help button home button Endocrine Society Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology, doi:10.1210/en.2004-0619
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Baker, A. E.
Right arrow Articles by Watters, J. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Baker, A. E.
Right arrow Articles by Watters, J. J.
Endocrinology Vol. 145, No. 11 5021-5032
Copyright © 2004 by The Endocrine Society

Estrogen Modulates Microglial Inflammatory Mediator Production via Interactions with Estrogen Receptor ß

Ann E. Baker, Vielska M. Brautigam and Jyoti J. Watters

Department of Comparative Biosciences (A.E.B., V.M.B., J.J.W.) and Program in Endocrinology and Reproductive Physiology (A.E.B., J.J.W.), University of Wisconsin, Madison, Wisconsin 53706

Address all correspondence and requests for reprints to: Jyoti J. Watters, Ph.D., Department of Comparative Biosciences, 2015 Linden Drive, Madison, Wisconsin 53706. E-mail: jjwatters{at}wisc.edu.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Estrogens are well known to exert antiinflammatory effects outside the central nervous system (CNS). They have also been shown to exert neuroprotective effects in the CNS after several types of injury, including neurodegeneration. However, the molecular mechanisms by which these effects occur remain unclear. Because microglial hyperactivation and their production of neurotoxins is associated with many types of brain injury for which estrogens are beneficial, we sought to investigate the ability of estrogen to modulate microglial function. Furthermore, because little is known regarding the role of each of the two known estrogen receptors (ERs) in microglia, our studies were designed to test the hypothesis that 17ß-estradiol (E2) exerts antiinflammatory effects in microglia, specifically via interactions with ERß. We tested this hypothesis using the murine microglial cell line BV-2, which naturally expresses only ERß. Our results indicate that not only does E2 decrease lipopolysaccharide (LPS)-stimulated nitric oxide (NO) production and inducible nitric oxide synthase (iNOS) expression, it also reduces the expression of cyclooxygenase-2, a target for estrogen that has not previously been reported for ERß. We also observed that LPS-stimulated TNF{alpha} mRNA was increased by estrogen. E2 exerts these effects within 30 min compared with typical estrogen transcriptional responses. Tamoxifen and ICI 182,780 differentially blocked the inhibitory effects of E2 on LPS-stimulated iNOS and cyclooxygenase-2. In addition, we show that E2 alters LPS-stimulated MAPK pathway activation, supporting the idea that alterations in the MAPKs may be a potential mechanism by which ERß mediates decreased microglial activation.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
MICROGLIA ARE IMMUNOCOMPETENT, macrophage-like cells residing in the central nervous system (CNS). Their activity is beneficial to neurons in that they produce growth factors that promote neuronal survival and viability; they also function to remove cellular debris from the brain during normal neuronal remodeling processes. However, microglial hyperactivation and their release of neurotoxic inflammatory mediators can also contribute to neuronal and glial cell death in many neurodegenerative diseases. Therefore, controlling microglial activation may have therapeutic benefit in diseases of the CNS. Because there are gender differences in the prevalence of some neurodegenerative diseases like multiple sclerosis (MS) and Alzheimer’s disease (AD), and estrogens have been shown to improve the symptoms of MS (1, 2, 3) and lower the risk of developing AD (4, 5, 6), estrogen may play a role in the etiology of these disorders; however, the mechanisms by which it exerts these neuroprotective effects remain poorly understood.

Because inflammation and microglial hyperactivation is a hallmark of MS, AD, and the pathology of many other neurodegenerative disorders, microglia may be an important target of estrogen action. Upon their activation, microglia release a number of potentially neurotoxic substances including cytokines such as TNF{alpha}; reactive oxygen species such as nitric oxide (NO) synthesized by inducible nitric oxide synthase (iNOS); and prostaglandins such as prostaglandin E2 (PGE2), synthesized by cyclooxygenase-2 (COX-2). Moreover, inhibiting the activities of iNOS and COX-2 ameliorates brain damage (7, 8, 9, 10, 11, 12), whereas the effects of TNF{alpha} have been reported to be both neurotoxic and neuroprotective (13, 14, 15, 16, 17). Although there is little information available regarding the effects of estrogen on microglia, 17ß-estradiol (E2) decreases microglial cell release of multiple inflammatory mediators including NO and PGE2 (18, 19). Bruce-Keller et al. (18) have further shown that E2 stimulates ERK-1/2 activation, which is involved in inhibiting superoxide production stimulated by phorbol ester. This change in MAPK activation, found to be important for superoxide production, is consistent with our previous findings regarding NO production in macrophages, where an increase in ERK activation correlated with decreased NO production (20, 21).

The differential roles of the two known estrogen receptor (ER) subtypes in mediating the antiinflammatory effects of E2 in microglia are not known. Both ER{alpha} and ERß are widely expressed in nonneuronal cells of the CNS (22, 23, 24, 25, 26, 27, 28), but their relative expression levels in microglia from different brain regions continue to be unexplored. The microglia used in the studies by Vegeto et al. (19) and Bruce-Keller et al. (18) were derived from forebrain and cerebral cortex respectively, and they were shown to express both ER{alpha} and ERß. A recent report indicates that ER{alpha} may mediate the antiinflammatory effects of E2 in lipopolysaccharide (LPS)-stimulated microglia residing in the hippocampus (CA1 region) (29); however, the broad distribution of ERß throughout the brain, in addition to the lack of available information regarding ER subtype expression profiles in microglia derived from other brain regions, suggests that the role of ERß in these processes merits investigation. We therefore examined the antiinflammatory effects of estrogen in microglia that express only ERß, and we report the molecular mechanisms whereby this receptor can mediate these actions.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Cell culture
Murine BV-2 (30) microglial cells were routinely cultured in DMEM (Cellgro, Herndon, VA) supplemented with 5% fetal bovine serum (Hyclone Logan, UT) and 100 U/ml penicillin/streptomycin (Cellgro) in 100-mm Sarstedt plates. For each experiment, cells were plated at a density of 7.5 x 104 to 1 x 105 in Sartstedt multiwell plates and were allowed to grow overnight. The next day, the cells were washed and steroid hormone-deprived in phenol red-free DMEM supplemented with 0.25% dextran-coated charcoal-stripped fetal bovine serum (CSFBS), 100 U/ml penicillin/streptomycin, and 100 U/ml glutamate. The cells were serum starved for 16 h, after which time the medium was changed to phenol red-free DMEM supplemented with 5% CSFBS, 100 U/ml penicillin/streptomycin, and 100 U/ml glutamate. The cells were then treated with: E2 (Sigma, St. Louis, MO); the short-acting ER agonist estriol (Steraloids Inc., Newport, RI); the selective ERß agonist diarylproprionitrile (DPN); the selective ER{alpha} agonist 4,4',4'-(4-propyl-[1H]-pyrazole-1,3,5-triyl)Trisphenol (PPT) (Tocris, Ellisville, MO); the phorbol ester PMA (phorbol 12-myristate 13-acetate), an ERK MAPK stimulus (100 nM; Sigma); the p38 MAPK stimulus anisomycin (10 ng/ml; Sigma); the partial ER agonist tamoxifen (Sigma); or the pure antiestrogen ICI 182,780 (AstraZeneca, Macclesfield, UK) as indicated in each experiment. After this time, the steroids were removed by washing with phenol red-free DMEM containing 5% CSFBS to facilitate greater steroid removal from the well. The cells were then activated by treatment with LPS serotype 0111:B4 (Sigma) for 16 h or as indicated in the figure. N9 microglia (31) were used as a control for ER expression, and were cultured in the same medium as described above for BV-2 cells, except using 10% FBS. In certain experiments as indicated, BV-2 cells were deprived of hormone by culturing them for 5–7 d in DMEM containing 5% CSFBS. The cells were then treated with 10 nM E2 for 24 h, followed by the addition of LPS (1 µg/ml) for 16–20 h. In these experiments, the E2 was not washed away, and the cells remained exposed to both LPS and E2 for the duration of the experiment. For experiments using the ER antagonists tamoxifen or ICI 182,780, the antagonist was given 30 min before stimulation with E2 for 1 h. Both agents were then washed out and the cells subsequently stimulated with LPS for 16–20 h. The mechanism of tamoxifen blockade of ER function involves its contact with the AF-2 domain, preventing AF-2 interaction with coactivators, whereas the inhibition of ER function by ICI 182,780 is thought to involve several mechanisms, including preventing ER interaction with coactivators as well as promoting ER degradation (32). For experiments using the MAPK inhibitors U0126 [a MAPK kinase (MEK)-ERK pathway inhibitor (Promega, Madison, WI)] or SB 202190 and SB 203580 (p38 MAPK inhibitors; Calbiochem, San Diego, CA), cells were pretreated with the inhibitors (10 µM) for 15 min before and during stimulation with PMA or anisomycin, or for 60 min before and during LPS treatment.

