This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Bodine, P. V. N. Articles by Komm, B. S. Search for Related Content PubMed PubMed Citation Articles by Bodine, P. V. N. Articles by Komm, B. S.

Suppression of Ligand-Dependent Estrogen Receptor Activity by Bone-Resorbing Cytokines in Human Osteoblasts

Women’s Health Research Institute, Wyeth-Ayerst Research, Inc., Radnor, Pennsylvania 19087

Address all correspondence and requests for reprints to: Dr. Peter V. N. Bodine, Women’s Health Research Institute, Wyeth-Ayerst Research, Inc., 145 King of Prussia Road, Radnor, Pennsylvania 19087. E-mail: bodinep{at}war.wyeth.com

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Estrogens are important for bone homeostasis and are classified as antiresorptive agents. One of the mechanisms for this effect is the inhibition of cytokine-induced bone resorption, which is mediated in part through an interaction between the estrogen receptor (ER) and nuclear factor (NF)-B in osteoblasts. We present evidence that bone-resorbing cytokines that activate NF-B conversely inhibit ligand-dependent ER activity in the conditionally immortalized human osteoblast cell line, HOB-03-CE6. Treatment of HOB-03-CE6 cells with 17ß-estradiol (17ß-E₂) up-regulated reporter gene activity [ERE-thymidine kinase (tk)-luciferase] 3- to 5-fold in a dose-dependent manner (EC₅₀ = 1.0 pM). However, cotreatment of the cells with 17ß-E₂ and increasing concentrations of either tumor necrosis factor- (TNF), interleukin-1 (IL-1), or IL-1ß completely suppressed ERE-tk-luciferase activity in a dose-dependent manner (IC₅₀ = 0.05–5.0 pM). On the other hand, treatment of the cells with growth factors either up-regulated or had no effect on ERE-tk-luciferase expression. Neither TNF, IL-1, nor IL-1ß treatment affected basal reporter gene activity in the cells, and the TNF effect was reversed by a neutralizing antibody to the cytokine. TNF treatment also suppressed ligand-dependent ER activity in MCF-7 human breast cancer cells, but not in Chinese hamster ovary cells that overexpressed human ER, even though both cell lines responded to the cytokine as measured by the up-regulation of NFB-tk-luciferase activity. TNF treatment did not affect the steady state levels of either ER or ERß messenger RNA expression by the HOB-03-CE6 cells, nor did it reduce [¹²⁵I]17ß-E₂ binding. Moreover, TNF treatment only weakly inhibited ligand-dependent glucocorticoid receptor activity in the HOB-03-CE6 cells. Bone-resorbing cytokines, which do not signal through the NF-B pathway, did not suppress ERE-tk-luciferase activity in HOB-03-CE6 cells. Treatment of the cells with 17ß-E₂ partially suppressed the activation of NF-B by TNF, but did not block cytokine-induced IL-6 secretion. Finally, cotreatment of HOB-03-CE6 cells with an antisense oligonucleotide to NF-B p50 partially reversed the suppression of ERE-tk-luciferase activity by TNF. In summary, these data provide evidence for a potent feedback inhibition of estrogen action in human osteoblasts that is at least partly mediated by the activation of NF-B.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
BONE remodeling is the process by which immature, damaged, or aged bone is replaced with new lamellar bone. This process involves the balanced coupling of osteoblastic bone formation with osteoclastic bone resorption (reviewed in Refs. 1, 2). One of the most important coupling factors is estrogen (reviewed in Ref. 3). The loss of estrogen at menopause creates an imbalance of osteoblastic and osteoclastic activities, such that bone resorption outpaces bone formation (1, 2, 3). If left unchecked, this situation leads to high turnover bone loss, postmenopausal or type I osteoporosis, and an increased risk of fractures (1, 2). In addition, estrogen deficiency may be a contributing factor to type II or senile osteoporosis, which affects both aging women and men (1).

Estrogens are considered to be the first line therapy for the treatment of postmenopausal osteoporosis (reviewed in Ref. 4). These steroid hormones are classified therapeutically as antiresorptive agents, because they inhibit osteoclast differentiation and activity (1, 2, 3). Although they have been observed to have direct suppressive effects on the cells of the osteoclast lineage (reviewed in Ref. 5), estrogens also have indirect inhibitory effects on bone resorption through cells of the osteoblast lineage (5, 6). One of the mechanisms by which this indirect suppression occurs is the inhibition of bone-resorbing cytokine action on the bone marrow stromal cells and osteoblasts (reviewed in Ref. 6). For example, 17ß-estradiol (17ß-E₂) has been reported to block the ability of tumor necrosis factor- (TNF) and interleukin-1ß (IL-1ß) to up-regulate IL-6 expression in several in vitro osteoblast models (7, 8, 9, 10). Other studies have shown that the estrogen receptor (ER) can interact directly with the mediators of TNF and IL-1ß action, namely nuclear factor (NF)-IL6 and NF-B (reviewed in Ref. 11), and that this interaction leads to a suppression of IL-6 up-regulation (8, 9, 12, 13). Conversely, in transfected U2-OS human osteosarcoma osteoblast-like cells overexpressing ER, cotransfection of NF-B (p65/RelA), or NF-IL6 (C/EPBß) expression vectors suppressed ER activity as measured by the trans-activation of an estrogen response element (ERE)-tk-luciferase reporter gene (9). These and other results indicated that the ER and NF-B/NF-IL6 were capable of mutually suppressing each other’s activity (9, 13). However, as this work involved overexpression of the ER and NF-B/NF-IL6 in an osteosarcoma cell line, it was not known whether the mutual suppression occurs in a more physiologically relevant in vitro setting, i.e. human osteoblasts that express both a normal phenotype and endogenous levels of functional receptors and transcription factors.

