Interaction of the Estrogen Receptors with the Fas Ligand Promoter in Human Monocytes¹

Autoimmune diseases are a heterogeneous group of disorders whose expression is influenced by multiple genes and environmental factors (1). Disease expression is also affected by age, gender, and reproductive status (2). Thus, autoimmune diseases such as multiple sclerosis, rheumatoid arthritis, systemic lupus erythematosus, and thyroiditis affect in a disproportionate number, more women than men, particularly during reproductive age (2). This predominance of autoimmune diseases among women suggests that sex hormones may modulate the immune response. Indeed, numerous studies have demonstrated the important role of sex hormones, mainly estrogen, on the regulation of immune function. Women tend to mount higher humoral and cell-mediated immunity in response to infections than men (3), and these differences may in part explain their susceptibility to autoimmune diseases (4, 5). Sex hormones regulate a wide variety of biological processes, including the development of cells of the immune system (6, 7). Estrogen is of particular interest because of its growing use in hormone replacement therapy for postmenopausal women, and the evidence indicating a pivotal role in the homeostasis of bone marrow cells (8).

Apoptosis is a physiological process of programmed cell death essential for the maintenance of cell and tissue homeostasis. Several genes are involved in the regulation of apoptosis, including members of the Bcl2 family and membrane receptors of the TNF family which include Fas. Fas ligand (FasL),³ which binds to Fas, is a proapoptotic type II transmembrane protein expressed in numerous cell types including epithelial and hemopoietic cells (9). Following the binding of FasL to the Fas receptor, apoptosis is rapidly induced through an autocrine/paracrine-signaling pathway (10). As a result, the intracellular death domain in Fas binds to Fas-associated death domain-containing protein, which then recruits the precursor form of the cysteine protease, caspase-8 (FLICE) (9, 11), forming the death-inducing signaling complex. The activation of caspase-8 or caspase-10 leads to its proteolytic activation through several cleavage steps. Active caspase-8 induces downstream activation of caspase-3, -6, and -7 which ultimately leads to cell death. Sensitivity toward Fas-mediated apoptosis depends on the cell type as determined by signaling pathways downstream of the death receptor which can be modulated by different proteins such as Bcl2, FLIP, chaperones, and kinases (12, 13).

There are numerous evidences that defects in apoptosis are, at least in part, responsible for the development of autoimmune diseases. The Fas/FasL system is a major mediator of the activation-induced cell death of T lymphocytes and its physiological relevance has been demonstrated by the lymphoproliferative disorder and autoimmune phenomena that is characterized by lpr and gld mice (14). However, increased monocyte numbers have also been associated with autoimmune diseases, suggesting that the Fas/FasL system may be important not only to T cell, but also in monocyte and macrophage homeostasis (15, 16). Monocytes produced in the bone morrow circulate in the peripheral blood and migrate into tissues where they differentiate, and according to their microenvironment, will exert their influence on the proliferation and apoptosis of surrounding cells, as well as controlling local immune responses (17, 18, 19). Monocytes play a major role in initiating, maintaining, and resolving host inflammatory responses by differentiating into tissue macrophages and dendritic cells and by releasing cell-signaling molecules (20).

FasL expression by monocytes has been shown to be a potential mechanism of peripheral T cell deletion, both under normal physiological conditions and in disease states (21, 22). Human monocytes normally express both Fas and FasL on the cell surface, and contain high intracellular levels of preformed FasL. FasL-expressing monocytes have been implicated in the pathogenesis of HIV infection as effector cells in bystander cell death of T lymphocytes (23). HIV-infected macrophages express FasL and kill uninfected T lymphocytes via Fas-mediated apoptosis (24). Similarly, monocyte expression of FasL may be an important factor in the regulation of leukocyte populations at inflamed sites, either through surface-expressed FasL or by the release of soluble or secreted FasL (21, 25). Furthermore, it has been shown that endogenous activation of the Fas pathway may induce spontaneous apoptosis of peripheral blood monocytes (26). We and others have shown that estrogen may affect the number of peripheral blood monocytes (27) as well as their progenitors in the bone marrow (28, 29).

