| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Research Service, Stratton Veterans Affairs Medical Center, the Ordway Research Institute, Inc. and the Wadsworth Center, New York State Department of Health, Albany, New York 12208
Address all correspondence and requests for reprints to: Faith B. Davis, M.D., Ordway Research Institute, Inc., 150 New Scotland Avenue, Albany, New York 12208. E-mail: fdavis{at}ordwayresearch.org.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, we studied whether thyroid hormone also promoted MAPK-mediated ER phosphorylation. Human breast cancer (MCF-7) cells were incubated with physiological concentrations of L-T4 or 17ß-estradiol (E2) for 15 min to 24 h, and nuclear ER and serine-118-phosphorylated ER were identified by Western blotting. Serine-118-phosphorylated ER was recovered at 15 min in nuclei of MCF-7 cells exposed to either T4 or E2. The T4 effect was apparent at 15 min and peaked at 2 h, whereas the E2 effect was maximal at 4–6 h. T4-agarose was as effective as T4 in causing phosphorylation of ER. T4 action on ER was inhibited by PD 98059, an inhibitor of ERK1/2 phosphorylation, and by tetraiodothyroacetic acid, a T4 analog that blocks cell surface-initiated actions of T4 but is not itself an agonist. Electrophoretic mobility shift assay of nuclear extracts from T4-treated and E2-treated cells showed similar specific protein-DNA-binding. Indexed by [3H]thymidine incorporation and nuclear proliferating cell nuclear antigen, MCF-7 cell proliferation was stimulated by T4 and T4-agarose to an extent comparable with the effect of E2. This T4 effect was blocked by either PD 98059 or ICI 182,780, an ER antagonist. Thus, T4, like E2, causes phosphorylation by MAPK of nuclear ER at serine-118 in MCF-7 cells and promotes cell proliferation through the ER by a MAPK-dependent pathway. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(1, 4, 5), the oncogene suppressor protein p53 (3), and the nuclear thyroid hormone receptor (TR) for T3 (TRß1) (2). The transcriptional activity of each of these proteins is changed by T4-directed, MAPK-mediated phosphorylation (1, 2, 3, 4, 5). It is also now apparent that thyroid hormone-activated MAPK may phosphorylate plasma membrane proteins. For example, T3 will activate the plasma membrane Na+/H+ antiporter by a nongenomic mechanism, which is MAPK dependent (6). Because ambient concentrations of thyroid hormone in the intact organism are relatively stable, these effects of the hormone mediated by signal transduction are thought by us to represent modulators of set points or "basal" activities of nuclear or plasma membrane proteins (7).
17ß-Estradiol (E2) has also been shown by several laboratories to cause serine phosphorylation of its nuclear receptor, estrogen receptor (ER) (8, 9, 10, 11), and phosphorylation of serine-118 of ER may result in a change in transcriptional activity of the receptor (10, 12). Serine phosphorylation caused by E2 has been suggested by some investigators (12, 13), but not all (14, 15), to be MAPK dependent, although there is variability in cell type, study design, and E2 concentration among these reports. Thyroid hormone has been shown to have estrogen-like effects in human breast cancer cells (16). In the present studies, we examined the possibility that thyroid hormone might mimic the action of E2 on MAPK-dependent serine phosphorylation of ER and on target cell proliferation.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and anti-ERK2, and monoclonal anti--actin and antiproliferating cell nuclear antigen (anti-PCNA) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA); polyclonal antiactivated MAPK (phospho-MAPK, pERK1/2) and monoclonal anti-pER (serine-118) (Cell Signaling, Beverly, MA). Antiphosphoserine was obtained from R & D Systems (Minneapolis, MN). Secondary antibodies (Dako, Carpinteria, CA) were either goat antirabbit (1:1000), rabbit antimouse (1:1000), or rabbit antigoat IgG (1:2000), depending on the origin of the primary antibody. [6-3H]thymidine was obtained from Amersham Biosciences (Piscataway, NJ).
Cell culture; preparation of nuclear fractions
Cells were cultured and nuclear fractions prepared as we have previously described (1, 2, 3). MCF-7 cells were grown in DMEM without phenol red. Cells were grown to confluence in DMEM supplemented with 5% fetal bovine serum, and then exposed for 72 h to DMEM without serum. Preparation of stock T4, agarose-T4, and tetrac solutions was as previously reported (2). E2 stock solutions were prepared in 100% ethanol and diluted 1:1000 for cell treatment. Hormones or analogs were then added to DMEM to achieve the final concentrations indicated. After treatment, the cells were harvested and nucleoproteins prepared as previously described (1, 2, 3).
