crucial to the proliferative action of the sex steroid. Further
supporting this role of the membrane ER, the ER antibody prevented
17-ß-E2 or E2-BSA stimulation of thymidine
incorporation by 74% and 86%, respectively, in CHO-ER cells (Fig. 4B, left). Progesterone or 17--E2 had no
effect on thymidine incorporation in either ER-expressing CHO
cell.
ERß, but Not ER, Stimulates c-Jun Kinase (JNK) Activity
To further investigate the effects of the two ERs in signal
transduction, the modulation of the activity of the important
mitogen-activated protein (MAP) kinase, JNK, was determined. This
kinase was activated by 17-ß-E2 but, surprisingly, only
in ERß-expressing cells (Fig. 5A). In
dose-related fashion, 17-ß-E2 stimulated a maximum 2-fold
increase above basal activity (lane 9), inhibited 85% by ICI 182,780.
No effect on JNK activity was observed in response to
17--E2. By contrast, in ER-expressing CHO cells, JNK
activity was inhibited by 17-ß-E2. Maximal inhibition was
55% of basal JNK activity at 10 nM 17-ß-E2
(lane 3), again prevented by ICI 182,780. Activation or inhibition was
most prominent at 15 min; based upon preliminary time course studies,
these same effects were seen by 10 min exposure to
17-ß-E2 (data not shown).
|
). Both compounds caused a
significant and comparable increase in JNK activity above control
(basal) levels, and the stimulations were approximately 65% inhibited
by ICI 182,780. These data support the ERß-expressing cell membrane
as mediating 17-ß-E2-activation of JNK. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
or ERß. We found that expression of cDNAs for either receptor
results in the production of a single transcript that gives rise to
both membrane and nuclear receptors. Membrane and nuclear receptors are
similarly sized and have nearly identical affinities for
17-ß-E2. We further report for the first time that ERß
can exist in the cell membrane. Both receptors enact a variety of
signal transduction events, some of which contribute to cell
proliferation. Our results indicate that the membrane ER is G protein
linked, and that this receptor should be considered to possibly
contribute to the cell biological functions of 17-ß-E2 on
normal target cells (9, 10, 11, 12, 13, 19, 21).
What is the structural nature of the membrane ER? Although we did not
experimentally examine this complicated issue, we speculate that the
protein must be very similar to the classical nuclear ER. Supporting
this position, both membrane and nuclear proteins are derived from the
same transcript. In target cells, alternatively spliced transcripts for
ER have been identified that give rise to ER proteins of different
lengths. However, here we demonstrate that a single cDNA and RNA are
capable of producing both membrane and nuclear receptors, although the
great majority of receptors are nuclear. Posttranslational modification
of some ER protein must occur to ensure targeting to the membrane.
Targeting modifications would likely include the addition of lipid
anchors such as glycosylphosphatidylinositol (31) and/or other lipid
modifications such as palmitoylation or myristolation; these
alterations would promote movement to the cell membrane. However, in
analyzing the structure of ER, there are no obvious palmitoylation
sites and only a few candidate myristolation sites. It is also not
obvious from a structural analysis how the ER would insert into the
cell membrane. Insertion would require a core hydrophobic region of the
receptor, and one potential region has been previously noted (18).
There is precedent for a subpopulation of ER to be posttranslationally
modified, glycosylated by N-acetylglucosamine (32).
Cytoskeletal and oncoproteins are substrates for this enzyme, and
several of these proteins protrude from the plasma membrane, suggesting
that this modification might play a role in the translocation of ER to
the membrane. However, our cross-linking studies show that the size of
both membrane and nuclear receptors is approximately the same. This
indicates that there can not be extensive modification of the membrane
ER protein, such as by the addition of bulky glycosyl groups which
substantially increase the receptor size and alter gel migration. A
subtle glycosylation of the membrane ER is more plausible, and this
analysis is ongoing.