The molecular mechanism by which estrogen regulates leukocyte homeostasis is unknown. This present study has investigated whether estrogen regulates the Fas/FasL apoptotic pathway in human monocytes and macrophages. Furthermore, the estrogen receptor (ER) isoforms present in human monocytes and macrophages have been characterized and the ligand-binding site sequenced. In this study, we report the presence of two regulatory regions at the FasL promoter that responds to the estrogen-ER complex. Furthermore, we show a correlation between ER isoform and stage of monocyte differentiation and cell survival. These results show a link between an essential apoptotic pathway, the Fas/FasL system, and sex hormones in the regulation of immune cell homeostasis and opens a new vein for understanding the impact of estrogen on autoimmunity.

Estradiol was purchased from Sigma-Aldrich (St. Louis, MO). RPMI 1640 and DMEM media were obtained from Life Technologies (New York, NY). FBS was purchased from Gemini Bio-Products (Calabasas, CA).

For the detection of ERs, the anti-ERα mAb, clone 6F11 from Novocastra (Newcastle, U.K.), and the anti-ERβ rabbit polyclonal Ab (Upstate Biotechnology, Lake Placid, NY) were used. The FITC-conjugated anti-idiotypic Ab (clone 1D5) which recognizes both ERα and ERβ was used for immunofluorescence and Western blot analysis (30). A rabbit anti-β-actin Ab was purchased from Sigma-Aldrich. The mouse anti-caspase-8 mAb was purchased from Oncogene (Boston, MA) and the anti-caspase-3 rabbit polyclonal Ab was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). FasL was detected using the mAbs, clone 33 (Transduction laboratories, Lexington, KY) and clone G247-4 (BD PharMingen, San Diego, CA). For detection of Fas, the mAb clone B-10 (Santa Cruz Biotechnology) was purchased. The anti-CD14 mAb was obtained from BD Biosciences (San Jose, CA).

All cell lines were purchased from American Type Culture Collection (Manassas, VA). The human promonocytic cell line, U937, and the human breast cancer cell line, T47D, were cultured in RPMI 1640 supplemented with 10% FBS, 0.1 mM MEM nonessential amino acids, 5 mM l-glutamine, 100 μg/ml streptomycin, and 100 IU/ml penicillin (Life Technologies) at 37°C and under 5% CO₂ in humidified air. The human breast cancer cell line, MCF-7, and the ovarian cancer cell line, Hey, were maintained in DMEM supplemented with 10% FBS, 0.1 mM MEM nonessential amino acids, 5 mM l-glutamine, 100 μg/ml streptomycin, and 100 IU/ml penicillin (Life Technologies) under the above conditions.

PBMC were purified from heparinized venous blood from volunteer donors, aged 19–26 years old. Blood samples were fractionated using Lymphocyte Separation medium (ICN Pharmaceuticals, Aurora, OH) according to the manufacturer’s instructions. The mononuclear cell interface was collected and washed once in an equal volume of PBS and then centrifugated at 1200 rpm for 10 min. The pellet was then incubated with anti-CD14 MicroBeads for 30 min at 4°C. The cell suspension was then applied to a selection column type MS⁺/RS⁺ (Miltenyi Biotec, Auburn, CA) and processed in the magnetic field. Following isolation, the monocyte-positive cell fraction was washed in RPMI 1640 and then either incubated in RPMI supplemented with 10% FBS or lysed for RNA and protein extraction.

PBMC (2 × 10⁶ cells/pellet) were stained with a PE-conjugated anti-CD14 mAb for 20–30 min at 4°C. The cells were washed twice with PBS/AZ (PBS containing sodium azide) by centrifugation for 5 min at 4°C at 1000 rpm. The cells were then fixed with cold acetone for 10 min and incubated with the FITC-conjugated anti-ER Ab, clone 1D5 (1 μg/ml; Ref. 30) for 45 min at 4°C. For isotype controls, cells were incubated with either mouse IgG1-PE or mouse IgG1-FITC (Zymed Laboratories, San Francisco, CA). Flow cytometry was performed using a FACScan (BD Biosciences, Palo Alto, CA) and data were analyzed using Lysis II software.

Total RNA and protein were prepared from cells using TRIzol reagent (Life Technologies) according to the manufacturer’s instructions. The TRIzol protocol was used to extract RNA and protein from the same cells. This method allowed us to study the same samples at both the mRNA and protein level.