Immunoblotting
Whole cell lysates were prepared by lysing BV-2 cells in 2x SDS-PAGE sample buffer without bromophenol blue. Nuclear extracts for the analyses of ERß were prepared from BV-2 and N9 cells as described previously (33). Protein concentrations were determined using the Micro-BCA Protein Assay (Pierce Biochemical Co., Rockford, IL). Proteins (25 µg/lane for iNOS and COX-2; 50 µg/lane for ERß) were separated by 10% SDS-PAGE (34) and transferred to Immobilon polyvinylidene difluoride membrane. The membranes were blocked overnight in 5% nonfat milk/TBST [10 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.05% Tween 20]. Anti-iNOS (BD Transduction Laboratories, Lexington, KY) and anti-COX-2 (Upstate, Lake Placid, NY) antibodies were used at a dilution of 1:2500 in 0.25% gelatin/TBST. ERß antibodies (Zymed Laboratories Inc., San Francisco, CA) were used at a final concentration of 1 µg/ml in 0.25% gelatin/TBST, and antiactive ERK and p38 antibodies (Cell Signaling Technology, Beverly, MA) were used at a final dilution of 1:1000 in 5% milk/TBST. The immunoreactive bands were visualized using secondary antibodies conjugated to horseradish peroxidase (Santa Cruz Biotechnology, Santa Cruz, CA) and chemiluminescent detection methods (Pierce) with the UVP image analysis system to collect and analyze the data obtained from the chemiluminescent light emission (UVP Inc., Upland, CA). To confirm equal protein loading, membranes were also probed with antibodies that react with the cytosolic proteins Grb-2 or ß-tubulin, or with an ERK-1 antibody that cross-reacts with ERK-2 (all used at 1:5000 in 5% milk/TBST; Santa Cruz Biotechnology) as indicated in the figures. Densitometric data were normalized against Grb-2 loading control proteins.

RT-PCR
Murine BV-2 or N9 microglia were grown in six-well plates as detailed above. Cells were lysed in 500 µl of Tri-Reagent (Sigma), and total RNA was harvested as described by the manufacturer’s protocol. RT-PCR was performed using 1 µg of total RNA as a template for the RT reaction using random hexamers and ImProm-II Reverse Transcriptase (Promega). Reactions were performed according to the manufacturer’s protocol. The cDNA was then used for PCR of the following genes: ERß: 5'-CAGTAACAAGGGCATGGAAC and 5'-GTACATGTCCCACTTCTGAC (expected size, 242 bp); ER{alpha}: 5'-TTCTGACCATCGACGCCAGAAT and 5'-CATCATGCCCACTTCGTAACAC (expected size 294 bp); and GAPDH: 5'-TGCAAGATCAAGACAGCTGCATCT and 5'-CAGTGGATGCAGGGATGATGTTCT (expected size 320 bp).

Real-time PCR
The effect of E2 on TNF{alpha} mRNA expression was measured after BV-2 cell stimulation with either vehicle or LPS (1 µg/ml) for 6 h. One microgram of total RNA was reverse transcribed as described above, and quantitative RT-PCR was performed by monitoring in real-time the increase in fluorescence of the SYBR-GREEN dye as described (35) using the TaqMan 7000 Sequence Detection System (Applied Biosystems, Foster City, CA). The relative amounts of each gene between samples were determined using two methods yielding identical results. The comparative threshold cycle method (36) was used and the values normalized to the relative levels of ß-actin. The relative amounts of each gene between samples were also determined using a standard curve of serial dilutions of the cDNA containing the highest amount of the gene. These values were then normalized to the relative amounts of ß-actin cDNA, which were obtained from a similar standard curve. The oligonucleotide primer sequences were as follows: TNF{alpha}, 5'-CATCTTCTCAAATTCGAGTGACAA and 5'-TGGGAGTAGACAAGGTACAACCC; ß-actin, 5' TGTCCACCTTCCAGCAGATGT and 5' AGCTCAGTAACAGTCCGCCTAGA.

Measurement of NO production
BV-2 microglia were stimulated with LPS for 16–20 h, after which time the medium was removed and assayed for nitrite accumulation using the Griess reagent as previously described (21). Nitrite is a stable breakdown product of NO, and its levels directly correlate with the levels of NO produced.

Statistical analysis
Statistical analyses were performed using two-way Student’s t tests or an ANOVA pre hoc test and the Dunnett or Bonferroni Multiple Comparison post hoc analyses. Statistical significance was set at the 95% confidence limit (P < 0.05). Quantitative data are expressed as the mean ± SEM of at least three independent experiments.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Estrogen decreases NO production and iNOS and COX-2 levels while increasing TNF{alpha} mRNA in BV-2 microglia
Because there have been no reports to our knowledge of estrogen action in BV-2 microglia, we sought first to ascertain their ability to be modulated by E2. As shown in Fig. 1AGo, when BV-2 microglia were exposed to E2 for 24 h before and during stimulation with LPS, E2 greatly reduced LPS-stimulated iNOS and COX-2 levels as measured by immunoblot analyses, suggesting that these cells are responsive to estrogen, and that the effects of E2 in this model system may have relevance to chronic brain pathologies. Similar results were observed when the cells had been exposed to E2 for 48 h before stimulation with LPS (data not shown). To investigate the potential mechanism(s) whereby E2 modulates BV-2 cell inflammatory capacity, we investigated the length of time required for E2 to elicit its effects on inflammatory mediator production. To this end, BV-2 cells were pretreated with 1 nM E2 for the times indicated in Fig. 1BGo, followed by estrogen removal (by washout). The cells were then immediately stimulated with LPS (1 µg/ml) for 16 h. The LPS-stimulated increase in nitrite accumulation was significantly reduced by E2 preexposure and washout for as little as 30 min (Fig. 1BGo). The data are expressed as percent of LPS-stimulated nitrite accumulation to illustrate the effects of estrogen on NO production from several separate experiments (n ≥ 3; the average nitrite levels obtained with LPS treatment in this paradigm were between 6 and 10 µM). The inhibition of LPS-stimulated NO production by E2 is consistent with others in the literature, both in macrophages and microglia (18, 19, 37, 38). We also measured the ability of E2 to alter the expression of the cytokine gene TNF{alpha} (Fig. 1CGo). The few reports of the actions of estrogen on TNF{alpha} expression in microglia (39, 40, 41), as well as others in peripheral immune cell types indicate that E2 has a suppressive or inhibitory effect on TNF{alpha} gene expression and/or release. Interestingly, we observed an increase in TNF{alpha} mRNA with E2 preexposure times of as little as 1 h. Immunoblot analyses of both iNOS and COX-2 levels in E2-treated BV-2 microglia show a marked reduction with somewhat different kinetics, when compared with LPS treatment alone (Fig. 1DGo, upper panel). This decrease occurs with as little as 30–60 min of E2 preexposure. As illustrated in the averaged data (n ≥ 3) in the lower panel of Fig. 1DGo, both iNOS and COX-2 protein levels are significantly reduced by E2 treatment. These inhibitory effects of E2 on COX-2 expression are consistent with the observations by Vegeto et al. (19), where PGE2 production in the presence of estrogen was greatly reduced. Interestingly, the inhibitory effects of E2 on iNOS and COX-2 were not observed when LPS and E2 were added simultaneously (data not shown), or when BV-2 microglia were exposed to E2 for 4 h or more before its withdrawal and stimulation with LPS (Fig. 1DGo). To ascertain the dose response of the E2 effect on the inhibition of iNOS and COX-2 protein levels, BV-2 cells were preexposed to the doses of E2 indicated in Fig. 1EGo for 1 h. The medium containing E2 was then removed and the cells were stimulated with LPS (1 µg/ml) for 16 h. Concentrations of E2 as low as 10–14 M began to exert a detectable and statistically significant decrease in iNOS levels. Statistical significance for inhibition of COX-2 was reached at 10–13 M E2. These concentrations of estrogen suggest that only several hundred receptors are likely necessary for the E2 effect in BV-2 microglia.