We recently reported the establishment of a conditionally immortalized mature osteoblast cell line from adult human bone that we termed HOB-03-CE6 (14). These cells were transformed with a temperature-sensitive simian virus 40 large T antigen (14) and, in contrast to osteosarcoma cells, are faithful to the proliferation/differentiation relationship at the nonpermissive temperature (39 C) when the T antigen mutant is inactivated (14, 15, 16). Moreover, unlike other estrogen-responsive osteoblast-like cell lines that overexpress ER (17, 18, 19, 20), this new cell line naturally expresses endogenous levels of functional receptors (14). The HOB-03-CE6 cells express both ER and ERß messenger RNAs (mRNAs) and contain approximately 1200 high affinity estrogen-binding sites/cell (both ER and ERß) (14). Like other osteoblast models (6), the HOB-03-CE6 cells also contain a functional NF-B/NF-IL6 pathway and respond potently to TNF, IL-1, and IL-1ß with a 10- to 100-fold increase in IL-6 secretion. Consequently, HOB-03-CE6 cells present us with the opportunity to study the regulation of endogenous ER activity in an important human target cell for estrogen action.

While screening a variety of growth factors, cytokines, and peptide hormones for the ability to regulate endogenous ER activity in HOB-03-CE6 cells, we observed that bone-forming growth factors such as transforming-growth factor-ß1 (TGFß1) up-regulated ERE-tk-luciferase expression in the absence of estrogen. In contrast, the bone-resorbing cytokines TNF, IL-1, and IL-1ß were potent suppressors of steroid-dependent ER function. As the latter observations had implications for the interaction between estrogens and cytokines during postmenopausal osteoporosis, we have pursued the mechanism of this inhibition. In this report, we describe the cytokine-mediated suppression of ligand-dependent ER activity in HOB-03-CE6 cells and present evidence that it involves activation of the transcription factor NF-B.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Materials
Except where noted, tissue culture and molecular biology reagents were purchased from Life Technologies (Grand Island, NY); other reagents and chemicals were purchased from either Sigma Chemical Co. (St. Louis, MO) or VWR (Philadelphia, PA). Recombinant human (rh) or purified human growth factors and cytokines were purchased from either R & D Systems (Minneapolis, MN) or Genzyme (Cambridge, MA). Neutralizing monoclonal antibodies to rhTNF and rhTNFß as well as enzyme-linked immunosorbent assay kits for human IL-6 were purchased from R & D Systems. The specific antiestrogen ICI-182,780 was obtained from Zeneca Pharmaceuticals (Wilmington, DE). [-³²P]ATP (3000 Ci/mmol) and [¹²⁵I]17ß-E₂ (2200 Ci/mmol) were purchased from New England Nuclear Corp. (Boston, MA).

The following cell culture media were used in these studies: HOB-03-CE6 growth medium: phenol red-free DMEM/Ham’s F-12 containing 10% (vol/vol) heat-inactivated FBS, 1% (vol/vol) penicillin-streptomycin, and 2 mM GlutaMAX-1; MCF-7 growth medium: DMEM/Ham’s F-12 containing 10% (vol/vol) heat-inactivated FBS, 1% (vol/vol) penicillin-streptomycin, and 2 mM GlutaMAX-1; CHO-hER growth medium: phenol red-free MEM containing 10% (vol/vol) charcoal-stripped FBS (HyClone Laboratories, Inc., Logan, UT), 1% (vol/vol) penicillin-streptomycin, 2 mM GlutaMAX-1, and 1 mg/ml geneticin; HOB-03-CE6 experimental medium: phenol red-free DMEM/Ham’s F-12 containing 2% (vol/vol) heat-inactivated charcoal-stripped FBS, 1% (vol/vol) penicillin-streptomycin, 2 mM GlutaMAX-1, 50 µM ascorbate-2-phosphate (Wako Pure Chemical Industries Ltd, Richmond, VA), and 10 nM menadione sodium bisulfite (vitamin K₃); and MCF-7 experimental medium: phenol red-free DMEM/Ham’s F-12 containing 10% (vol/vol) heat-inactivated charcoal-stripped FBS, 1% (vol/vol) penicillin-streptomycin, and 2 mM GlutaMAX-1.

Initiation site-directed antisense, sense, and scrambled phosphorothioate oligonucleotides to the human NF-B subunits p65/RelA, p50, and c-Rel were synthesized on an Applied Biosystems 392/394 DNA synthesizer (Foster City, CA) using phosphoramidite chemistry as previously described (21).

Cell culture
Conditionally immortalized HOB-03-CE6 cells were cultured at 34 C (the permissive temperature for cell division) in HOB-03-CE6 growth medium as previously described (14). MCF-7 human breast carcinoma cells were purchased from American Type Culture Collection (Rockville, MD) and were cultured in MCF-7 growth medium at 37 C. Chinese hamster ovary (CHO)-hER cells, which overexpress hER, were cultured in CHO-hER growth medium at 37 C. These cells were stably transfected with a cytomegalovirus-hER expression vector that also encoded for the neomycin resistance gene (Bhat, R. A., and B. S. Komm, unpublished results).

Measurements of ER, glucocorticoid receptor (GR), and NF-B activities
ER activity was measured by the trans-activation of an adenoviral (Ad5)-ERE-tk-luciferase reporter gene construct as previously described (14). Similarly, GR activity was measured by the trans-activation of an Ad5-glucocorticoid response element (GRE)-luciferase reporter gene construct, whereas NF-B activity was measured by the trans-activation of an Ad5-NFB response element-tk-luciferase reporter gene construct. HOB-03-CE6 cells were infected with approximately 200 plaque-forming units (PFU)/cell of virus, whereas the MCF-7 and CHO-hER cells were infected with about 100 and 50 PFU/cell, respectively. Under these conditions, essentially 100% of the cells were infected with virus and expressed the reporter genes. The Ad5-GRE-luciferase virus contained two copies of the GRE from the tyrosine aminotransferase promoter (5'-TGTACAGGATGTTCT-3') linked to the luciferase gene (22). The Ad5-NFB response element-tk-luciferase virus contained three copies of the NF-B binding site from the major histocompatibity complex class I promoter (5'-AGATCTGGGGAATCCCC-3') linked to the tk promoter (-105 to +10) and the luciferase gene (23). Both of these constructs were inserted in place of the E1a adenoviral gene. The luciferase assay was performed as previously described (14).