Additional proof of similarity between membrane and nuclear ER was
provided by our studies. An antibody raised against the LBD of the
nuclear ER competes for binding of labeled 17-ß-E2 to
the putative ER cell membrane (shown here and in Ref. 19). Pappas
et al. (18) previously showed that antibodies raised against
the classical ER identify a membrane ER. Thus, by ligand affinity,
receptor protein size, and immunological/shared epitope criteria, the
membrane and nuclear receptors are similar. These findings presuppose
that our membrane preparations are not contaminated with cytosolic ER
(our isolation procedure precludes nuclear ER). Although we previously
established the lack of cytosol in our typical membrane preparations
(33), we cannot exclude very small amounts of cytosolic membrane
(e.g. endoplasmic reticulum) in our samples. However, here
we show that G proteins and adenylate cyclase are activated by
17-ß-E2 in membrane preparations and, according to
current knowledge, this cannot result from cytosolic membrane
binding.
We report that both expressed ERs are capable of stimulating a modest
but consistent increase in nuclear thymidine incorporation. Since the
cells are not normally responsive to 17-ß-E2, it is
perhaps remarkable to see this effect. However, it is known that many
signal transduction molecules are present in CHO, and hence these cells
can serve as a model to understand some of the effects of ER. We found
that ER and ERß comparably activate ERK activity, and that this is
necessary for their proliferative effects, based upon reversal by a
soluble MEK inhibitor, PD98059; MEK is the consensus upstream activator
of ERK (30). The stimulation of ERK activity by 17-ß-E2
has been previously shown in MCF-7 cells (29), mediated through
undetermined subtype or membrane/nuclear receptors. In these same
cells, 17-ß-E2 was shown to trigger a signal cascade
including the activation of the Src and Ras tyrosine kinases, and the
phosphorylation of the adapter protein, Shc. Similar signal
transduction in response to progesterone has been reported to be
mediated through ER (34). The activation of ERK by
17-ß-E2 has been attributed previously to signaling
through a membrane ER in a neuroblastoma cell line (20), and we showed
that when 17-ß-E2 is an antigrowth factor, this effect is
mediated through membrane ER-induced inhibition of ERK activation by
growth factors (19). Furthermore, inhibition of transcription by ER can
be, in part, mediated through interrupting stimulatory signaling
through ERK in endothelial cells (21).
Thus, although ER stimulates signal transduction in some cells, the receptor inhibits these same molecules in other cells. This likely requires the assembly of different intermediary signal molecules in different cell types, activated by the membrane receptor. Understanding the mechanisms involved should provide insight into the differential effects of 17-ß-E2, including transcriptional modulation. Negative transcriptional regulation may be an important function of the membrane ER (21).
As a positive effector of growth-related genes, the membrane ER might
activate growth factor receptor tyrosine kinases (TK). Most notably, ER
might activate the EGFR, which then enacts signal transduction to
stimulate ERK activity (35). ERK nuclear translocation and activation
of targets such as c-fos (36) or egr-1 (37) would
then activate cell proliferation. Here, we found that added EGF could
not activate ERK in the transfected CHO-ER cells, consistent with
the absence of EGFR in these cells. Further, when the CHO-ER cells
were incubated with 17-ß-E2 in the presence (or absence)
of the specific EGFR-TK antagonist, tyrphostin AG127 (38), there was no
difference in ERK activation (data not shown). We therefore propose
that Gq activation by 17-ß-E2 in transfected CHO cells
leads to ERK activation, as has been demonstrated for a variety of G
protein-coupled transmembrane receptors (39).
By investigating additional signal transduction effects, we found that
ERß is capable of activating JNK, while ER inhibits basal JNK
activity. To our knowledge, this is the first report that ER modulates
this proline-directed serine/threonine kinase in any model. Further,
this is one of the few differential effects of the two ERs reported to
date. It was recently found that ER activates AP-1-dependent
transcription, while ERß inhibits this action of AP-1 (40). It is
unlikely that activation of JNK by ERß (which we show here)
contributes to the previously described inhibition of AP-1-induced
transcription (40): AP-1 is composed of fos/jun family heterodimers,
and c-jun is a natural substrate for JNK activity, which usually
enhances AP-1 activity. On the other hand, stimulation of this MAP
kinase by ERß could contribute to cell proliferation, because it is
now believed that activation of JNK by growth factors contributes to
this function (41). These findings may be especially relevant to
natural target cells expressing predominantly ERß.