Immunoblotting; immunoprecipitation
The techniques have been described in earlier publications from our laboratory (1, 2, 3, 4, 5). After treatment of cells with thyroid hormone analogs or E2, the cells were washed with PBS and lysed in hypotonic buffer. The resulting crude nuclear extracts were exposed to high-salt buffer for preparation of nuclear fractions. Equal aliquots of nucleoproteins (25 µg) from each sample were separated on discontinuous SDS-PAGE (10% resolving gels) and then transferred by electroblotting to Immobilon (Millipore, Bedford, MA) membranes. After blocking with 5% milk in Tris-buffered saline containing 0.1% Tween, the membranes were incubated at room temperature with selected antibodies for 2 h or overnight. Immunoreactive proteins were detected by chemiluminescence (ECL, Amersham Life Science, Arlington Heights, IL) and integrated OD (IOD) of bands were scanned and quantitated as previously described (17). For immunoprecipitation, 200 µg protein from each nuclear sample was used, followed by separation of the immunoprecipitated proteins by PAGE and immunoblotting as previously described (1, 2, 3, 4, 5). Results shown are representative of three or more similar experiments.
EMSA
EMSA of nuclear extracts was carried out by a modification of our previously published method (17). In brief, nuclear extracts (10 µg protein) were incubated in a 25-µl total reaction vol that contained 10 mM Tris (pH 7.5), 50 mM NaCl, 1.0 mM MgCl2, 0.5 mM EDTA, 0.5 mM dithiothreitol, 4% glycerol, and 0.05 µg/µl poly(dI-dC) (Promega, Madison, WI). [
-32P]ATP-labeled oligonucleotide was added to the total reaction mixture, and the latter was incubated for 20 min at room temperature. Samples were loaded onto 4% polyacrylamide gels in low-ionic-strength buffer (22.3 mM Tris, 22.2 mM borate, 0.5 mM EDTA) and run at 15 V/cm with cooling. Gels were then dried and exposed to x-ray film, and the radioautograph was analyzed. The estrogen response element (ERE) consensus binding oligonucleotide was obtained from Santa Cruz Biotechnology, Inc. (5'-GGA TCT AGG TCA CTG TGA CCC CGG ATC-3'), and an SP-1 consensus oligonucleotide was purchased from Promega.
[3H]thymidine uptake
The technique was modified from that of Cao et al. (18). MCF-7 cells were seeded in 24-well plastic tissue culture plates (6 x 104 cells/well). Cells were incubated for 48 h in medium containing 10% hormone-stripped fetal bovine serum (19). [3H]thymidine was then added as 1 µCi/ml, and the cells were treated with T4 (10–7 M) or E2 (10–10 M) for 1, 2, 4, 6, or 24 h. Control samples contained no added thyroid hormone or E2. Monolayers were washed three times with PBS, followed by treatment with cold 5% trichloroacetic acid for 30 min at 4 C to lyse the cells and precipitate cellular DNA. After washing the precipitate twice with cold absolute ethanol, the DNA was solubilized in 2% sodium dodecyl sulfate. Radioactivity in the solubilized DNA was then counted.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, upper panel, (pERK1/2) and summarized in the graph below. Phosphorylated MAPK in nuclear fractions was seen as early as 15 min after exposure of cells to thyroid hormone; the maximum response was obtained at 2 h, and the signal declined thereafter. Prior studies with T4-A have indicated that this hormone analog causes activation of the MAPK-dependent signal transduction cascade in a manner similar to that of T4 (1, 2). Experiments in MCF-7 cells were performed with T4-A (10–7 M), and results in Fig. 1B demonstrate nuclear accumulation of pERK1/2 with this T4 analog, comparable to that with T4.
|
), as others have noted (12, 20). The pERK1/2 nuclear accumulation induced by E2, like that caused by T4, appeared at 15 min, but persisted for 24 h.
Serine phosphorylation of ER by T4 and E2
T4 caused an increase in nuclear abundance of serine-118-phosphorylated ER in MCF-7 cells that was apparent at 15 min of hormone exposure and was maximal by 2 h (Fig. 1A). Similar effects on serine-118 phosphorylation of ER were seen with T4-A (Fig. 1B).
E2 treatment of MCF-7 cells also resulted in nuclear accumulation of serine-118-phosphorylated ER (Fig. 1C). The increase in phosphorylated nuclear ER also appeared by 15 min and was maximal at 4–6 h. The antibody used to detect phosphorylated ER in these experiments was directed against a phosphoserine at residue 118 of ER. Serine-118 phosphorylation of ER appears to play a significant role in DNA-binding (8) and transcriptional activation by the receptor (9, 10).