RT-PCR for ER and ERß
RNA isolation and quantitative or semiquantitative RT-PCR analysis of human ER and ERß mRNA were performed as previously described (14, 24). Quantitative RT-PCR for ER was performed as follows. A standard RNA competitor was synthesized that shared primers with wild-type ER but was 86 bp shorter. The primer used in the RT reaction was 5'-CAGCATGTCGAAGATC-3', whereas the forward primer was 5'-GGAGACATGAGAGCTGCCAACC-3'. These primers yielded 350- and 436-bp products of standard and wild-type ER, respectively. The RT reaction was performed for 45 min at 42 C in 50 mM Tris-HCl, pH 8.3, containing 75 mM KCl, 3 mM MgCl₂, 1 mM dithiothreitol, 1 mM deoxyribonucleotide triphosphates (dNTPs), 30 pmol reverse primer, and 200 U Superscript II reverse transcriptase. After heat inactivation, 35 cycles of PCR were performed in 26 mM Tris-HCl, pH 8.4, containing 55 mM KCl, 1.35 mM MgCl₂, 0.2 mM dithiothreitol, 0.2 mM dNTPs, 30 pmol forward primer, and 2.5 U Taq DNA polymerase using an annealing temperature of 64 C. PCR products were separated by HPLC on a Ranin Dynamax System (Woburn, MA) using a DNASep column (Sarasep, Santa Clara, CA). The products were eluted at 50 C with 0.1 M triethylammonium acetate (Fluka Chemical Co., Ronkonkoma, NY) and 25% (vol/vol) acetonitrile and were quantified by absorbance at 254 nm.

Semiquantitative RT-PCR for ERß was performed as follows. RT reactions were performed at 42 C with 0.5 µg RNA in 1 x PCR buffer containing 5 mM MgCl₂, 2.5 µM ERß-specific reverse primer (5'-GCAGAAGTG-AGCATCCCTCTTTG-3'), 0.5 µM glyceraldehyde-3-phosphate dehydrogenase (GAPDH)-specific reverse primer (5'-CACCCT-GTTGCTGTAGCCATATTC-3'), 1 mM dNTPs, 20 U RNasin (Promega Corp., Madison, WI), and 200 U Superscript II reverse transcriptase. After heat inactivation, PCR was initiated by adding a master mix containing ERß-specific forward primer (5'-CAGCATTCCCAGCAATGTCAC-3') and GAPDH-specific forward primer (5'-GACATCAAGAAGGTGGTGAAGCAG-3') directly to the RT reaction. The final concentration of reagents in the PCR reaction was as follows: 0.5 mM each ERß-specific primer, 0.1 mM of each GAPDH-specific primer, 1 x PCR buffer, 0.2 mM dNTPs, 1.5 mM MgCl₂, and 0.5 U Taq DNA polymerase. Two-step PCR was carried out in a Stratagene Robocycler (La Jolla, CA) for 25 cycles using an annealing temperature of 64 C. Samples were fractionated on 1.5% (wt/vol) agarose gels and transferred to Hybond-N⁺ (Amersham, Arlington Heights, IL) by alkali Southern blotting in 0.4 N NaOH and 0.6 M NaCl. Blots were prehybridized at 42 C in Rapid-Hyb buffer (Amersham). Internal oligonucleotide probes specific for the ERß (5'-CCGCATACAGATGTGATAACTGGCGATGGA-3') and GAPDH (5'-GCTGTTGAA-GTCACAGGAGACAACCTGGT-3') complementary DNA fragments were end labeled with [-³²P]ATP using polynucleotide kinase. Probes were added to the prehybridization solution at 3.0 x 10⁶ cpm/ml and incubated at 42 C. Blots were washed twice in 2 x NaCl-sodium citrate (SSC) and 0.1% (wt/vol) SDS at room temperature and then twice in 0.2 x SSC and 0.1% SDS at 42 C. Blots were exposed to x-ray film and were quantified using a Molecular Dynamics, Inc. PhosphorImager SI (Sunnyvale, CA). The primers for ERß amplified a 280-bp region from nucleotides 27–306 of the message (25).

Steroid binding assay
The [¹²⁵I]17ß-E₂ binding assays was performed as previously described (14). Dishes of cells were rinsed twice with PBS and once with lysis buffer (10 mM Tris-HCl, pH 7.5, containing 1 mM EDTA). The cells were then scraped from the dishes and cooled on ice, and a lysate was prepared using a hand-held homogenizer (model 395, Dremel Moto-Tool, Racine, WI). The lysate was centrifuged at 100,000 x g for 60 min, and aliquots of cytosol were mixed with 0.21 nM [¹²⁵I]17ß-E₂ in the presence or absence of a 1000-fold molar excess diethylstilbesterol. After incubation for 2 h at room temperature, cold 5% (wt/vol) dextran-coated charcoal was added (final concentration, 2.1%, wt/vol), and the mixture was centrifuged at 850 x g. Aliquots of the supernatant were removed, placed into a scintillation vial, and counted in a Beckman Coulter, Inc. LS 6500 (Fullerton, CA) using Ready Protein Plus cocktail (Beckman Coulter, Inc.). The protein concentration of the cytosol was estimated as previously described (14).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Regulation of endogenous ER activity in HOB-03-CE6 cells
In addition to being regulated by a natural ligand such as 17ß-E₂ or a variety of other estrogenic and antiestrogenic compounds, the ER is also regulated in a ligand-independent manner by growth factors or agents that elevate intracellular cAMP levels (reviewed in Ref. 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36). The mechanism for this ligand-independent regulation of the ER is thought to result from the phosphorylation of the receptor by various protein kinases (26, 35, 37, 38), and this activation is typically blocked by cotreatment with antiestrogens (26, 27, 29, 30, 33, 34, 35, 37, 38). As these studies were performed with primary rodent uterine cells (28, 29, 31, 33, 34, 35) or human breast cancer cells (27, 30, 32, 37), we were interested in learning whether the endogenous ER in a human osteoblast was also regulated in a similar manner. Therefore, HOB-03-CE6 cells were treated with insulin-like growth factor I (IGF-I), epidermal growth factor (EGF), platelet-derived growth factor (PDGF), TGFß1, IL-1ß, or TNF in the absence or presence of 1.0 nM 17ß-E₂ and/or 100 nM ICI-182,780 (a specific antiestrogen). At the end of the 48-h treatments, ER activity was measured by the trans-activation of Ad5-ERE-tk-luciferase, which is a simple, sensitive, and selective functional assay for the receptor. A 48-h treatment is optimal for the up-regulation of reporter gene expression by estrogens in this cell line (14).