View larger version (33K):
[in this window]
[in a new window]
  		FIG. 1. LPS-stimulated iNOS and COX-2 expression decreases with E2. A, Representative immunoblot (of at least three separate experiments) for iNOS and COX-2 expression in BV-2 microglial cells exposed to vehicle (ethanol) or E2 (10 nM) for 24 h preceding LPS exposure for 20 h. E2 was present during the LPS exposure. Each treatment (LPS or LPS + E2) is shown in duplicate. ß-Tubulin is shown as a loading control. B, Measurement of nitrite in the medium of BV-2 microglial cell cultures pretreated with E2 (1 nM) or ethanol (vehicle) for the times indicated. The E2 was washed out and the cells were exposed to LPS (1 µg/ml) for 16 h. All treatments were performed in hormone-free, phenol red-free DMEM. Values are expressed as the percent of LPS stimulation, and the data represent the means ± SEM of at least three separate experiments. C, Quantitative PCR analysis of TNF{alpha} expression (n = 3) after a time course of E2 pretreatment and washout before LPS exposure for 6 h. The data shown are the means ± SEM of three individual experiments, and they are expressed as the fold change relative to LPS. D, Representative immunoblot for iNOS and COX-2 expression in BV-2 microglial cells pretreated with E2 for the indicated times, preceding LPS exposure for 16 h. E2 had no effect on iNOS or COX-2 levels alone and Grb-2 is shown as a loading control. The lower panel shows the densitometric analysis of iNOS and COX-2 immunoreactivity obtained from immunoblot analyses (n ≥ 4) performed as shown in the upper panel, after E2 pretreatment for the times indicated. E, Dose response of E2 inhibition of LPS-stimulated iNOS and COX-2. BV-2 microglia were exposed to the indicated concentrations of E2 for 1 h, followed by washout before stimulation with LPS. Quantification of iNOS and COX-2 immunoreactivity obtained from the densitometric analyses of immunoblot studies (n ≥ 3). Values are expressed as percent of LPS and represent the mean ± SEM of at least three separate experiments. *, P < 0.05; **, P < 0.01 vs. LPS treatment.

 
ERß, not ER{alpha}, is detectably expressed by BV-2 microglia
Because we had observed significant effects of E2 on BV-2 microglial production of inflammatory mediators, we used RT-PCR and immunoblot analyses to establish the identity of the ERs expressed in these cells (Fig. 2Go). We used N9 microglia as a control for both ER{alpha} and ERß expression as had been previously reported (18). By RT-PCR, BV-2 cells expressed only detectable levels of ERß, suggesting that our observed effects with E2 were likely mediated through ERß (Fig. 2AGo). The expression of ERß by BV-2 cells was confirmed at the protein level by immunoblot analyses of nuclear extracts from N9 and BV-2 microglia (Fig. 2BGo). The BV-2 microglial cell line thus provides an excellent model for the study of inflammatory modulation by estrogen, specifically via interactions with ERß in microglia.



View larger version (31K):
[in this window]
[in a new window]
  		FIG. 2. The microglial cell line BV-2 expresses only detectable levels of ERß, not ER{alpha}. A, Representative RT-PCR analysis for ER{alpha} and ERß expression in BV-2 and N9 microglial cell lines. Expected length of the amplification product is 294 bp for ER{alpha} and 242 bp for ERß. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is shown as a loading control. B, Representative immunoblot for ERß expression in nuclear lysates from N9 and BV-2 microglia.

 
The ERß-selective ligand DPN mimics the inhibitory effect of E2 on LPS-stimulated NO, iNOS, and COX-2 levels in BV-2 microglia
Figure 3AGo demonstrates the ability of DPN (1 nM), a selective ERß ligand (42), to reduce NO production stimulated by LPS in BV-2 microglia. Consistent with our observations with E2 (Fig. 1AGo), a significant decrease in nitrite accumulation was observed upon DPN preexposure and removal for as little as 30 min. The data are expressed as percent of LPS-stimulated nitrite accumulation to illustrate the results from multiple separate experiments (n ≥ 3; the average nitrite levels obtained with LPS treatment were between 6 and 10 µM). LPS-stimulated iNOS and COX-2 protein levels were also examined in response to DPN preexposure and washout, for the times indicated in the figure (Fig. 3BGo, upper panel). As was observed with E2 treatment, DPN also markedly reduced iNOS and COX-2 expression within 30 min. The lower panel of Fig. 3BGo illustrates the quantification of the averaged densitometric data (n ≥ 3). DPN significantly inhibited LPS-stimulated iNOS and COX-2 within 30 min, and this inhibitory effect was lost between 4 and 8 h after E2 washout/withdrawal. A dose response of the inhibitory effect of DPN on iNOS and COX-2 protein levels is shown in Fig. 3CGo. Unlike the concentrations of E2 required to attenuate iNOS expression (Fig. 1DGo), higher concentrations of DPN were required. The requirement for more DPN to exert the same effect as E2, is consistent with the lower affinity of DPN for ERß compared with that of E2 (43).



View larger version (30K):
[in this window]
[in a new window]
  		FIG. 3. LPS-stimulated iNOS and COX-2 expression decreases with DPN pretreatment, an ERß selective ligand. A, Measurement of nitrite in the medium of BV-2 microglial cell cultures pretreated with DPN (1 nM) or ethanol (vehicle) for the times indicated. The DPN was washed out and the cells were exposed to LPS (1 µg/ml) for 16 h. All treatments were done in hormone-free, phenol red-free DMEM. Values are expressed as the percent of LPS stimulation, and the data represent the means ± SEM of at least three separate experiments. B, Representative immunoblot for iNOS and COX-2 expression in BV-2 microglial cells pretreated with DPN for the indicated times, before LPS exposure for 16 h. DPN had no effect on iNOS or COX-2 levels alone and Grb-2 is shown as a loading control. The lower panel shows the densitometric analysis of iNOS and COX-2 immunoblots (n ≥ 4) after DPN pretreatment for the times indicated performed as shown in the upper panel. C, Dose response of DPN inhibition of LPS-stimulated iNOS and COX-2. Quantification of iNOS and COX-2 immunoreactivity (n ≥ 3) obtained from the densitometric analyses of immunoblot studies after DPN pretreatment at the concentrations indicated. Values are expressed as percent of LPS and represent the mean ± SEM of at least three separate experiments. *, P < 0.05; **, P < 0.01 vs. LPS treatment.