Based upon signal transduction events triggered by ER in a variety of
cell types (9, 10, 11, 12, 13), it can be speculated that 17-ß-E2
activates Gs and Gq by binding to a cell membrane receptor. Here, we
provide the first direct evidence that this occurs. In
membrane preparations, we showed that each expressed ER stimulates the
activation of both Gs and Gq, as well as adenylate cyclase in the
membrane and increased production of IP3 determined from whole cells.
The stimulation of most of these events was significantly inhibited by
the specific ER antagonist ICI 182,780, and in CHO-ER cells, H222
antibody directed against the LBD prevented G protein activation.
Interestingly, we showed that 17-ß-E2 activates Gs,
and that this is reversed by ICI 182,780; however, we also found that
this ER antagonist independently stimulates adenylate cyclase activity
in ER-expressing CHO cells. These seemingly dichotomous findings
indicate that ICI 182,780 is capable of inhibiting ER-mediated G
protein activation but can also stimulate adenylate cyclase, perhaps by
reducing the Michaelis-Menton constant (Km) of the enzyme.
Alternatively, ICI might prevent cAMP degradation. However, we think
that this is unlikely since Aronica et al. (9) also showed
that ER antagonists stimulate cAMP production but have no effect on
phosphodiesterase activity.
These data strongly support the idea that membrane ER couple in some way to G proteins and transduce intracellular signals. The structure of the membrane ER that potentially allows contact with G proteins awaits isolation of the protein, but it is very unlikely that the protein is structurally similar to a typical, heptahelical G protein-coupled receptor. We (19, 21) and others (9, 10, 13, 20, 29) have previously implicated intracellular signaling by ER as contributing significantly to the cellular effects of the sex steroid. The present demonstration of ER-induced G protein activation supports and lends mechanistic details to those previous studies.
One caveat to our results is that we have used an artificial system.
However, cells expressing only ER or ERß are not readily available
for culture, nor are there available discrete receptor antagonists or
ligands for each individual receptor. Therefore, it is difficult
to separate out the independent effects of the two receptor isoforms in
native cells at this time. Studies of cells from ER knockout animals
will be of use in further defining the relative contributions of the
two receptors. However, our model lends itself to identifying and
implicating the membrane form of each ER to signal and thereby
potentially contribute to the cell physiology modulated by ER.
In summary, we present the first direct evidence that membrane and
nuclear ER arise from a single transcript, implying that novel
posttranslational processing occurs on a minority of these proteins.
How this occurs, and how the ER translocates and inserts into the cell
membrane requires further study. We also show that a membrane ERß can
exist, that it can signal, and that there are both common and unique
signal transduction pathways enacted by membrane ER and ERß. The
intracellular actions of ER and other members of the steroid receptor
superfamily arise from both genomic and nongenomic effects. Direct
interaction between progesterone and G protein-coupled cell membrane
receptors has been shown recently (42). Signal transduction through
cell membrane receptors may regulate discrete actions of steroid
ligands and may play a significant role in the cross-talk between
tyrosine kinase receptors, such as EGF, and ERs (35). Just as several
signal transduction systems target and amplify nuclear events (such as
c-fos expression), there is likely to be cooperation between
both signal transduction and direct transcriptional effects of the
membrane and nuclear ERs, respectively (21). The challenge is to
identify and define the unique and interactive contributions of each
mechanism.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Transient Transfections
CHO-K1 cells were grown to 40–50% confluence and then
transiently transfected with 0.5–10 µg of fusion plasmids, depending
on the plate size and the amount of cells; the plasmids included
pcDNA3mER encoding nucleotides 17–2001 (43) and pSG5mERß encoding
nucleotides 12–1469 (23) (kindly provided by Dr. K. Korach and Dr.