Effect of inhibition of the MAPK signal transduction pathway on serine phosphorylation of nuclear ER, and on nuclear complex formation between serine-phosphorylated ER and MAPK induced by T4 and E2
We have shown previously in several cell lines that activation of MAPK by T4 is inhibited by the MAPK kinase-associated inhibitor, PD (1, 2, 3, 4, 5). We therefore asked whether serine phosphorylation of ER and activation of MAPK by T4 and E2 were subject to inhibition by PD in MCF-7 cells. As shown in the upper immunoblot of Fig. 2A, PD (30 µM) inhibited phosphorylation of ERKs 1 and 2 (MAPK) by both hormones. In association with this effect of PD, we also observed inhibition of serine-118 phosphorylation of ER by PD in the presence of either T4 or E2 (Fig. 2A, lower immunoblot). These experiments do not exclude the possibility that serines in addition to residue 118 were also subject to hormone-activated MAPK phosphorylation, but do indicate that serine-118 is at least one target of activated MAPK, as reported by others (11, 12).
|
, we immunoprecipitated nuclear proteins with either anti-ER or antiphosphoserine antibodies, separated the immunoprecipitated proteins by PAGE, and immunoblotted the proteins with antiphosphoserine or anti-ER, respectively. Similar results were obtained with both approaches (Fig. 2B) in that serine phosphorylation of ER was demonstrated in cells treated with either T4 or E2, and both effects were blocked by PD.
In prior studies (2, 3), we have shown that T4 causes nuclear translocation and coimmunoprecipitation, or complexing, of activated MAPK and the endogenous nuclear TRß1. In the present studies, nuclear extracts of hormone-treated cells were immunoprecipitated with anti-phosphoMAPK and the resulting proteins separated by PAGE and immunoblotted with anti-serine-118-phosphorylated ER. Shown in a representative study in Fig. 3, treatment of MCF-7 cells with either T4 or E2 led to nuclear coimmunoprecipitation of phosphorylated MAPK and serine-118-phosphorylated ER. This nuclear complex is evident with both T4 and E2 treatment (lanes 3 and 5, respectively). The cells were treated with T4 and E2 in the presence or absence of PD. With the resulting inhibition of MAPK activation, there was little evidence of hormone-induced coimmunoprecipitation of ER and activated MAPK (lanes 4 and 6). Additional nuclear samples from the cells in this study were immunoblotted with anti-pERK1/2 without immunoprecipitation. The lower immunoblot shows nuclear accumulation of pERK1/2 in cells treated with either hormone, and inhibition of that effect with PD.
|
(Fig. 4A), as seen above in Fig. 1, A and B. Protein A-agarose served as a control and did not activate MAPK or cause ER phosphorylation (results not shown).
|
phosphorylation (Fig. 4B). Tetrac, alone, did not cause serine phosphorylation of ER, because it is not an agonist, but is known to displace T4 from erythrocyte membranes (21). L-T3 is the natural ligand of the nuclear TR (22, 23). We compared T3 and T4 as MAPK activators and found that, at 10–7 M, they were both effective in 15 min (results not shown). When phosphorylation of ER was examined, however, 10–6 M T3 was only minimally effective, whereas 10–9 and 10–7 M T4 caused serine-118 phosphorylation. The addition of 1 mM 6-n-propyl-2-thiouracil to cells treated with T4 or T4-A did not alter the response to the hormone, indicating that T4 was effective without conversion to T3 by 5'-deiodinase activity (results not shown).
EMSA of nuclear extracts from cells exposed to T4 or to E2
Shown in Fig. 5, nuclear extracts obtained from T4-treated and E2-treated MCF-7 cells revealed similar binding of nucleoproteins to labeled ERE oligonucleotide. Specific protein-DNA interaction is shown as band b in lanes 2 and 4 with T4 and E2, respectively. Concurrent treatment of cells with PD and either hormone reduced specific protein-DNA interaction (band b, lanes 3 and 5). Addition of excess unlabeled ERE oligonucleotide reduced the signal at band b in nuclear extracts from E2-treated cells (lane 7), whereas excess unlabeled SP-1 (control) oligonucleotide had no effect on the distribution of oligonucleotide-binding in E2-treated cells (lane 6). Band a in the figure represents nonspecific protein-DNA interaction that is unaffected by addition of excess unlabeled SP-1 or ERE oligonucleotides.
|
are similar progressive increases in radiolabeled thymidine uptake by cells over 4 h with or without hormone treatment. In 6 and 24 h, there were significant increases in growth of cells treated with either T4 or E2. The effects of both hormones were reduced by ICI to the level of untreated control cells at 24 h.