As shown in Fig. 1, treatment of HOB-03-CE6 cells with the growth factors IGF-I, EGF, and PDGF had no effect on ERE-tk-luciferase activity in the absence or presence of 17ß-E₂. This was in contrast to results with rodent uterine and MCF-7 breast cancer cells, where IGF-I and EGF have been reported to up-regulate ER activity in the absence of estrogen (26, 34, 35, 38). On the other hand, treatment of HOB-03-CE6 cells with the bone-forming agent TGFß1 up-regulated ER activity in a ligand-independent manner (Fig. 1); curiously, however, these effects were incompletely blocked by cotreatment with a 100-fold molar excess of ICI-182,780.

View larger version (33K): [in this window] [in a new window]   Figure 1. Regulation of endogenous ER activity by peptide hormones, growth factors, and cytokines in HOB-03-CE6 cells. HOB-03-CE6 cells were seeded with HOB-03-CE6 growth medium at 17,000 cells/well into 96-well plates and incubated overnight at 34 C. The cells were then infected with Ad5-ERE-tk-luciferase, rinsed with PBS, and incubated for 48 h at 39 C (the nonpermissive temperature when cell division halts) with HOB-03-CE6 experimental medium containing 1) vehicle (0.1% ethanol; No Treatment), 2) 1.0 nM 17ß-E2, 3) 100 nM ICI-182,780 (a specific antiestrogen), or 4) 1.0 nM 17ß-E2 and ICI. The cells were also treated in the absence (Control) or presence of 100 nM IGF-I, 10 nM EGF, 10 nM PDGF, 0.08 nM TGFß1, 0.83 nM IL-1ß, or 8.3 nM TNF. After treatment, the cells were assayed for luciferase activity as previously described (14 ). In these experiments, treatment of the cells with 1.0 nM 17ß-E2 up-regulated luciferase expression 3- to 4-fold. The results are presented as the mean ± SEM (n = 8). *, P < 0.05, as determined by Dunnett’s ANOVA compared with the corresponding control values.

TNF suppresses ligand-dependent ER activity in HOB-03-CE6 cells
TNF is a very potent and efficacious suppressor of ligand-dependent ER activity in the HOB-03-CE6 cells. As shown in Fig. 2A, treatment of HOB-03-CE6 cells for 48 h with 1.0 nM 17ß-E₂ up-regulated Ad5-ERE-tk-luciferase activity 5- to 6-fold. However, when the cells were cotreated with estrogen and TNF, the cytokine completely suppressed ERE-tk-luciferase activity in a dose-dependent manner with an IC₅₀ of 3.8 pM. In contrast, cotreatment of the cells with vehicle (0.1% ethanol) and TNF did not affect basal luciferase activity. This result indicated that cytokine treatment did not simply inhibit basal promoter function or enzyme activity, but, instead, suppressed ligand-dependent receptor action. In support of these data, cotreatment of the HOB-03-CE6 cells with 100 pM TNF and increasing concentrations of 17ß-E₂ caused a complete suppression of estrogen-dependent ERE-tk-luciferase activity at each dose of steroid (Fig. 2B).

View larger version (20K): [in this window] [in a new window]   Figure 2. Suppression of ligand-dependent ER activity by TNF in HOB-03-CE6 cells. HOB-03-CE6 cells were seeded with HOB-03-CE6 growth medium at 17,000 cells/well into 96-well plates and incubated overnight at 34 C. The cells were then infected with Ad5-ERE-tk-luciferase, rinsed with PBS, and incubated for 48 h at 39 C with HOB-03-CE6 experimental medium containing either vehicle (0.1% ethanol) or 1.0 nM 17ß-E2 in the absence or presence of increasing concentrations of TNF (A). Alternatively, the cells were treated with either vehicle or increasing concentrations of 17ß-E2 in the absence or presence of 100 pM TNF (B). After treatment, the cells were assayed for luciferase activity as previously described (14 ). The results are presented as the mean ± SEM (n = 8). *, P < 0.05, as determined by Dunnett’s ANOVA, compared with the corresponding no treatment control.

TNF also suppresses ligand-dependent ER activity in MCF-7 cells, but not in CHO-hER cells
To determine the cellular selectivity for the suppression of ligand-dependent ER activity by TNF, we examined the effects of this cytokine on receptor function in two additional cell lines: MCF-7 human breast cancer cells, which express high endogenous levels of ER (50,000–150,000 receptors/cell) (39), and CHO-hER cells, which overexpress hER at levels comparable to those in MCF-7 cells (100,000 ERs/cell) (Bhat, R. A., H. A. Harris, and B. S. Komm, unpublished results). As depicted in Fig. 3A, treatment of MCF-7 cells for 48 h with a broad concentration range of TNF resulted in a dose-dependent suppression of estrogen-dependent Ad5-ERE-tk-luciferase activity, whereas basal luciferase expression was not altered. The IC₅₀ for this suppression was 8.3 pM, which was similar to the potency of this cytokine for the inhibition of ligand-dependent ER activity in HOB-03-CE6 cells. In contrast, treatment of CHO-hER cells with TNF had no effect on estrogen-activated ERE-tk-luciferase expression (Fig. 3B). These results underscore the observation that TNF treatment did not affect basal luciferase activity and indicated that this response was, to the extent tested, cell type selective.