J. A. Gustafsson via Dr. Korach, respectively) or respective
backbone vectors. Transfections were done with Lipofectamine Reagent
(GIBCO-BRL, Grand Island, NY); cells were incubated with liposome-DNA
complexes at 37 C for 5 h, followed by overnight recovery in
DMEM-F12 medium containing 10% FBS. Then, before experimental
treatment, cells were synchronized in serum-free DMEM-F12 for 24 h
and then treated with 17-ß-E2 and/or related compounds.
Cotransfections with a green fluorescent protein expression vector
(Promega, Madison, WI) indicated 80–90% efficiency of transfection,
corroborated by the labeling of cells with FITC-E2-BSA. In
other studies, CHO cells were cotransfected with pcDNA3mER and
ERE-SV40 Luciferase (kindly provided by Dr. B. Gehm) and then incubated
with 10 nM 17-ß-E2 or 100 nM
E2-BSA (Sigma Chemical Co., St Louis, MO) for 10 min and
24 h, and luciferase activity was determined as previously
described (44). The concentration of E2-BSA was calculated
from the number of E2 molecules attached to each BSA
molecule.
Northern Analysis
Total RNA was extracted from the CHO cells transfected with
ER or ERß plasmids, using Tri-Reagent-LS (Molecular Research
Center, Inc., Cincinnati, OH). RNA (25 µg) from each CHO-ER
type was separated by electrophoresis on a denaturing 1% agarose gel,
transferred to nitrocellulose, and prehybridized, as we previously
described (19, 45). The blots were hybridized for 12 h at 65 C
with 32P-labeled, antisense cRNAs for ER or ERß. The
probes were transcribed from a rat cDNA template (ER), nucleotides
81–398 (46) (kindly provided by Dr. E. Spreafico) and a mouse cDNA
template (ERß) (pCRSK+, nucleotides 49–310) (47) (kindly
provided by Dr. K. Korach), using T7 and T3 RNA polymerases,
respectively. Hybridization bands were quantified by laser
densitometry, after autoradiography. Sense probes produced no
hybridization.
Cross-Linking
For cross-linking studies,
[125I]17ß-E2 (2200 Ci/mmol) was bound to
CHO-ER or CHO-ERß. After binding, nuclear pellets and cell
membrane pellets were washed twice with PBS and then cross-linked by
incubating with 4% formaldehyde (Mannich Reaction) (48) at 57 C for
24 h. After washing, the samples were solubilized in sample buffer
(80 mM Tris-HCl, pH 6.8, 10% glycerol, 1% SDS, 0.025%
bromphenol blue). Proteins were denatured at 95 C for 5 min and then
loaded onto 8% polyacrylamide gel and electrophoresed. Gels were dried
and then subjected to autoradiography.
Binding Studies
Cells were grown on 100-mm petri dishes in DMEM-F12 without
phenol red. Twenty four hours after transfection with ER or ERß
constructs, the cultures were washed three times with PBS; cells were
then lysed in buffer A (50 mM Tris-HCl, pH 7.5, 5
mM EDTA, 100 nM NaCl, 50 mM NaF,
100 µM phenylmethylsulfonyl fluoride, protease inhibitor
cocktail, and 0.2% Triton X-100). Nuclear pellets were collected
through low-speed centrifugation. The supernatants were centrifuged at
100,000 x g for 30 min to pellet cell membranes. Both
pellets were washed twice, once with buffer A and once without
detergent. Fifty microliters of membrane proteins from CHO-ER or
CHO-ERß cells were incubated in buffer A without Triton X-100 and
with 0.5% BSA, increasing concentrations of unlabeled
17-ß-E2 (10-11 M to
10-7 M), and
[3H]-17-ß-E2 (specific activity, 80
Ci/mmol, pH 7.5) (New England Nuclear, Boston, MA) at 37 C for 45 min,
similar to binding studies previously described by us (49, 50). The
nuclear and cell membrane pellets were washed three times by
centrifugation to remove unbound isotope and then quantified by
ß-scintillation counting. Binding studies were repeated three times,
and data were combined for Scatchard analysis.
Membrane ER Labeling Studies
Nontransfected CHO and CHO-ER cells were grown on coverslips.