 
The steroid hormone profile of iNOS and COX-2 inhibition is consistent with an ERß-mediated effect
Although detectable levels of ER{alpha} expression were not observed in BV-2 cells (Fig. 2Go), we used PPT, a selective ER{alpha} agonist that has a 410 times lower affinity for ERß than it does for ER{alpha} (44), to test the specificity of our proposed ERß effects. To do this, BV-2 microglia were pretreated with E2, DPN, or PPT (all at 1 nM) for 1 h, after which time the hormones were washed away and the cells stimulated with LPS (1 µg/ml) for 16 h. As indicated in the quantified densitometric data (n ≥ 3) shown in Fig. 4AGo, E2 and DPN significantly reduced LPS-stimulated iNOS and COX-2 levels, whereas PPT, the ER{alpha} selective ligand, was without effect. These data suggest that ER{alpha}, even if present at levels below the limit of our detection, is likely not mediating the effects of estrogen in this system. Furthermore, they support the idea that estrogen can exert very potent modulatory effects on microglial inflammatory capacity via interactions with ERß. To further test the steroid hormone receptor specificity of the observed effects, we treated cells in the same paradigm described above, with E2, cholesterol, 17{alpha}-estradiol, and dihydrotestosterone (DHT) (all at 1 nM). Because BV-2 cells do not express detectable levels of progesterone or glucocorticoid receptors (Baker, A. E., and J. J. Watters, unpublished observations), we did not include agonists for these receptors in the studies reported here. Figure 4BGo shows the quantified densitometry data (n ≥ 3) of LPS-stimulated iNOS and COX-2 protein levels in the presence of the indicated steroids/compounds. E2 treatment reduced iNOS and COX-2 levels as seen before; however, cholesterol, 17{alpha}-estradiol, and DHT were all without effect, supporting the idea that the effects of estrogen (and DPN) are specifically mediated by ERs.



View larger version (18K):
[in this window]
[in a new window]
  		FIG. 4. The inhibitory effect of estrogen is specific to E2 and DPN. A, Comparison of the inhibition of LPS-stimulated iNOS and COX-2 protein levels by ER-specific ligands. Quantification of iNOS and COX-2 immunoreactivity obtained from the densitometric analyses of immunoblot studies (n ≥ 3) after a 1-h pretreatment with either ethanol (vehicle), E2 (1 nM), DPN (1 nM), or PPT (1 nM, an ER{alpha}-specific ligand). The ligands were removed and the cells were exposed to LPS (1 µg/ml) for 16 h. All treatments were done in hormone-free, phenol red-free DMEM. B, Comparison of other steroid hormones and cholesterol with E2 to inhibit LPS-stimulated iNOS and COX-2. Quantification of iNOS and COX-2 immunoreactivity obtained from the densitometric analyses of immunoblot studies (n ≥ 4) after a 1-h pretreatment and washout of ethanol (vehicle), E2 (1 nM), cholesterol (1 nM), 17{alpha}-estradiol (1 nM), or DHT (1 nM). Treatments without LPS did not alter iNOS or COX-2 levels. Values are expressed as percent of LPS and represent the mean ± SEM of at least four separate experiments. *, P < 0.05; **, P < 0.01 vs. LPS treatment.

 
Tamoxifen and ICI 182,780 have differential effects on E2-mediated inhibition of iNOS and COX-2
We further tested the idea that the effects of estrogen were mediated by ERs using the ER antagonists tamoxifen and ICI 182,780. BV-2 microglia were pretreated with vehicle, tamoxifen (100 nM), or ICI 182,780 (100 nM) for 30 min before the addition of E2 (1 nM) for 1 h, after which time the medium containing both the E2 and the antagonist was removed. The cells were then stimulated with LPS (1 µg/ml) for 16 h. iNOS and COX-2 protein levels were ascertained by immunoblot analyses and the quantified data obtained by densitometric analyses (n ≥ 3) are illustrated in Fig. 5BGo. The decrease in iNOS expression elicited by E2 was unaffected by the presence of ICI 182,780, but it was completely reversed in the presence of tamoxifen. Interestingly, both tamoxifen and ICI 182,780 effectively prevented the inhibitory actions of E2 on LPS-stimulated COX-2 levels, suggesting that E2 modulates expression of the iNOS and COX-2 genes differently. Using both higher and lower doses of ICI 182,780 failed to block the effects of E2 on iNOS levels (data not shown), implying that this lack of blockade is not due to an inappropriate dose of antagonist.



View larger version (27K):
[in this window]
[in a new window]
  		FIG. 5. E2 inhibition of LPS-stimulated iNOS and COX-2 is differentially blocked by the ER antagonists tamoxifen and ICI 182,780. A, Representative immunoblot for iNOS and COX-2 expression in BV-2 microglial cells pretreated in the presence and absence of the ER antagonists ICI 182,780 (ICI, 100 nM) or tamoxifen (Tam, 100 nM) for 30 min before treatment with E2 (1 nM) for 1 h. Both the E2 and the ER antagonists were removed by washout, and the cells were subsequently exposed to LPS (1 µg/ml) for 16 h. Grb-2 is shown as a loading control. B, Quantification of iNOS and COX-2 immunoreactivity obtained from the densitometric analyses of immunoblot studies (n ≥ 3) performed as shown in panel A. Values are expressed as percent of LPS and represent the mean ± SEM of at least three separate experiments. *, P < 0.05, vs. LPS treatment.

 
The short-acting estrogen estriol exerts similar effects to E2 on LPS-stimulated iNOS and COX-2 protein levels
Because we had observed a rapid response to E2 (the maximal inhibitory effects occurred within 30–60 min) with regard to decreasing iNOS and COX-2 protein levels, we confirmed these effects using an estrogen that has a shorter on-off rate than E2, to ensure that any residual E2 that might be remaining in the tissue culture wells after washing the cells could not account for our observations. We thus performed experiments similar to those described above where the cells were exposed to estriol for 1 h at the doses indicated in the figure (Fig. 6AGo). The estriol was then washed out of the wells using phenol red-free DMEM containing 5% CSFBS. All concentrations of estriol tested (between 10–8 and 10–10 M) effectively and similarly decreased iNOS and COX-2 protein levels. In addition, like E2, estriol exerted inhibitory effects on LPS-stimulated iNOS and COX-2 levels within 30–60 min (data not shown). Figure 6BGo shows the densitometric quantification of several independent experiments (n ≥ 3) and indicates that all concentrations of estriol used resulted in a statistically significant reduction in LPS-stimulated iNOS and COX-2 protein levels. These data lend support to the idea that estrogen exerts its inhibitory effects on iNOS and COX-2 expression within a short time period (30–60 min) and that the observed inhibitory effects of estrogen are likely not due to residual estrogen remaining in the wells after washout because the short-acting estrogen estriol exerts comparable effects within the same time period.



View larger version (31K):
[in this window]
[in a new window]
  		FIG. 6. LPS-stimulated iNOS and COX-2 expression decreases with estriol pretreatment. A, Representative immunoblot for iNOS and COX-2 expression in BV-2 microglial cells pretreated for 1 h with estriol at the indicated concentrations, which was removed before addition of LPS (1 µg/ml) for 16 h. Grb-2 is shown as a loading control. B, Dose response of estriol inhibition of LPS-stimulated iNOS and COX-2. Quantification of iNOS and COX-2 immunoreactivity obtained from the densitometric analyses of immunoblot studies (n ≥ 3) performed as shown in panel A. Values are expressed as percent of LPS and represent the mean ± SEM of at least three separate experiments. *, P < 0.05; **, P < 0.01 vs. LPS treatment.

 
The p38 and ERK-1/-2 MAPK pathways are critical regulators of LPS-stimulated iNOS and COX-2 expression in hormone-deprived BV-2 microglia
Because the data in the literature are conflicting with regard to which MAPK enzymes are specifically critical for LPS-stimulated iNOS expression (45, 46, 47, 48, 49), and the role of these enzymes in microglia that have been deprived of steroid hormones has not been reported, we tested the importance of the MAPKs ERK-1/-2 and p38 in regulating the expression of iNOS in response to LPS stimulation of hormone-deprived BV-2 microglia. We used the pharmacologic inhibitors U0126 (a MEK-1/-2 inhibitor that prevents activation of the MEK/ERK pathway), and SB 202190 and SB 203580 (p38 MAPK inhibitors) to investigate the role of these MAPK pathways in regulating the expression of the iNOS protein. Although each inhibitor exerted no detectable effect on iNOS expression by themselves, in the presence of LPS, the MEK inhibitor U0126 resulted in a strong increase in iNOS protein levels, whereas the p38 inhibitors SB 202190 and SB 203580 blocked LPS-induced expression of these proteins (Fig. 7AGo). Similar results were observed on nitrite production (data not shown). The densitometric data obtained from multiple experiments (n = 3) are shown in Fig. 7BGo. The robust increase in iNOS protein levels with MEK/ERK pathway inhibition is consistent with our previous observations on iNOS protein expression in macrophages (20, 21) and suggests that activation of the MEK/ERK pathway is inhibitory to LPS-stimulated iNOS protein expression. To ensure that the pharmacologic inhibitors are interfering with the MAPK pathways intended, the data in Fig. 7CGo show the specificity of each compound on p38 and ERK-1/-2 activation, detected using phosphorylation state-specific antibodies that recognize the enzymatically active forms of ERK-1/-2 (first blot) and p38 (third blot). As expected, U0126 blocked PMA-stimulated ERK-1/-2 activation while exerting no effect on anisomycin-stimulated p38 MAPK activation; and SB 202190 and SB 203580 inhibited p38 activation in response to anisomycin while exerting no effect on PMA-stimulated ERK-1/-2 activation. To verify protein loading, the blots were stripped and reprobed with an antibody that recognizes total ERK proteins (second and fourth blots).