View larger version (18K): [in this window] [in a new window]   Figure 3. TNF suppresses ligand-dependent ER activity in MCF-7 cells, but not in CHO-hER cells. MCF-7 human breast cancer cells (which express high endogenous levels of ER) were seeded with MCF-7 growth medium at 25,000 cells/well into 96-well plates (A), whereas CHO-hER cells (which stably overexpress hER) were seeded with CHO-hER growth medium at 50,000 cells/well into 96-well plates (B); both cell lines were incubated overnight at 37 C. The cells were then infected with Ad5-ERE-tk-luciferase, rinsed with PBS, and incubated for 48 h at 37 C with either MCF-7 experimental medium or CHO-hER growth medium containing vehicle (0.1% ethanol) or 1.0 nM 17ß-E2 in the absence or presence of increasing concentrations of TNF. After treatment, the cells were assayed for luciferase activity as previously described (14 ). The results are presented as the mean ± SEM (n = 8). *, P < 0.05, as determined by Dunnett’s ANOVA, compared with the corresponding no treatment control.

View larger version (19K): [in this window] [in a new window]   Figure 4. TNF activates NF-B in both MCF-7 and CHO-hER cells. MCF-7 cells were seeded with MCF-7 growth medium at 25,000 cells/well into 96-well plates (A), whereas CHO-hER cells were seeded with CHO-hER growth medium at 50,000 cells/well into 96-well plates (B); both cell lines were incubated overnight at 37 C. The cells were then infected with approximately 200 PFU/cell of Ad5-NFB-tk-luciferase, rinsed with PBS, and incubated for 48 h at 37 C with either MCF-7 experimental medium or CHO-hER growth medium containing increasing concentrations of TNF. After treatment, the cells were assayed for luciferase activity as previously described (14 ). The results are presented as the mean ± SEM (n = 8). *, P < 0.05, as determined by Dunnett’s ANOVA, compared with the corresponding no treatment control. The untreated control cells had essentially the same response as the lowest concentration of cytokine (data not shown).

TNF treatment does not suppress ER mRNA expression or block estradiol binding in HOB-03-CE6 cells

View larger version (24K): [in this window] [in a new window]   Figure 5. TNF treatment does not affect the steady state levels of ER or ERß mRNA expression by HOB-03-CE6 cells. Confluent 100-mm dishes of HOB-03-CE6 cells (50,000 cells/cm2) were rinsed with PBS and incubated for 48 h at 39 C with HOB-03-CE6 experimental medium containing either vehicle (medium, control) or 1.0 nM TNF. Total cellular RNA was then isolated, and quantitative RT-PCR was performed for ER mRNA (A) or semiquantitative RT-PCR was performed for ERß mRNA (B) as described in Materials and Methods. The equations for the lines in A were as follows: control, y = -0.976 x - 17.764 (r2 = -0.999); TNF, y = -1.054 x - 19.184 (r2 = -0.999).

View larger version (14K): [in this window] [in a new window]   Figure 6. TNF treatment does not block estradiol binding by HOB-03-CE6 cells. Confluent 150-mm dishes of HOB-03-CE6 cells (50,000 cells/cm2) were rinsed with PBS and incubated for 48 h at 39 C with HOB-03-CE6 experimental medium containing either vehicle (medium, control) or 0.1 nM TNF. Cytosolic extracts were then prepared, and [125I]17ß-E2 binding was performed as described in Materials and Methods. The results are presented as the mean ± SD (n = 3). *, P < 0.05, as determined by Dunnett’s ANOVA, compared with the corresponding nonspecific binding controls.

TNF is only a weak suppressor of ligand-dependent GR activity in HOB-03-CE6 cells

View larger version (19K): [in this window] [in a new window]   Figure 7. TNF is only a weak suppresser of ligand-dependent glucocorticoid receptor activity in HOB-03-CE6 cells. HOB-03-CE6 cells were seeded with HOB-03-CE6 growth medium at 17,000 cells/well into 96-well plates and incubated overnight at 34 C. The cells were then infected with Ad5-GRE-luciferase, rinsed with PBS, and incubated for 48 h at 39 C with HOB-03-CE6 experimental medium containing either vehicle (0.1% ethanol) or increasing concentrations of Dex in the absence or presence of 100 pM TNF (A). Conversely, the cells were treated with either vehicle or 50 nM Dex in the absence or presence of increasing concentrations of TNF (B). After treatment, the cells were assayed for luciferase activity as previously described (14 ). The results are presented as the mean ± SEM (n = 8). *, P < 0.05, as determined by Dunnett’s ANOVA, compared with the corresponding no treatment control.

IL-1 and IL-1ß, but not IL-11 or leukemia inhibitory factor (LIF), also suppress ligand-dependent ER activity in HOB-03-CE6 cells

View larger version (27K): [in this window] [in a new window]   Figure 8. Suppression of ligand-dependent ER activity by IL-1 and IL-1ß, but not by IL-11 or LIF, in HOB-03-CE6 cells. HOB-03-CE6 cells were seeded with HOB-03-CE6 growth medium at 17,000 cells/well into 96-well plates and incubated overnight at 34 C. The cells were then infected with Ad5-ERE-tk-luciferase, rinsed with PBS, and incubated for 48 h at 39 C with HOB-03-CE6 experimental medium containing either vehicle (0.1% ethanol) or 1.0 nM 17ß-E2 in the absence or presence of increasing concentrations of IL-, IL-ß, IL-11, or LIF. After treatment, the cells were assayed for luciferase activity as previously described (14 ). The results are presented as the mean ± SEM (n = 8). *, P < 0.05, as determined by Dunnett’s ANOVA, compared with the corresponding no treatment control. The untreated control cells had essentially the same response as the lowest concentration of cytokine (data not shown).