The cells were incubated at 4 C for 30 min with FITC-conjugated
E2-BSA (Sigma). The E2-BSA-FITC compound has
previously been shown to label a membrane ER in several cell types (18, 19). For competition studies, cells were preincubated for 5 min with
17-ß-E2, 10 nM, or ICI 182,780 1
µM, or antibodies to the LBD of the ER (H-222) (18) or N
terminus of the receptor protein (ER-21); this was followed by the
addition of E2-BSA-FITC. After labeling, the cells were
fixed for 3 min in freshly prepared 4% paraformaldehyde and mounted
for microscopic evaluations. Some cells were also permeabilized with
0.2% Triton X-100 to allow the labeling of the nuclear pool of ER.
Antibody was diluted 1:100 from an original concentration of 0.56 mg/ml
and was used for labeling and functional studies as noted.
Kinase Assays
ERK and JUN kinase assays were carried out as previously
described (51, 52). CHO-ER or -ERß was incubated with various
steroids alone, or coincubated with ER antagonist for 10 or 15 min,
based upon our preliminary time course studies. Cell lysate (200 µl)
was then added to Erk2 or Jnk-1 antiserum (Santa Cruz Biotechnology,
Santa Cruz, CA) (10 µl) conjugated to prewashed protein A-Sepharose;
the mixture was then incubated for 2 h at 4 C in microfuge tubes.
After washing, each bead-antibody-antigen complex was then incubated
for 30 min at 30 C with 40 µl kinase buffer (25 mM HEPES,
10 mM MgAc, 40 µM ATP, 2 mM
dithiothreitol), 10 µCi 32P-ATP, and myelin basic
protein or glutathione-S-transferase-c-jun
(Santa Cruz), as substrate for kinase activity. Reactions were
terminated by adding SDS reducing buffer, samples were boiled, and the
proteins were separated by SDS-PAGE. After autoradiography, the bands
were compared by laser densitometry. Western blot documented similar
amounts of kinase protein loaded under each condition.
Nuclear Thymidine Incorporation
Subconfluent CHO-ER were synchronized for 24 h in
serum-free media. All cells were then incubated for 20 h in the
absence or presence of 17-ß-E2 or E2-BSA,
with or without ICI 182,780 or the H222 antibody. In some conditions,
the MEK inhibitor, PD 98059, 20 µM (kindly provided by
Dr. A. Saltiel) was added to the incubation mixture 30 min before the
steroid. After 20 h, 0.5 µCi of [3H]thymidine was
added for 4 more hours, as previously described (19). Cells were then
washed in cold HBSS, incubated for 10 min with 10% trichloroacetic
acid at 4 C to precipitate the nuclear incorporated thymidine, washed,
and lysed with 0.2 N NaOH overnight, after which the
lysates were counted in a liquid scintillation ß-counter. Experiments
were repeated three to four times, and data were combined for analysis
by ANOVA plus post hoc test (Scheffe’s).
G Protein Activation Assays
Cell membranes from CHO-ER were prepared and analyzed for lack
of cytosol as we previously reported (32). Membrane aliquots (20 µg)
were resuspended in 50 mM Tris-HCl buffer, with 1
µMGTP, 100 µM Mg++ and
incubated with 30 nM [S35] GTPS (Sigma),
in the presence of 10 nM 17-ß-E2 for 5 min at
30 C. The incubation was terminated by adding 600 µl ice-cold 50
mM Tris-HCl, 20 mM MgCl2, 0.5%
Nonidet P-40, and 100 µM GTP. After 30 min, the extract
was placed into microfuge tubes containing 2 µl of nonimmune serum
preincubated with 150 µl of a 10% suspension of pansorbin cells
(Calbiochem, San Diego, CA.). Nonspecifically bound proteins were
removed by centrifugation after 20 min. The supernatant was then
incubated with Gs or Gq -subunit antibody (Calbiochem), preincubated
with 5% protein A Sepharose. Immunoprecipitants were washed in
extraction buffer without detergent and boiled with SDS, and equal
protein aliquots from each condition were separated by gel
electrophoresis. Other aliquots were quantified by scintillation
counting; each condition was prepared in duplicate, and the data from
three to four separate experiments were combined.