View larger version (30K):
[in this window]
[in a new window]
  		FIG. 7. The p38 and ERK-1/-2 pathways are important regulators of LPS-stimulated iNOS and COX-2 expression in hormone-deprived microglia. Cells were pretreated with the MAPK inhibitors: U0126 (10 µM), a MEK/ERK pathway inhibitor and the p38 inhibitors SB 202190 and SB 203580 (10 µM each) for 60 min before and during LPS stimulation. A, Representative immunoblot of LPS-stimulated iNOS expression in BV-2 microglial cells. The cells were exposed to the inhibitors for the duration of the experiment (16 h). The inhibitors alone did not affect iNOS expression and Grb-2 is shown as a loading control. B, Quantification of iNOS and COX-2 immunoreactivity obtained from the densitometric analyses of immunoblot studies (n = 3) performed with the MAPK inhibitors as shown in panel A. Values are expressed as percent of LPS and represent the mean ± SEM of three separate experiments. *, P < 0.05; **, P < 0.01 vs. LPS treatment. C, BV-2 cells were treated with either 100 nM PMA or 10 ng/ml anisomycin for 15 min in the presence or absence of the MAPK inhibitors U0126, SB 202190, or SB 203580 (15 min pretreatment, each at 10 µM final concentration). The inhibitors remained for the duration of the 15 min treatment with PMA or anisomycin. First panel, Active ERK-1/-2; third panel, active p38. The blots were stripped and reprobed for total ERK immunoreactivity as a loading control (second and fourth panels).

 
Estrogen increases LPS-stimulated ERK-1/-2 activation while exerting no detectable effect on p38 MAPK activation
Given the rapidity with which estrogen exerts its inhibitory effects on LPS-stimulated iNOS protein expression, and having observed interesting regulation of LPS-stimulated iNOS/NO by MAPKs in hormone-deprived microglia, we reasoned that modulation of the MAPK pathway may be a potential target for the actions of estrogen. Thus, we tested the effect of E2 (1 nM for 1 h pretreatment and washout before stimulation with LPS for the indicated times) on MAPK activation. Short times of estrogen treatment on its own (between 5 and 60 min) had no effect on either ERK-1/-2 or p38 MAPK activation (data not shown). LPS induced the activation ERK-1 and ERK-2 in BV-2 microglia, an effect that was initially observed between 15 and 30 min (Fig. 8AGo). ERK-1/-2 activation remained elevated at 60 min, but it returned to baseline by 2 h (data not shown). In the presence of estrogen, LPS-stimulated ERK-1/-2 activation was augmented between 45 and 60 min. Figure 8Go, B and C, show by graphical representation that the densitometry of ERK-1 (Fig. 8BGo) and ERK-2 (Fig. 8CGo) activation obtained from several experiments performed in this same way (n ≥ 3). E2 pretreatment for 1 h, followed by washout, significantly increased both ERK-1 and ERK-2 activation in response to LPS stimulation at 60 min. The effect of estrogen was no longer observed at 2 h (data not shown). In contrast, p38 activation was unaffected (Fig. 8Go, D and E) by 1 h E2 pretreatment and washout. A representative immunoblot of active p38 stimulated by LPS is shown in Fig. 8DGo. LPS initiates measurable p38 activation within 15 min, an effect that returns to baseline between 30 and 45 min. These data are consistent with our observations that ERK activation is inhibitory to iNOS expression, and they support a possible mechanism by which estrogen may exert partial inhibitory effects on LPS-induced iNOS levels.



View larger version (28K):
[in this window]
[in a new window]
  		FIG. 8. LPS-stimulated ERK-1/-2 activation, but not p38 activation, increases with E2 pretreatment. A, Representative immunoblot of LPS-stimulated ERK-1/-2 activation after 1 h vehicle (ethanol) or E2 (1 nM) pretreatment and removal. An antibody to total ERK-1 (recognizing both phosphorylated and nonphosphorylated enzyme forms) that also cross-reacts with ERK-2 is shown as a loading control. Quantification of active ERK-1 (B) and active ERK-2 immunoreactivity (C) obtained from the densitometric analyses of immunoblot studies (n ≥ 3) as performed in panel A. Values are expressed as percent of LPS and represent the mean ± SEM of at least three separate experiments. D, Representative immunoblot of a time course of LPS-stimulated p38 activation in BV-2 microglial cells after 1 h vehicle (ethanol) or E2 (1 nM) pretreatment and removal. Grb-2 is shown as a loading control. E, Quantification of active p38 immunoreactivity obtained from the densitometric analyses of immunoblot studies (n = 3) as performed in panel D. *, P < 0.05; **, P < 0.01 vs. LPS treatment.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Our data demonstrate that E2 can exert modulatory effects on microglial inflammatory capacity via interactions with ERß. The microglial cell line BV-2 provides a unique model system for the study of the functions of ERß apart from ER{alpha} because it naturally lacks detectable expression of the latter. These cells thus make available valuable information about the ability of ERß to modulate the inflammatory status of microglia, information that is presently lacking with regard to this ER isoform. Our results show that activation of ERß leads to decreased NO production and iNOS expression in response to LPS stimulation of BV-2 microglia, as well as decreased expression of COX-2. To our knowledge, this is the first report of COX-2 as a target for ERß action, and it provides a mechanism for the decreased production of PGE2 in response to E2 observed by Vegeto et al. (19). The inhibitory effects of E2 on iNOS and COX-2 are differentially sensitive to blockade by ER antagonists; both ICI 182,780 and tamoxifen block the effects of estrogen on COX-2, whereas only tamoxifen blocks the effects of E2 on LPS-stimulated iNOS. Furthermore, we observe that E2 increases LPS-mediated ERK-1/-2 activation; a pathway that arbitrates inhibitory effects on iNOS expression, providing a potential mechanism whereby E2 decreases microglial inflammatory mediator production.

The ability of estrogen to modulate microglial inflammatory capacity via interactions with both ERß as well as ER{alpha} has important implications for the biological role of ERs in neuroinflammatory diseases and potentially provides an additional target for pharmacological interventions aimed at selectively modulating estrogen signal transduction pathways in the CNS. Although it is conceivable that the effects of estrogen on iNOS and COX-2 levels in BV-2 microglia may represent a compensatory mechanism for the lack of ER{alpha} expression, a recent article by Carswell et al. (50) would argue against such a possibility. This group demonstrated that, in a mouse model of cerebral ischemia, administration of the ERß-selective ligand DPN afforded significant neuroprotection against ischemic damage, whereas the ER{alpha}-selective ligand PPT did not. Together with the present results, the data suggest that ERß in its own right may mediate many of the important neuroprotective effects of estrogen and may indicate that ERß need not necessarily be functioning solely in a compensatory manner due to the lack of ER{alpha} expression/function.