|
, both E2 and T4 increased PCNA abundance in nuclear fractions of MCF-7 cells, and the hormone effects were diminished by PD treatment. These results indicate that both T4 and E2 increased cell proliferation by an ERK1/2-dependent mechanism. The effects of T4 and T4-A (10–7 M) were similar, and both were inhibited by PD (Fig. 7B). To test whether this effect on proliferation required intact ER function, we treated cells with ICI together with either T4 or E2, and measured nuclear PCNA (Fig. 5C). The ER inhibitor substantially blocked the effects of both hormones on cell proliferation.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, although these authors did not fully establish a mechanism by which MAPK was activated. Several other laboratories have confirmed that both ER and ERß are substrates for MAPK (13, 27). It is also known that E2 activates MAPK (12, 13, 28). In addition to these similar effects of the two hormones on MAPK activation, thyroid hormone is known to have estrogen-like actions in human breast cancer cells (16). The possibility that activation by thyroid hormone of MAPK might lead to serine phosphorylation of ER prompted the studies reported here. It is clear from these experiments, carried out in MCF-7 cells, that T4 indeed can promote MAPK-dependent serine phosphorylation of serine-118 of ER. We also confirm that E2 activates MAPK (ERK1/2) in MCF-7 cells. The time courses of stimulation of phosphorylation of ER by T4 and by E2 in MCF-7 cells in our studies were similar in onset, although the E2 effect was more sustained.
In these experiments involving thyroid hormone, it was necessary to establish that there were downstream consequence(s) of serine-118 phosphorylation of ER caused by T4. The experimental approach we used was to index cell proliferation by two methods (18, 29, 30) and to determine the contribution of T4- and E2-activated ER to cell proliferation. To establish that ER was involved, we added ICI to hormone-treated cells and found that T4-directed increases in nuclear abundance of PCNA and [3H]thymidine uptake by cells were reduced by ICI. This result indicates that the effect of T4 on cell proliferation is ER dependent. Consistent with the results of these studies of cell proliferation, we also found, on gel shift assays of nuclear extracts, that thyroid hormone treatment of cells resulted in increased specific protein-binding of an ERE oligonucleotide. The T4 effect was similar to the effect of E2 in this assay system. The specific protein-DNA interactions in nuclear extracts from both E2- and T4-treated cells were inhibited when the MAPK cascade was interrupted by PD.
Although we have established elsewhere that activated MAPK directly phosphorylates TRß1 in vitro (2, 31), the present studies indicate that serine phosphorylation of ER is MAPK dependent, and suggest that ERKs 1 and 2 directly phosphorylate ER. As indicated above, others have demonstrated serine-118 phosphorylation of ER in vitro by activated MAPK (12). We have recently identified the docking site for activated MAPK on TR: residues 128–133, KGFFRR, in the DNA-binding domain (31). Similar basic amino acid-enriched motifs exist in other members of the nuclear superfamily of hormone receptors, including ER (residues 206–211, KAFFKR) and ERß, as well as the glucocorticoid and progesterone receptors (Table 1). These structural similarities suggest that most or all of these receptors are potential MAPK substrates and raise the possibility, as yet untested, that the function of other receptor superfamily members may be affected by thyroid hormone.
|
phosphorylation that we describe here contributes to a set-point for transcriptional activity of the receptor-E2 complex. We have proposed elsewhere (7) that thyroid hormone is a contributor to basal activity of certain plasma membrane transport proteins, such as the Na+/H+ exchanger (32). It is also possible that, in the context of constant circulating levels of endogenous estrogen, the clinical states of hyperthyroidism or hypothyroidism might alter the impact of E2 on certain target organs. Thus, in the absence of estrogen or in the setting of a constant ambient concentration of estrogen, clinically important changes in thyroid hormone levels (hypothyroidism or hyperthyroidism) could result in functional alterations that are estrogen-like (33) (hypoestrogenism or hyperestrogenism, respectively) in their effects.
In contrast to the present description of apparently overlapping MAPK-dependent effects of thyroid hormone and estrogen at the cell surface, direct interaction of the nuclear receptors for thyroid hormone and estrogen has been reported by others (34, 35). For example, Dinda et al. (16) have shown that thyroid hormone can activate MAPK by a genomic mechanism in which the hormone (T3)-TR complex binds to an ERE and activates E2-dependent transcription of factors such as EGF that, on an autocrine basis, then activate MAPK. In T4-treated cells, we can recover complexes of TR, ER, and activated MAPK from nuclear fractions (H.-Y.L., F.B.D., P.J.D.; unpublished observations). In the model of thyroid hormone action reported here, an action that appears to be initiated at the cell surface, it is possible that, downstream of MAPK activation, the ER-dependent cell proliferation response that we describe may depend on formation of heterodimeric complexes of ER with TR.
|
Footnotes |
|---|
Abbreviations: E2, 17ß-Estradiol; ER, estrogen receptor; ERE, estrogen response element; ICI, ICI 182,780; PCNA, proliferating cell nuclear antigen; PD, PD 98059; T4-A, T4-agarose; tetrac, tetraiodothyroacetic acid; TR, thyroid hormone receptor.