IP3 and Adenylate Cyclase Activity
The generation of IP3 was assayed as follows.
ER-expressing CHO cells were cultured in 100-mm petri dishes followed
by synchronization in the absence of serum for 24 h. The cells
were washed in PBS (pH 7.4) and incubated in DMEM-F12 (with 20
mM HEPES, 0.1% BSA, and 10 mM LiCl) at 37 C
for 10 min. 17-ß-E2, 10 nM, was added to the
cells for 15 sec, with or without ICI 182,780 or H222 (for CHO-ER),
which were added 10 min before 17-ß-E2. The reaction was
stopped by adding cold 5% perchloric acid and kept on ice for 20 min.
The media were centrifuged at 15,000 x g for 1 min,
and the supernatant was neutralized with 60 mM HEPES/1.5
M KOH for 60 min on ice. Insoluble KClO4 was
removed by centrifugation at 12,000 x g for 15 min at
4 C. The IP3 content in the supernatant was measured by kit
(New England Nuclear, Boston, MA).
Adenylate cyclase activity in the membrane was determined as follows.
CHO-ER or ERß membranes were prepared (32), and cAMP generation as
a function of cyclase activity was measured. For the cAMP assay, equal
amounts of cell membrane suspension (containing the adenylate cyclase)
and 17-ß-E2, 10 nM, with or without ER
antagonists, were added to assay buffer to a final total volume of 200
µl. The buffer contained 100 mM KCl, 20 mM
MgCl, 4 mM isobutylmethylxanthine, 0.8
mM EDTA, 1 mM GTP, 1 mM ATP, 20
mM phosphocreatine, and 1 mg/ml creatine phosphokinase in
0.2 M Tris-HCl, pH 7.5. After vortexing, the mixture was
incubated in a 37 C shaking water bath for 20 min, and then the enzyme
activity was stopped by boiling for 3 min at 95 C. The mixtures were
centrifuged at 6500 rpm for 10 min, and from each sample, 100 µl of
the supernatant were diluted with acetate buffer, pH 6.2, for the RIA.
The inter- and intraassay coefficients of variation of the RIA were
always less than 10%.
|
FOOTNOTES |
|---|
This work was supported by a Merit Review Grant from the Veterans Administration, NIH Grant NS-30521 (E.R.L.), National Cancer Institute Grant CA-02897, and United States Army Research Medical Corp. Grant DAMD-17–04-J4228 (G.L.G.).
Received for publication September 11, 1998. Revision received October 30, 1998. Accepted for publication November 5, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
,25-dihydroxyvitamin D3 in basal-lateral membranes of chick
intestinal epithelium and relationship to transcaltachia. J Biol
Chem 269:23750–23756 and ß) expression in
the excurrent ducts of the adult male rat reproductive tract. J
Androl 18:602–611 and ERß at AP1 sites. Science 277:1508–1510This article has been cited by other articles:
 |
|
 |
|
  J. Teng, Z.-Y. Wang, E. R. Prossnitz, and D. E. Bjorling The G Protein-Coupled Receptor GPR30 Inhibits Human Urothelial Cell Proliferation Endocrinology, August 1, 2008; 149(8): 4024 - 4034. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Titolo, C. M. Mayer, S. S. Dhillon, F. Cai, and D. D. Belsham Estrogen Facilitates both Phosphatidylinositol 3-Kinase/Akt and ERK1/2 Mitogen-Activated Protein Kinase Membrane Signaling Required for Long-Term Neuropeptide Y Transcriptional Regulation in Clonal, Immortalized Neurons J. Neurosci., June 18, 2008; 28(25): 6473 - 6482. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kurokawa, M. Tamagawa, N. Harada, S.-i. Honda, C.-X. Bai, H. Nakaya, and T. Furukawa Acute effects of oestrogen on the guinea pig and human IKr channels and drug-induced prolongation of cardiac repolarization J. Physiol., June 15, 2008; 586(12): 2961 - 2973. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