The reason for the loss of estrogen inhibition of both iNOS and COX-2 between 2 and 4 h after E2 has been removed from the cells is not yet clear, but it may involve the down-regulation of ERß by estrogen, either at the transcriptional level as has been shown in the ovary (51), or perhaps at the protein level due to proteasome-mediated degradation (52). Estradiol also induces the down-regulation of ER{alpha}, by multiple mechanisms; this has been well described (reviewed in Ref. 53). Alternatively, this loss of response could reflect a change in the kinetics of LPS-stimulated iNOS and COX-2 protein expression, where estradiol shifts the time-response curve of LPS to the right such that it simply takes longer for LPS to exert its maximal biologic effect. When the cells are not constantly exposed to E2, the inhibitory signaling pathways elicited by E2/ERß interactions appear to function most strongly before 2–4 h; however, constant exposure to E2 (for 24 h or more before LPS stimulation) permits long-term measurable reductions in iNOS and COX-2 expression. These data suggest that, therapeutically, E2 may lack beneficial CNS effects if given at the time of onset or after inflammatory reactions in the brain begin. This is consistent with the observations of Vegeto et al. (19) and with data from human clinical trials, indicating that estrogen replacement therapy lacked beneficial cognitive effects in women who were already affected with AD (54); however, it delayed the onset of AD in healthy perimenopausal women (4, 55, 56).

Another interesting finding of the present study is that E2 causes an increase in LPS-stimulated TNF{alpha} mRNA. TNF{alpha} has been reported to exert both neurotoxic and neuroprotective effects in the brain (13, 14, 15, 16, 17); however, the significance of its increase by E2/ERß in these studies is presently unclear. In microglia expressing both ER isoforms (39, 40, 41) and other monocytic and/or lymphoid cell types (40, 57, 58), estrogen has been shown to decrease TNF{alpha} mRNA and release. It is possible that the increase in TNF{alpha} mRNA that we observe simply reflects the individual contributions of ERß vs. those of ER{alpha} and ERß combined. This explanation is plausible in light of the reports of ERß inhibiting ER{alpha} function (59, 60, 61). A recent study in retinal neurons after ischemia has demonstrated that one potential mechanism by which TNF{alpha} can exert both damaging and protective effects depends upon which TNF receptor is activated (17). In the retinal system, activation of TNFR1 was shown to mediate neuronal toxicity, whereas TNFRII activation exerted protective effects. Both neurons (62) and microglia (63) have been shown to express TNF receptors of both subtypes. It is conceivable therefore, that the TNFR whose activity prevails in a given brain region, combined with the relative expression levels of ERß vs. ER{alpha} there, will be involved in determining the detrimental or beneficial effects of TNF{alpha}.

The current model for ER antagonism by tamoxifen involves preventing AF-2 domain interaction with coactivators that initiate activation of basal transcription machinery (reviewed in Ref. 32). Because ERß is thought to possess both a weak AF-1 domain compared with that of ER{alpha}, as well as a repressive function in the A/B domain (59, 60, 61), tamoxifen blockade of ligand-sensitive AF-2 activity in ERß may be the mechanism of tamoxifen antagonism of our effect on iNOS and COX-2. Although it is possible that tamoxifen is acting as an agonist in BV-2 cells to induce the expression of an inhibitory protein that subsequently interferes with iNOS and COX-2 expression, this seems unlikely due to the fact that tamoxifen behaves as a complete antagonist of ERß/ERE interactions in cells expressing only ERß (60). However, at activator protein-1 (AP-1) sites, tamoxifen interactions with ERß can be agonistic (64), suggesting that, if tamoxifen is behaving as a partial agonist, it may be doing so via non-ERE-dependent mechanisms. Pure antiestrogens such as ICI 182,780 are thought to prevent ER{alpha}-mediated transcriptional activity by several mechanisms: 1) preventing ER{alpha} interaction with transcriptional coactivators; 2) promoting nuclear to cytoplasmic shuttling; 3) increasing proteasome-mediated degradation of the antagonist bound ER{alpha}; and 4) retarding intranuclear mobility of ER{alpha} such that it becomes tightly associated with a subnuclear compartment (reviewed in Ref. 32). On the contrary, the mechanism of pure antiestrogen blockade of ERß function is not at all clear. IC1 182,780 blocks the actions of E2 on ERE-mediated gene transcription, but it has variable effects on other nonclassical types of estrogen responses, such as those at AP-1 sites (64, 65, 66). Perhaps this implies a non-ERE-dependent mechanism of E2/ERß action on iNOS levels in microglia. It is interesting to note that both the iNOS and COX-2 promoters contain AP-1 binding sites that are important for regulating their expression (49, 67), and canonical EREs have not been found in the cloned promoter regions of either the human or murine iNOS or COX-2 genes (Watters, J. J., unpublished observations), although both genes appear to be regulated by different E2/ERß mechanisms in BV-2 microglia.

Although the lack of effect of ICI 182,780 in blocking E2-mediated reductions in iNOS levels was unexpected given the previous report by Vegeto et al. (19), our results are not inconsistent with reports of some nonclassical effects of estrogen. For example, we have previously reported in neuroblastoma cells that E2-stimulated MAPK activation, c-fos, and cAMP response element-mediated gene transcription are unable to be blocked by tamoxifen or ICI 182,780 (68, 69). Others have also reported differential effects of ER antagonists on various E2-induced endpoints, some of which are hypothesized to occur via nonclassical mechanisms (64, 65, 66). Several of these studies report that pure antiestrogens can exert agonist-like effects (65, 66, 69), especially when the effects are mediated via ERß (65). In addition, we found that E2 effectively inhibited LPS-stimulated iNOS levels at 10–14 M, whereas it required 10–13 M to significantly inhibit COX-2 levels. Together with our observations that tamoxifen and ICI 182,780 both antagonized E2-mediated inhibition of LPS-stimulated COX-2 levels, but the effects on iNOS were blocked only by tamoxifen, it is likely that the mechanisms through which estrogen is acting to exert its inhibitory effects on the expression of these two proteins are different.

Because we observed actions of ER ligands that required exposure for only 30–60 min, and ICI 182,780 did not antagonize the effects on iNOS, we hypothesized that estrogen may be acting in part, in a non-ERE-dependent manner. Furthermore, because we and others have observed that MAPKs are critical regulators of inflammatory gene expression (Fig. 7Go; and Refs. 21 , 49 , 70 , and 71) and estrogens have been shown to stimulate MAPK pathways both in microglia (18) and other cell types (Ref. 68 and reviewed in Ref. 72), we tested the ability of estrogen, via interactions with ERß, to modulate MAPK pathways in BV-2 microglia. Although E2 did not initiate p38 or ERK-1/-2 activation on its own, it did augment ERK-1/-2 activation in response to LPS stimulation. In contrast, in cells expressing both ER subtypes (N9 and primary rat cortical microglia), Bruce-Keller et al. found that E2 could, on its own, initiate ERK-1/-2 activation (18). One explanation for this difference could be that ER{alpha} is responsible for estrogen-stimulated MAPK activation. Nonetheless, our observation that E2 increases LPS-stimulated ERK activation is commensurate with decreased iNOS and NO levels and is remarkably consistent with the model that we have proposed previously in macrophages, wherein the activation of ERK-1/-2 is inhibitory to inflammatory gene expression (20, 21).

Our working model of the interaction between E2 and LPS signaling in microglia is illustrated in Fig. 9Go. To definitively test the importance of estrogen inhibition of the ERK pathway with regard to microglial iNOS/NO production as proposed in our model, it is necessary to prevent ERK pathway activation only during the time of estrogen exposure (i.e. 1 h) because long-term interference with this signaling network alters LPS signaling in microglia (Fig. 7Go). Because the currently available MEK/ERK pathway inhibitors (U0126 and PD98059) have extremely long-lasting inhibitory effects on ERK activation even after washout of the pharmacologic agents (Ref. 21 ; and our unpublished observations), it is not possible at present to selectively inhibit ERK activation only during the 1 h of estrogen pretreatment. If the proposed model were correct, then blockade of the ERK pathway during estrogen treatment would prevent the ability of estrogen to decrease LPS-stimulated iNOS/NO production. This is an important experiment that will be necessary to validate our model.