Received March 9, 2004.
Accepted for publication March 23, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-induced HLA-DR expression. J Immunol 161:843–849 are mediated by 3',5'-cyclic-adenosine monophosphate-dependent protein kinase II. Endocrinology 145:1708–1717This article has been cited by other articles:
 |
|
 |
|
  H.-Y. Lin, H.-Y. Tang, T. Keating, Y.-H. Wu, A. Shih, D. Hammond, M. Sun, A. Hercbergs, F. B. Davis, and P. J. Davis Resveratrol is pro-apoptotic and thyroid hormone is anti-apoptotic in glioma cells: both actions are integrin and ERK mediated Carcinogenesis, January 1, 2008; 29(1): 62 - 69. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. R. Hammes and E. R. Levin Extranuclear Steroid Receptors: Nature and Actions Endocr. Rev., December 1, 2007; 28(7): 726 - 741. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Verga Falzacappa, E. Petrucci, V. Patriarca, S. Michienzi, A. Stigliano, E. Brunetti, V. Toscano, and S. Misiti Thyroid hormone receptor TR{beta}1 mediates Akt activation by T3 in pancreatic {beta} cells J. Mol. Endocrinol., February 1, 2007; 38(2): 221 - 233. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Germain, B. Staels, C. Dacquet, M. Spedding, and V. Laudet Overview of Nomenclature of Nuclear Receptors Pharmacol. Rev., December 1, 2006; 58(4): 685 - 704. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-Y. Lin, L. Lansing, J.-M. Merillon, F. B. Davis, H.-Y. Tang, A. Shih, X. Vitrac, S. Krisa, T. Keating, H. J. Cao, et al. Integrin {alpha}V{beta}3 contains a receptor site for resveratrol FASEB J, August 1, 2006; 20(10): 1742 - 1744. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. B. Davis, H.-Y. Tang, A. Shih, T. Keating, L. Lansing, A. Hercbergs, R. A. Fenstermaker, A. Mousa, S. A. Mousa, P. J. Davis, et al. Acting via a Cell Surface Receptor, Thyroid Hormone Is a Growth Factor for Glioma Cells. Cancer Res., July 15, 2006; 66(14): 7270 - 7275. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. W. Woodmansee, J. M. Kerr, E. A. Tucker, J. R. Mitchell, D. J. Haakinson, D. F. Gordon, E. C. Ridgway, and W. M. Wood The Proliferative Status of Thyrotropes Is Dependent on Modulation of Specific Cell Cycle Regulators by Thyroid Hormone Endocrinology, January 1, 2006; 147(1): 272 - 282. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Incerpi Thyroid Hormones: Rapid Reply by Surface Delivery Only Endocrinology, July 1, 2005; 146(7): 2861 - 2863. [Full Text] [PDF]
|
|
|
|
|
|
J. J. Bergh, H.-Y. Lin, L. Lansing, S. N. Mohamed, F. B. Davis, S. Mousa, and P. J. Davis Integrin {alpha}V{beta}3 Contains a Cell Surface Receptor Site for Thyroid Hormone that Is Linked to Activation of Mitogen-Activated Protein Kinase and Induction of Angiogenesis Endocrinology, July 1, 2005; 146(7): 2864 - 2871. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Zhao, H. Lorenc, H. Stephenson, Y. J. Wang, D. Witherspoon, B. Katzenellenbogen, D. Pfaff, and N. Vasudevan Thyroid hormone can increase estrogen-mediated transcription from a consensus estrogen response element in neuroblastoma cells PNAS, March 29, 2005; 102(13): 4890 - 4895. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. D'Arezzo, S. Incerpi, F. B. Davis, F. Acconcia, M. Marino, R. N. Farias, and P. J. Davis Rapid Nongenomic Effects of 3,5,3'-Triiodo-L-Thyronine on the Intracellular pH of L-6 Myoblasts Are Mediated by Intracellular Calcium Mobilization and Kinase Pathways Endocrinology, December 1, 2004; 145(12): 5694 - 5703. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