TNF and estrogen mutually suppress each other’s activity in HOB-03-CE6 cells

View larger version (19K): [in this window] [in a new window]   Figure 9. TNF and 17ß-E2 mutually suppress each other’s activity in the HOB-03-CE6 cells. HOB-03-CE6 cells were seeded with HOB-03-CE6 growth medium at 17,000 cells/well into 96-well plates and incubated overnight at 34 C. The cells were then infected with Ad5-ERE-tk-luciferase (A) or Ad5-NFB-tk-luciferase (B), rinsed with PBS, and incubated for 48 h at 39 C with HOB-03-CE6 experimental medium containing either vehicle (0.1% ethanol) or 1.0 nM 17ß-E2 in the absence or presence of increasing concentrations of TNF. After treatment, the cells were assayed for luciferase activity as previously described (14 ). The results are presented as the mean ± SEM (n = 8). *, P < 0.05, as determined by Dunnett’s ANOVA, compared with the corresponding no treatment control.

View larger version (17K): [in this window] [in a new window]   Figure 10. Estrogen does not block TNF- or IL-1ß-induced IL-6 secretion from HOB-03-CE6 cells. HOB-03-CE6 cells were seeded with HOB-03-CE6 growth medium at 17,000 cells/well into 96-well plates and incubated overnight at 34 C. The cells were then treated for 24 h at 39 C with HOB-03-CE6 experimental medium containing either vehicle (0.1% ethanol) or 10 nM 17ß-E2 in the absence or presence of increasing concentrations of TNF (A). Alternatively, the cells were treated for 16 h at 39 C with HOB-03-CE6 experimental medium containing either vehicle or 1.0 nM 17ß-E2 in the absence or presence of increasing concentrations of IL-1ß (B). After treatment, the conditioned medium was assayed for the concentration of IL-6 using an enzyme-linked immunosorbent assay, and the cells were analyzed for protein concentration as previously described (15 ). The results are presented as the mean ± SEM (n = 8). *, P < 0.05, as determined by Dunnett’s ANOVA, compared with the corresponding no treatment control. The untreated control cells had essentially the same response as the lowest concentration of cytokine (data not shown).

Evidence that NF-B p50 is involved in the suppression of ligand-dependent ER activity by TNF in HOB-03-CE6 cells

View larger version (18K): [in this window] [in a new window]   Figure 11. An antisense oligonucleotide to NF-B p50 partially reverses the suppression of ligand-dependent ER activity by TNF in HOB-03-CE6 cells. HOB-03-CE6 cells were seeded with HOB-03-CE6 growth medium at 17,000 cells/well into 96-well plates and incubated overnight at 34 C. The cells were then infected with Ad5-ERE-tk-luciferase, rinsed with PBS, and incubated for 24 h at 39 C with HOB-03-CE6 experimental medium containing either vehicle (0.1% ethanol) or 1.0 nM 17ß-E2 in the absence or presence of 3.8 µM of initiation-site directed antisense (AS), sense (S), or scrambled (Scram) phosphorothioate oligonucleotides to the human NF-B subunits p65/RelA, p50, and c-Rel (21 ). The cells were then treated for an additional 24 h in the absence or presence of 10 pM TNF. After treatment, the cells were assayed for luciferase activity as previously described (14 ). In these experiments, treatment of the cells with 1.0 nM 17ß-E2 up-regulated luciferase expression 3- to 4-fold. The results are presented as the mean ± SEM (n = 8). *, P < 0.05, as determined by Dunnett’s ANOVA, compared with the E2 and TNF control.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In these studies, we used the trans-activation of Ad5-ERE-tk-luciferase as a sensitive and selective assay of endogenous ER activity in HOB-03-CE6 cells. With this assay, we showed that treatment of the cells with a bone-forming growth factor (TGFß1) up-regulated ER activity in the absence of estrogen, whereas treatment of the cells with bone-resorbing cytokines (TNF, IL-1, and IL-1ß) down-regulated estrogen-dependent receptor function. Overall, these observations suggest a correlation between bone remodeling and the regulation of osteoblast ER activity. However, we should note that other growth factors and cytokines (i.e. IGF-I, EGF, PDGF, IL-11, and LIF) were inactive as receptor regulators in this assay, indicating that this is not a generalized phenomenon. One apparent connection between these results is that TGFß1 signals through a serine/threonine kinase receptor (reviewed in Ref. 44), whereas TNF, IL-1, and IL-1ß signal through NF-B (11).

The suppression of HOB-03-CE6 ER activity by TNF, IL-1, and IL-1ß was dose related and resulted in complete inhibition of estrogen-dependent ERE-tk-luciferase expression without affecting basal enzyme activity. In the case of IL-1 and IL-1ß, the IC₅₀ values for this effect (0.05–0.8 pM) were within the circulating concentration range for these cytokines in women (45), suggesting that the suppression of osteoblast ER activity by IL-1 and IL-1ß is potentially a physiologically relevant event. Although serum levels of TNF in humans are low (0.15 pM) (46), the concentration of this cytokine in the bone microenvironment after estrogen loss may be much higher (47). For TNF, the inhibition of ligand-dependent receptor function was due to the peptide itself, as it was reversed by preincubation with a neutralizing monoclonal antibody to TNF, but not by one to TNFß. Interestingly, this suppression did not appear to involve the prevention of DNA binding, as a DNA electrophoretic mobility shift assay failed to demonstrate inhibition of receptor binding to an ERE after TNF treatment (data not shown). This observation was not entirely surprising, because interaction between the ER and NF-B does not always result in altered DNA binding as measured by this technique (48). It was not possible to demonstrate that TNF, IL-1, and IL-1ß also suppress endogenous estrogenic responses in HOB-03-CE6 cells. Although several estrogen-responsive genes have been identified in these cells (14), the expression of these genes is also regulated by the cytokines independently of steroid hormone. For example, estrogen up-regulated alkaline phosphatase expression, but down-regulated basal IL-6 mRNA levels in HOB-03-CE6 cells (14), while TNF, IL-1, and IL-1ß had the opposite effects (the current study and data not shown). However, the inability of 17ß-E₂ to suppress TNF-, IL-1-, and IL-1ß-induced IL-6 secretion from HOB-03-CE6 cells may result from the prior inactivation of ligand-dependent ER activity after cytokine treatment.