View larger version (9K):
[in this window]
[in a new window]
  		FIG. 9. Proposed model of E2 action on LPS-stimulated iNOS production. A, In the absence of E2, LPS stimulates both inhibitory (ERK-1/-2) and stimulatory (p38) MAPK pathways. With regard to iNOS expression, the stimulatory pathway prevails; however, in the presence of agents that inhibit ERK pathway activation (such as U0126) the inhibitory nature of this pathway is revealed such that an even greater stimulatory effect of LPS on iNOS and NO production is observed. B, In the presence of E2, p38 activation in response to LPS stimulation is unaffected. However, the inhibitory ERK pathway is augmented such that now the inhibitory pathway becomes stronger, causing a decrease in the ability of LPS to induce iNOS and NO production.

 
Although it is not known whether the effects of MAPKs on the overall expression of iNOS and COX-2 are necessarily manifested at the level of gene transcription, one possible mechanism by which the MAPKs may be involved in altering iNOS and COX-2 gene expression is by modulating the function of transcription factors involved in the expression of these genes (e.g. AP-1 and cAMP response element binding protein). Very recently, another mechanism whereby MAPKs can control the expression of genes in hippocampal neurons at the protein level has been described (73). ERK-1/-2 activation can regulate the phosphorylation of translation factors, impacting their ability to translate mRNA. In these ways, ERK-1/-2 can potentially effect the expression of genes both at the transcriptional and translational levels. Additional studies are required to ascertain the exact mechanism(s) of ERK action in our model system. There are likely several mechanisms employed by E2 to decrease LPS-stimulated iNOS and COX-2 protein levels in microglia; however, certainly an effect on MAPK pathways is one possible mode of action. Because recognizable EREs have not been found in the iNOS or COX-2 promoters, interactions of ERs with other transcription factors is conceivable. Such interactions may include ERß complexing with AP-1 proteins, perhaps preventing their ability to bind their cognate response elements.

In summary, these studies indicate that ERß can mediate many of the estrogen responses reported in microglia that have been previously attributed to the actions of ER{alpha}. We have also shown that the ability of ER ligands to decrease the levels of iNOS and COX-2 proteins occur by a mechanism that is rather quick, and in a manner that is differentially sensitive to blockade by ICI 182,780, suggesting that perhaps some of the antiinflammatory effects of E2 may not be mediated by classical ER/ERE interactions. These data also support the idea that E2 can modulate an inhibitory LPS-stimulated MAPK pathway that may be important for its ability to reduce microglial inflammatory capacity.


   Acknowledgments
 
We thank Ms. Gina Accola for excellent technical support and help as well as Ms. Carly Sauter and Drs. Fern Murdoch and Michael Fritsch for providing assistance with steroid receptor primer sequences and PCR conditions. In addition, we also thank Drs. Jack Gorski, Linda Schuler, Jennifer Gutzman (University of Wisconsin, Madison, WI), and Benita Katzenellenbogen (University of Illinois, Urbana-Champaign, IL) for helpful insights, advice, and suggestions with this work. Lastly, we are grateful to Drs. Elisabetta Blasi (University of Perugia, Perugia, Italy), Gary Weisman (University of Missouri, Columbia, MO), and Paola Ricciardi-Castagnoli (University of Milano, Milano, Italy) for providing the cell lines used in this study.


   Footnotes
 
This work was supported by the Wisconsin Alumni Research Foundation, a School of Veterinary Medicine Basic Science Award, and an anonymous private donation.

Abbreviations: AD, Alzheimer’s disease; AP-1, activator protein-1; CNS, central nervous system; COX-2, cyclooxygenase-2; CSFBS, charcoal-stripped fetal bovine serum; DHT, dihydrotestosterone; DPN, diarylproprionitrile; E2, 17ß-estradiol; ER, estrogen receptor; iNOS, inducible nitric oxide synthase; LPS, lipopolysaccharide; MEK, MAPK kinase; MS, multiple sclerosis; NO, nitric oxide; PGE2, prostaglandin E2; PMA, phorbol 12-myristate 13-acetate; PPT, 4,4',4'-(4-propyl-[1H]pyrazole1,3,5-triyl)Trisphenol.

Received May 14, 2004.