Details of the characterization for the RT-PCR have been described previously (31, 32). Briefly, reverse transcription was performed using the RT-PCR kit from Amersham Pharmacia Biotech (Piscataway, NJ) according to the manufacturer’s directions. cDNA synthesis was performed with pd(N)₆ 0.2 μg and 5 μg total RNA. The primers used for amplification of rat FasL have been described previously (33) and have the following sequence: ATAGGATCCATGTTTCTGCTCCTTCCACCTACAGAAGGA-3′, downstream: 5′-ATAGAATTCTGACCAAGAGAGAGCTCAGATACGTTGAC-3′. Each PCR cycle consisted of a denaturation step at 94°C for 30 s, annealing at 52°C for 30 s, and elongation at 72°C for 1 min for 10 cycles, followed by 35 cycles modified by a cumulative 5-s increase of extension time per cycle. The size of the product was 569 bp. The Fas and FasL signals were measured by a densitometer and standardized against the β-actin signal using a digital imaging and analysis system (AlphaEase; Alpha Innotech, San Leandro, CA). The linearity of the system was determined using serial dilution of cDNA, and the regulation of dilution factor on amplified cDNA was linear (y = 2881.125x − 785.75) and the correlation coefficient was r = 0.994 (31).

The reporter plasmid pGL3-basic luciferase (Promega, Madison, WI) containing the FasL promoter (pFasL-Luc) was a kind gift from Dr. X. Liu (Department of Virus and Cancer, Danish Cancer Society, Aarhus, Denmark) (34). Mutations at the AP-1 site in the pFasL-Luc construct were introduced using the QuikChange Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA). The primers for the mutagenesis reactions were: sense primer, 5′-GAATA AAGAT GTCAG GTGTA GTGAC TTATG CCTAT AATCC cagc-3′ and antisense primer, 5′-GCTG GGATT ATAGG CATAA GTCAC TACAC CTGAC ATCTT TATTC-3′. Mutations at estrogen-recognizing element (ERE) site in the pFasL-Luc construct were performed using the following primers: sense (5′-GGTGA TCGGC ATATT AGGGT AAATG GTAGT TGTGT GTGGG G-3′), and anti-sense (5′-CCCCA CACAC AACTA C CATT TACCC TAATA TGCCG ATCAC C-3′). Underlined nucleotides represent the mutated sites compares with the wild-type FasL promoter sequence. The mutations were verified by automated sequence analysis (Keck Facility, Yale University, New Haven, CT).

Proteins were separated by SDS-PAGE using 10% polyacrylamide gels and transferred to nitrocellulose membranes. Membranes were stained with Ponceau red before the Ab incubation and the protein bands were photographed using a Kodak digital camera (Kodak, Rochester, NY). Similarly, the gel was stained with Coomassie blue and then photographed. Following immunoblotting, the intensity of the bands was normalized to the concentration of proteins present in the membranes and in the gel. The 1D Image Analysis software (Scientific Imaging Kodak, Rochester, NY) was used for this purpose. Immunoblotting was performed after blocking the membranes with 5% powdered milk in PBS containing 0.05% Tween 20. The blots were initially incubated with Ab against either FasL (mAb clone 33 at 1/1000 dilution) or Fas (mAb B-10 at 1/500 dilution) for 4 h at room temperature. Fas- or FasL-specific signals were detected using a peroxidase-conjugated horse anti-mouse secondary Ab (Vector Laboratories, Burlingame, CA) and Chemiluminescence Reagent Plus (PerkinElmer, Wellesley, MA). Because the specificity of clone 33 for FasL has been questioned, we confirmed the results using clone G247-4 from BD PharMingen. No significant differences were found between the two Abs in their ability to detect FasL. All experiments were repeated at least three times and the intensity of the signals were analyzed using a digital imaging analysis system (1D Image Analysis software). β-Actin was used as internal control, in addition to Ponseau red, to validate the amount of protein loaded onto the gels.

Isolated monocytes or U937 cells were induced to differentiate into a macrophage-like phenotype by incubation for 48–72 h with PMA at a concentration of 6 ng/ml. For estrogen treatment, cells were incubated in media without phenol red, supplemented with stripped serum in the presence or absence of 17-β estradiol. Cell viability was determined using the CellTiter 96 assay (Promega) according to the manufacturer’s instructions.

For the immunolocalization of ER, monocytes were first fixed in adhesion slides with cold methanol at −20°C for 10 min. Slides were then washed for 10 min with PBS and subsequently incubated in a humidified chamber with Abs against either ERα (clone 6F11), ERβ (rabbit polyclonal IgG), or both ERα and ERβ (clone 1D5) as previously described (35). For localization of FasL expression, cells were fixed with 4% paraformaldehyde for 15 min at room temperature, washed with PBS, and then incubated with anti-FasL mAb clone N-20 (Santa Cruz Biotechnology).