TNF was also a potent and efficacious suppressor of ligand-dependent ER activity in MCF-7 breast cancer cells, but the cytokine had no effect on receptor function in CHO-hER cells, even though both cell lines responded to TNF as measured by the trans-activation of Ad5-NFB-tk-luciferase activity. This observation initially suggested to us that TNF may inhibit endogenous ER gene expression or steroid binding in HOB-03-CE6 and MCF-7 cells. However, this did not appear to be the case in HOB-03-CE6 cells at least, as cytokine treatment had no effect on either ER or ERß mRNA levels or [¹²⁵I]17ß-E₂ binding. Previous reports have also shown an interaction between NF-B and the ER in MCF-7 cells (8, 49, 50), and TNF has been reported to antagonize estrogen-stimulated proliferation of these cells (51).

In contrast to the potent and efficacious suppression of ligand-dependent ER activity, TNF treatment of the HOB-03-CE6 cells did not have a dramatic effect on ligand-dependent GR activity, suggesting that this inhibition was at least partly receptor selective in these cells. One explanation for this difference may be that human osteoblasts express approximately 10 times more GRs per cell than ERs (14, 52), and therefore, the relative ratio of NF-B to the ER is greater than that to the GR. Alternatively, this selectivity may result from the differential expression and/or activity of NF-B subtypes in HOB-03-CE6 cells. A recent study using cotransfected COS-1 (simian virus 40-transformed African green monkey kidney) cells demonstrated that p65, but not p50, was responsible for the mutual antagonism observed between GR and NF-B (53). These researchers also showed that p65 repressed progesterone receptor, AR, and ER function as well, but that only the GR (and to lesser extent the AR) effectively inhibited NF-B activity in these cells. Thus, the mutual suppression of steroid receptor and NF-B activities appears to be receptor, cell type, and NF-B subtype selective (53). These observations in HOB-03-CE6 cells are interesting in light of the fact that pharmacological concentrations of glucocorticoids also potentiate bone loss (54), albeit by a different mechanism than cytokine-induced osteopenia (6).

Finally, with the use of initiation site-directed antisense oligonucleotides to several NF-B subunits (21), we have demonstrated that incubation of the HOB-03-CE6 cells with an antisense oligonucleotide to p50 partially reversed the suppressive effects of TNF on ligand-dependent ER activity, whereas other antisense, sense, and scrambled oligonucleotides had no effect. These data provide preliminary evidence that activation of NF-B p50 may be involved in this suppression. Interestingly, NF-B1, which is a homodimer of p50 subunits, may function primarily as a transcriptional repressor (11). However, this work with the HOB-03-CE6 cells differs from previous reports using cotransfected U2-OS (human osteosarcoma) and HeLa (human cervical carcinoma) cells (9, 12). In these studies, the suppression of ER activity (also measured by the trans-activation of ERE-tk-luciferase) was shown to be mediated by p65/RelA and NF-IL6. These discrepancies may result from a number of factors, including the cell types used in these studies as well as endogenous expression (HOB-03-CE6 cells) vs. exogenous overexpression (9, 12). NF-IL6 and p65/RelA as well as c-Rel have also been shown to be involved in the down-regulation of IL-6 promoter activity by the ER in various cell lines (9, 12, 49). Although treatment of HOB-03-CE6 cells with 17ß-E₂ down-regulated basal IL-6 message levels (14) and NF-B promoter activity, we do not know which form(s) of NF-B/NF-IL6 is involved in this process. As the level of repression of ligand-dependent HOB-03-CE6 ER activity by TNF was greater than the level of inhibition of cytokine-dependent NF-B function by 17ß-E₂, it is conceivable that different NF-B dimers are involved in this mutual suppression. This conclusion is supported by the failure of estrogen to block cytokine-induced IL-6 production in HOB-03-CE6 cells. Thus, although sequestration of p65/RelA, c-Rel or NF-IL6 by the ER may lead to the suppression of NFB-tk-luciferase activity or the down-regulation of basal IL-6 expression (14), the corresponding activation of NF-B p50 by TNF appears to result in the inhibition of ligand-dependent ER activity. The inhibition of estrogen-activated receptor function by TNF, IL-1, and IL-1ß may also help to explain some of the controversy surrounding the role of IL-6 in estrogen-mediated bone loss (7, 55, 56, 57). Although interactions between the ER and NF-B/NF-IL6 or the suppression of IL-6 expression and secretion by 17ß-E₂ have been readily observable with overexpression (9, 10, 12, 13, 48), these events have been difficult to reproducibly detect in cells such as normal human osteoblasts, which express low receptor levels (55, 56, 57).

In summary, we have shown that TGFß1, a growth factor involved in bone formation, activates nonliganded ER activity in human osteoblasts, whereas bone-resorbing cytokines that signal through NF-B suppress ligand-dependent receptor function in these cells. These observations potentially place the ER at a pivotal junction between agents that modulate bone formation and bone resorption, and further support the critical role that estrogen plays in skeletal homeostasis (1). Perhaps more importantly, these results provide evidence for a potent feedback inhibition of estrogen action in human osteoblasts that is mediated by bone-resorbing cytokines such as TNF, IL-1, and IL-1ß. This, in turn, may be related to the pathophysiology of diseases such as postmenopausal osteoporosis, which result in high turnover bone loss.

	Acknowledgments

 
We thank Dr. Ramesh Bhat, Mr. Matt Bookler, Ms. Ruth Henderson, Ms. Deborah McMahon, Ms. Helga Ponce-de-Leon, Ms. Joan Scott, and Ms. Barbara Stauffer for technical assistance. We also thank Drs. Douglas Harnish and John Robinson for critically reviewing the manuscript.