Accepted for publication July 9, 2004.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Kim S, Liva SM, Dalal MA, Verity MA, Voskuhl RR 1999 Estriol ameliorates autoimmune demyelinating disease: implications for multiple sclerosis. Neurology 52:1230–1238[Abstract/Free Full Text]
  2. Sicotte NL, Liva SM, Klutch R, Pfeiffer P, Bouvier S, Odesa S, Wu TC, Voskuhl RR 2002 Treatment of multiple sclerosis with the pregnancy hormone estriol. Ann Neurol 52:421–428[CrossRef][Medline]
  3. Soldan SS, Retuerto AI, Sicotte NL, Voskuhl RR 2003 Immune modulation in multiple sclerosis patients treated with the pregnancy hormone estriol. J Immunol 171:6267–6274[Abstract/Free Full Text]
  4. Kawas C, Resnick S, Morrison A, Brookmeyer R, Corrada M, Zonderman A, Bacal C, Lingle DD, Metter E 1997 A prospective study of estrogen replacement therapy and the risk of developing Alzheimer’s disease: the Baltimore Longitudinal Study of Aging. Neurology [Erratum (1998) 51:654] 48:1517–1521
  5. LeBlanc ES, Janowsky J, Chan BK, Nelson HD 2001 Hormone replacement therapy and cognition: systematic review and meta-analysis. JAMA 285:1489–1499[Abstract/Free Full Text]
  6. Zandi PP, Carlson MC, Plassman BL, Welsh-Bohmer KA, Mayer LS, Steffens DC, Breitner JC; Cache County Memory Study Investigators 2002 Hormone replacement therapy and incidence of Alzheimer disease in older women: the Cache County Study. JAMA 288:2123–2129[Abstract/Free Full Text]
  7. Iadecola C, Zhang F, Casey R, Nagayama M, Ross ME 1997 Delayed reduction of ischemic brain injury and neurological deficits in mice lacking the inducible nitric oxide synthase gene. J Neurosci 17:9157–9164[Abstract/Free Full Text]
  8. Lecanu L, Verrecchia C, Margaill I, Boulu RG, Plotkine M 1998 iNOS contribution to the NMDA-induced excitotoxic lesion in the rat striatum. Br J Pharmacol 125:584–590[CrossRef][Medline]
  9. Parmentier-Batteur S, Bohme GA, Lerouet D, Zhou-Ding L, Beray V, Margaill I, Plotkine M 2001 Antisense oligodeoxynucleotide to inducible nitric oxide synthase protects against transient focal cerebral ischemia-induced brain injury. J Cereb Blood Flow Metabolism 21:15–21[CrossRef][Medline]
  10. Acarin L, Peluffo H, Gonzalez B, Castellano B 2002 Expression of inducible nitric oxide synthase and cyclooxygenase-2 after excitotoxic damage to the immature rat brain. J Neurosci Res 68:745–754[CrossRef][Medline]
  11. Nogawa S, Zhang F, Ross ME, Iadecola C 1997 Cyclo-oxygenase-2 gene expression in neurons contributes to ischemic brain damage. J Neurosci 17:2746–2755[Abstract/Free Full Text]
  12. Iadecola C, Niwa K, Nogawa S, Zhao X, Nagayama M, Araki E, Morham S, Ross ME 2001 Reduced susceptibility to ischemic brain injury and N-methyl-D-aspartate-mediated neurotoxicity in cyclooxygenase-2-deficient mice. Proc Natl Acad Sci USA 98:1294–1299[Abstract/Free Full Text]
  13. Carlson NG, Bacchi A, Rogers SW, Gahring LC 1998 Nicotine blocks TNF-{alpha}-mediated neuroprotection to NMDA by an {alpha}-bungarotoxin-sensitive pathway. J Neurobiol 35:29–36[CrossRef][Medline]
  14. Carlson NG, Wieggel WA, Chen J, Bacchi A, Rogers SW, Gahring LC 1999 Inflammatory cytokines IL-1{alpha}, IL-1ß, IL-6, and TNF-{alpha} impart neuroprotection to an excitotoxin through distinct pathways. J Immunol 163:3963–3968[Abstract/Free Full Text]
  15. Stoll G, Jander S, Schroeter M 2000 Cytokines in CNS disorders: neurotoxicity versus neuroprotection. J Neural Transm Suppl 59:81–89[Medline]
  16. Viel JJ, McManus DQ, Smith SS, Brewer GJ 2001 Age- and concentration-dependent neuroprotection and toxicity by TNF in cortical neurons from ß-amyloid. J Neurosci Res 64:454–465[CrossRef][Medline]
  17. Fontaine V, Mohand-Said S, Hanoteau N, Fuchs C, Pfizenmaier K, Eisel U 2002 Neurodegenerative and neuroprotective effects of tumor necrosis factor (TNF) in retinal ischemia: opposite roles of TNF receptor 1 and TNF receptor 2. J Neurosci 22:RC216
  18. Bruce-Keller AJ, Keeling JL, Keller JN, Huang FF, Camondola S, Mattson MP 2000 Antiinflammatory effects of estrogen on microglial activation. Endocrinology 141:3646–3656[Abstract/Free Full Text]
  19. Vegeto E, Bonincontro C, Pollio G, Sala A, Viappiani S, Nardi F, Brusadelli A, Viviani B, Ciana P, Maggi A 2001 Estrogen prevents the lipopolysaccharide-induced inflammatory response in microglia. J Neurosci 21:1809–1818[Abstract/Free Full Text]
  20. Aga M, Watters JJ, Pfeiffer ZA, Wiepz GJ, Sommer JA, Bertics PJ 2004 Evidence for nucleotide receptor modulation of cross talk between MAP kinase and NF-{kappa}B signaling pathways in murine RAW 264.7 macrophages. Am J Physiol Cell Physiol 286:C923–C930
  21. Watters JJ, Sommer JA, Pfeiffer ZA, Prabhu U, Guerra AN, Bertics PJ 2002 A differential role for the mitogen-activated protein kinases in lipopolysaccharide signaling: the MEK/ERK pathway is not essential for nitric oxide and interleukin 1ß production. J Biol Chem 277:9077–9087[Abstract/Free Full Text]
  22. Gudino-Cabrera G, Nieto-Sampedro M 1999 Estrogen receptor immunoreactivity in Schwann-like brain macroglia. J Neurobiol 40:458–470[CrossRef][Medline]
  23. Azcoitia I, Sierra A, Garcia-Segura LM 1999 Localization of estrogen receptor ß-immunoreactivity in astrocytes of the adult rat brain. Glia 26:260–267[CrossRef][Medline]
  24. Gudino-Cabrera G, Nieto-Sampedro M 2000 Schwann-like macroglia in adult rat brain. Glia 30:49–63[CrossRef][Medline]
  25. Jakab RL, Wong JK, Belcher SM 2001 Estrogen receptor ß immunoreactivity in differentiating cells of the developing rat cerebellum. J Comp Neurol 430:396–409[CrossRef][Medline]
  26. Blurton-Jones M, Tuszynski MH 2001 Reactive astrocytes express estrogen receptors in the injured primate brain. J Comp Neurol 433:115–123[CrossRef][Medline]
  27. Dhandapani KM, Brann DW 2002 Estrogen-astrocyte interactions: implications for neuroprotection. BMC Neurosci 3:6–12[CrossRef][Medline]
  28. Garcia-Ovejero D, Veiga S, Garcia-Segura LM, Doncarlos LL 2002 Glial expression of estrogen and androgen receptors after rat brain injury. J Comp Neurol 450:256–271[CrossRef][Medline]
  29. Vegeto E, Belcredito S, Etteri S, Ghisletti S, Brusadelli A, Meda C, Krust A, Dupont S, Ciana P, Chambon P, Maggi A 2003 Estrogen receptor-{alpha} mediates the brain antiinflammatory activity of estradiol. Proc Natl Acad Sci USA 100:9614–9619[Abstract/Free Full Text]
  30. Blasi E, Barluzzi R, Bocchini V, Mazzolla R, Bistoni F 1990 Immortalization of murine microglial cells by a v-raf/v-myc carrying retrovirus. J Neuroimmunol 27:229–237[CrossRef][Medline]
  31. Righi M, Mori L, De Libero G, Sironi M, Biondi A, Mantovani A, Donini SD, Ricciardi-Castagnoli P 1989 Monokine production by microglial cell clones. Eur J Immunol 19:1443–1448[Medline]
  32. Smith CL, O’Malley BW 2004 Coregulator function: a key to understanding tissue specificity of selective receptor modulators. Endocr Rev 25:45–71[Abstract/Free Full Text]
  33. Watters JJ, Wilkinson CW, Dorsa DM 1996 Glucocorticoid regulation of vasopressin V1a receptors in rat forebrain. Brain Res Mol Brain Res 38:276–284[Medline]
  34. Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685[CrossRef][Medline]
  35. Morrison TB, Weis JJ, Wittwer CT 1998 Quantification of low-copy transcripts by continuous SYBR Green I monitoring during amplification. Biotechniques 24:954–962[Medline]
  36. Livak KJ, Schmittgen TD 2001 Analysis of relative gene expression data using real-time quantitative PCR and the 2(-{Delta}{Delta}C(T)) method. Methods 25:402–408[CrossRef][Medline]
  37. Hayashi T, Yamada K, Esaki T, Muto E, Chaudhuri G, Iguchi A 1998 Physiological concentrations of 17ß-estradiol inhibit the synthesis of nitric oxide synthase in macrophages via a receptor-mediated system. J Cardiovasc Pharmacol 31:292–298[CrossRef][Medline]
  38. Vegeto E, Pollio G, Ciana P, Maggi A 2000 Estrogen blocks inducible nitric oxide synthase accumulation in LPS-activated microglia cells. Exp Gerontol 35:1309–1316[CrossRef][Medline]
  39. Cuzzocrea S, Mazzon E, Sautebin L, Dugo L, Serraino I, De Sarro A, Caputi AP 2001 The protective role of endogenous estrogens in carrageenan-induced lung injury in the rat. Mol Med 7:478–487[Medline]
  40. Ito A, Buenafe AC, Matejuk A, Ito A, Buenafe AC, Matejuk A 2002 Estrogen inhibits systemic T cell expression of TNF-{alpha} and recruitment of TNF-{alpha}(+) T cells and macrophages into the CNS of mice developing experimental encephalomyelitis. Clin Immunol 102:275–282[CrossRef][Medline]
  41. Matsuda J, Vanier MT, Saito Y, Suzuki K 2001 Dramatic phenotypic improvement during pregnancy in a genetic leukodystrophy: estrogen appears to be a critical factor. Hum Mol Genet 10:2709–2715[Abstract/Free Full Text]
  42. Meyers MJ, Sun J, Carlson KE, Marriner GA, Katzenellenbogen BS, Katzenellenbogen JA 2001 Estrogen receptor-ß potency-selective ligands: structure-activity relationship studies of diarylpropionitriles and their acetylene and polar analogues. J Med Chem 44:4230–4251[CrossRef][Medline]
  43. Sun J, Baudry J, Katzenellenbogen JA, Katzenellenbogen BS 2003 Molecular basis for the subtype discrimination of the estrogen receptor-ß-selective ligand, diarylpropionitrile. Mol Endocrinol 17:247–258[Abstract/Free Full Text]
  44. Stauffer SR, Coletta CJ, Tedesco R, Nishiguchi G, Carlson K, Sun J, Katzenellenbogen BS, Katzenellenbogen JA 2000 Pyrazole ligands: structure-affinity/activity relationships and estrogen receptor-{alpha}-selective agonists. J Med Chem 43:4934–4947[CrossRef][Medline]
  45. Ajizian SJ, English BK, Meals EA 1999 Specific inhibitors of p38 and extracellular signal-regulated kinase mitogen-activated protein kinase pathways block inducible nitric oxide synthase and tumor necrosis factor accumulation in murine macrophages stimulated with lipopolysaccharide and interferon-{gamma}. J Infect Dis 179:939–944[CrossRef]