The presence of apoptotic cells was determined using Hoechst 33342 staining (Molecular Probes, Eugene, OR). Cells were incubated with 1 μg/ml Hoechst 33342 for 20 min at room temperature. Following washing with PBS, the cells were observed under a Zeiss fluorescent microscope (Zeiss, Oberkochen, Germany) and pictures captured using the OpenLab image system (Improvision, Lexiton, MA).

We first evaluated the expression of ER in human peripheral blood monocytes using FACS analysis. For this purpose, we used PE-conjugated anti-CD14 as a monocyte marker and FITC-conjugated clone 1D5 that recognizes both ERα and ERβ (30). As shown in Fig. 1,A, A1, 88% of the monocyte population was positive for ER. No specific monocyte staining was observed with the isotype control (Fig. 1 A, A2).

Because in classical target organs ER are located in the cell nuclei, we evaluated the intracellular localization of ER in monocytes by immunofluorescence. Peripheral blood monocytes, U937 cells, and differentiated U937 cells were stained with the anti-ER mAb clone 1D5. In all the studied cells, positive immunoreactivity was found in the nuclei of the cell, corresponding with the classical localization of ER (Fig. 1 B).

To determine whether human monocytes express the classical ER, we cloned and sequenced the ligand binding site of ER using an RNA transcript from human monocytes and compared it with the wild-type sequence of the human ER-ligand-binding domain (36). PCR amplification showed a similar size in the product obtained from monocytic ER and the wild-type human ER. The bands were isolated and subsequently sequenced. The DNA sequence data showed 100% homology between the ligand binding site of ER derived from human monocytes and the ligand-binding domain of wild-type human ER (Fig. 2).

To characterize the ER isoforms present in monocytes and monocyte-derived macrophages, we performed Western blot analysis using specific Abs against ERα (clone F-6) and ERβ (rabbit polyclonal). The breast cancer cell lines MCF-7 (ERα positive) and T47D (both ERα and ERβ) were used as positive controls. As shown in Fig. 2,B, the monocytic cell line U937 expressed only ERβ (Fig. 2,B, lane 1). However, when these cells underwent differentiation into macrophages following treatment with PMA, the ER expression profile changed. Thus, ERα was dominantly present in monocyte-derived macrophages, while ERβ expression decreased or was absent (Fig. 2,B, lane 2). T47D cells expressed both ERα and ERβ (Fig. 2 B, lane 3).

Once the presence of ER had been demonstrated in monocytes and monocyte-derived macrophages, we then studied their function. First, we assessed the effect of estrogen on cell growth and viability. Cell viability was determined using CellTiter 96 proliferation assay in the U937 cells. Treatment of U937 cells with estradiol decreased the number of viable cells in a time-dependent manner. Hence, 56% of the cells were alive following 24 h of treatment with estradiol. After 48 h in the presence of estradiol, 20% of cells were alive and 11% were viable after 72 h (Fig. 3,A). To test whether this effect was ER-mediated, tamoxifen, an ER antagonist, was used. As shown in Fig. 3 B, following a 24-h incubation with estradiol, the presence of tamoxifen not only blocked the effect of estrogen on U937 cell viability, but reversed it toward the induction of cell proliferation (70% increase in relation to the control, p = 0.025). Interestingly, treatment of U937 cells with tamoxifen alone induced some cell proliferation, and this may be due to its potential agonistic effects.

Monocytes undergo differentiation into tissue macrophages upon appropriate stimulation and many of their characteristics change according to their stage of differentiation and tissue location. Therefore, we tested whether the stage of monocyte differentiation and ER isoform expression would influence estrogen-mediated effects. For this, U937 cells were differentiated into macrophages by treating the cells with PMA for 72 h. The differentiated cells were then treated with estradiol and cell viability was determined. Contrary to our observations with undifferentiated U937 cells (Fig. 3), no cell death was observed in either the untreated or estrogen-treated PMA-differentiated U937 cells (data not shown). As shown in Fig. 4,A, microscopic observation of undifferentiated U937 cells treated with estradiol (10⁻⁸ M) for 24 h presented morphological changes characteristic of apoptosis (Fig. 4,Ab). These morphological changes were confirmed using Hoechst 33342 stain, a specific marker for DNA damage following apoptosis (see arrows in Fig. 4,A, b and d). However, no apoptotic cells were seen in monocyte-differentiated macrophage cultures following treatment with estrogen (Fig. 4 B).