Received July 14, 1998.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Riggs BL, Khosla S, Melton III LJ 1998 A unitary model for involutional osteoporosis: estrogen deficiency causes both type I and type II osteoporosis in postmenopausal women and contributes to bone loss in aging men. J Bone Miner Res 13:763–773[CrossRef][Medline]
2. Dempster DW, Lindsay R 1993 Pathogenesis of osteoporosis. Lancet 341:797–801[CrossRef][Medline]
3. Horowitz MC 1993 Cytokines and estrogen in bone: anti-osteoporotic effects. Science 260:626–627[Free Full Text]
4. Eastell R 1998 Treatment of postmenopausal osteoporosis. N Engl J Med 338:736–746[Free Full Text]
5. Turner RT, Riggs BL, Spelsberg TC 1994 Skeletal effects of estrogen. Endocr Rev 15:275–300[CrossRef][Medline]
6. Manolagas SC, Jilka RL 1995 Bone marrow, cytokines, and bone remodeling: emerging insights into the pathophysiology of osteoporosis. N Engl J Med 332:305–311[Free Full Text]
7. Girasole G, Jilka RL, Passeri G, Boswell S, Boder G, Williams DC, Manolagas SC 1992 17ß-Estradiol inhibits interleukin-6 production by bone marrow-derived stromal cells and osteoblastic cells in vitro: a potential mechanism for the anti-osteoporotic effect of estrogens. J Clin Invest 89:883–891
8. Galien R, Evans HF, Garcia T 1996 Involvement of CCAAT/enchancer binding protein and nuclear factor-B binding sites in interleukin-6 promoter inhibition by estrogens. Mol Endocrinol 10:713–722[Abstract]
9. Stein B, Yang MX 1995 Repression of the interleukin-6 promoter by estrogen receptor is mediated by NF-B and C/EBPß. Mol Cell Biol 15:4971–4979[Abstract]
10. Kassem M, Harris SA, Spelsberg TC, Riggs BL 1996 Estrogen inhibits interleukin-6 production and gene expression in a human osteoblastic cell line with high levels of estrogen receptors. J Bone Miner Res 11:193–199[Medline]
11. Baldwin AS 1996 The NF-B and IB proteins: new discoveries and insights. Annu Rev Immunol 14:649–681[CrossRef][Medline]
12. Pottratz ST, Bellido T, Mocharla H, Crabb D, Manolagas SC 1994 17ß-Estradiol inhibits expression of human interleukin-6 promoter-reporter constructs by a receptor-dependent mechanism. J Clin Invest 93:944–950
13. Ray A, Prefontaine KE, Ray P 1994 Down-modulation of interleukin-6 gene expression by 17ß-estradiol in the absence of high-affinity DNA binding by the estrogen receptor. J Biol Chem 269:121940–12946
14. Bodine PVN, Green J, Harris HA, Bhat RA, Stein GS, Lian JB, Komm BS 1997 Functional properties of a conditionally-phenotypic, estrogen-responsive, human osteoblast cell line. J Cell Biochem 65:368–387[CrossRef][Medline]
15. Bodine PVN, TrailSmith M, Komm BS 1996 Development and characterization of a conditionally-transformed adult human osteoblastic cell line. J Bone Miner Res 11:806–819[Medline]
16. Stein GS, Lian JB 1993 Molecular mechanisms mediating proliferation/differentiation interrelationships during progressive development of the osteoblastic phenotype. Endocr Rev 14:424–442[CrossRef][Medline]
17. Migliaccio S, Wetsel WC, Fox WM, Washburn TF, Korach KS 1993 Endogenous protein kinase-C activation in osteoblast-like cells modulates responsiveness to estrogen and estrogen receptor levels. Mol Endocrinol 7:1133–1143[Abstract]
18. Watts CKW, King RJB 1994 Overexpression of estrogen receptor in HTB 96 human osteosarcoma cells results in estrogen-induced growth inhibition and receptor cross talk. J Bone Miner Res 9:1251–1258[Medline]
19. Huo B, Dossing DA, Dimuzio MT 1995 Generation and characterization of a human osteosarcoma cell line stably transfected with the human estrogen receptor gene. J Bone Miner Res 10:769–781[Medline]
20. Harris SA, Tau KR, Enger RJ, Toft DO, Riggs BL, Spelsberg TC 1995 Estrogen response in the hFOB1.19 human fetal osteoblastic cell line stably transfected with the human estrogen receptor gene. J Cell Biochem 59:193–201[CrossRef][Medline]
21. Roshak AK, Jackson JR, McGough K, Chabot-Fletcher M, Mochan E, Marshall LA 1996 Manipulation of distinct NFB proteins alters interleukin-1ß-induced human rheumatoid synovial fibroblast prostaglandin E₂ formation. J Biol Chem 271:31496–31501[Abstract/Free Full Text]
22. Allgood VE, Oakley RH, Cidlowski JA 1993 Modulation by vitamin B6 of glucocorticoid receptor-mediated gene expression requires transcription factors in addition to the glucocorticoid receptor. J Biol Chem 268:20870–20876[Abstract/Free Full Text]
23. David-Watine B, Israel A, Kourilsky P 1990 The regulation and expression of MHC class I genes. Immunol Today 11:286–292[CrossRef][Medline]
24. Bodine PVN, Henderson RA, Green J, Aronow M, Owen T, Stein GS, Lian JB, Komm BS 1998 Estrogen receptor- is developmentally regulated during osteoblast differentiation and contributes to selective responsiveness of gene expression. Endocrinology 139:2048–2057[Abstract/Free Full Text]
25. Mosselman S, Polman J, Dijkema R 1996 ERß: identification and characterization of a novel human estrogen receptor. FEBS Lett 392:49–53[CrossRef][Medline]
26. Katzenellenbogen BS 1996 Estrogen receptors: bioactivities and interactions with cell signaling pathways. Biol Reprod 54:287–293[Abstract]
27. Vignon F, Bouton MM, Rochefort H</