We then evaluated the possible apoptotic pathways involved in monocyte cell death following treatment with estrogen. Using Western blot analysis, we determined caspase-8 and caspase-3 activation in monocytes and monocyte-derived macrophages. The active forms of caspase-8 (p43 and p28; Fig. 5,A) and caspase-3 (p20 and p17; Fig. 5,B) were clearly detected in undifferentiated U937 cells (monocytes) following treatment with estradiol, but not in the untreated controls. As expected, we did not observe activation of either caspase-8 or caspase-3 in PMA-differentiated U937 cells (macrophages) following treatment with estradiol (Fig. 5).

Caspase-8 is activated by membrane receptors of the TNFR superfamily, which includes Fas. The Fas/FasL system has been shown to play an important role in monocyte homeostasis. Therefore, we evaluated whether estrogen could affect monocyte cell survival by regulating Fas or FasL expression. Estrogen treatment increased FasL but not Fas protein expression in U937 cells (Fig. 6,A). Estradiol at concentrations of 10⁻⁶ and 10⁻⁸ M increased FasL expression by 80 and 60%, respectively. No differences in FasL expression were found following treatment with estradiol at concentrations of 10⁻¹⁰ or 10⁻¹² M in relation to the control (data not shown). This estrogenic effect on FasL expression was time-dependent, with the highest expression found at 24 h following treatment (Fig. 6 A).

The ER, ERα and ERβ, are transcriptional factors that induce the transcription of target genes by binding to cis-acting enhancer elements within the promoter region of such responsive genes. To determine whether the effect of estrogen on FasL expression was through direct action on its promoter region, we first looked for the presence of ERE at the FasL gene using a computerized gene homology program from the National Institutes of Health (Bethesda, MD). Motifs resembling the consensus ERE were found at the promoter region of the FasL gene located at nucleotides 743–752 bp. The FasL ERE consisted of two palindromic arms separated by 3-bp “bridge sequence”. One of the arms, GGTCA, has perfect homology to the canonical ERE while the second arm has two mismatches. A second pathway for transcriptional regulation by ERα and ERβ is through the AP-1 enhancer element. Accordingly, and as previously described (37), we recognized a complete AP-1 sequence, TTAGTCAG, located at nucleotides 1061–1053 bp of the FasL promoter region.

The presence of an ERE and the above described in vitro effect of estrogen on monocytes clearly suggests that estrogen might act directly upon the activity of the FasL gene. To test this hypothesis, we conducted transient expression assays in cultured monocytes or in the ERα-positive breast carcinoma cell line, MCF-7, by using a luciferase reporter gene containing the FasL promoter (pGL3-FasL-Luc). We chose MCF-7 cells as a model system due to their higher transfection efficiency and their high levels of ER expression. The activity of the pGL3-FasL-Luc construct was markedly increased when transfected into either monocytes or MCF7 cells and exposed to estradiol during 24 h (150% increase; Fig. 6,B). The specificity of the ER-ERE interaction in the FasL promoter was determined by adding tamoxifen to the cultures. As shown in Fig. 6 B, tamoxifen blocked the estrogen-induced FasL promoter activity. As a negative control we used a construct containing a wild type progesterone-recognizing element introduced in a SVM promoter next to the luciferase reporter gene. No change on baseline activity was detected in the control plasmid when treated with estrogen (data not shown).

Because we found two possible target regions for the ER effect at the FasL promoter, we introduced mutations either in the ERE (pGL3-FasL ERE), AP-1 (pGL3-FasL AP-1), or both regions (pGL3-FasL ERE-AP-1) of the FasL promoter. The constructs were tested first in MCF-7 cells that express ERα. Thus, mutation on the ERE segment abolished 100% FasL-induced activity following estradiol treatment compared with the wild type. The absence of a functional AP-1 also effected the overall activity of the promoter construct by significantly decreasing (∼40%) the estrogen-induced activity of FasL promoter. Double mutation for ERE and AP-1 sites completely abolished the estrogenic effect on FasL promoter, similar to what was found with the ERE single mutant (Fig. 6,C). To determine whether the effect of the AP-1 region is ER isoform-dependent, we transfected the FasL promoter constructs into ovarian cancer Hey cells expressing only ERβ (38) and evaluated the effect of estrogen on FasL promoter activity. Similar to what we found with the ERα-positive cells, the absence of a functional ERE abolished the estrogenic effect. However, in the ERβ cells, the absence of a functional AP-1 region did not alter the effect of estrogen on FasL promoter (Fig. 6 C).

We wanted to confirm whether estrogen-induced apoptosis of monocytes is indeed through the Fas pathway by blocking the activation of the pathway. Therefore, U937 cells were treated with estradiol in the presence or absence of a blocking Ab against FasL or the caspase-3 inhibitor Ac-DEVD-CHO. Cell viability was determined by CellTiter 96 after 24 h of treatment. The addition of the blocking anti-FasL Ab decreased the apoptotic effect of estradiol by 44% (p = 0.01) on the monocytic cell line. More striking was the effect of the caspase-3 inhibitor, which suppressed by 96%, estrogen-induced apoptosis (data not shown).

Estrogen is more than a sex hormone; it is a transcriptional factor regulating important biological functions in almost every tissue of the body. Although its role in the regulation of the immune system has been known for several decades (7), the target cells and the mechanisms of estrogen action are still unclear. In this study, we report the identification of two estrogen-related regulatory regions in the FasL promote: a positive ERE and AP-1 motif. Similarly, we show definitive evidence for the presence of ER in human monocytes and macrophages. Furthermore, we demonstrate the control of monocyte viability by estrogen through the regulation of the Fas/FasL system. This effect was found to be dependent on the stage of monocyte cell differentiation and ER isoform expression.

Regulation of apoptosis is an important mechanism for the control of monocyte cell numbers and thus the magnitude of an immune response to infection, wound healing, and tissue remodeling (39). The accumulation of activated monocytes is involved in pathogenesis of autoimmune diseases such as arteriosclerosis, vasculitis, and rheumatoid arthritis, demonstrating the existence of a delicate balance, where deviation to either side can have severe consequences. The Fas/FasL system is believed to play an important role in normal monocyte physiology (16). Monocytes are able to undergo apoptosis via an autocrine or paracrine mechanism that is dependent upon the expression of both Fas and FasL (40). This has been confirmed by in vitro studies which have shown that anti-Fas or anti-FasL neutralizing Abs can block monocyte spontaneous apoptosis during differentiation (40).

Mice mutated for Fas (lpr) (41) or FasL (gld) (42) display autoimmune disorders including glomerulonephritis, lymphoproliferation, autoantibody production, hypergammaglobulinemia, vasculitis, and arthritis. The majority of these disorders are associated with increased monocyte numbers (43, 44). This is supported by the work of Fukuyama et al. (46), who have shown that restoration of Fas to the T cell population in lpr mice block lymphoproliferation but not other aspects of autoimmune disease (45), suggesting that T cells are not the only cell type responsible for the development of autoimmune disease in these animals. Furthermore, mice deficient for Src homology 2 domain containing inositol polyphosphate phosphatase or heterozygotic for phosphatase and tensin homolog detected on chromosome ten (46) display an increased number of monocytes and have similar autoimmune disease as seen in the lpr and gld mice. All these data strongly indicate that the factors regulating the Fas/FasL pathway are important in the pathophysiology of autoimmunity.

Although the elucidation of the Fas-mediated cell death cascade has been widely investigated (7, 47, 48), very little is known about the regulation of FasL expression, especially by hormonal factors such as sex hormones. In a previous study, we showed that during menopause the number of circulating monocytes significantly increased in relation to the number found in premenopausal women. When estrogen replacement therapy was administrated to postmenopausal women, the number of monocytes decreased in proportion to their estrogen levels (27). In agreement with our previous study, estrogen treatment of monocytes, but not macrophages, induced an increase in FasL expression followed by apoptosis.

We found using the monocytic U937 cell line that the estrogen-induced apoptotic effect was limited to the bone marrow-derived undifferentiated monocytes, indicating that the role of estrogen in the immune system is at the developmental stage. This is in agreement with other studies showing a decrease in the bone marrow cell population following estrogen treatment. Similarly, the T cell population of the thymus decreases significantly as levels of estrogen increase (6). This effect on T cells is restricted to the thymus and to the undifferentiated double positive CD4⁺CD8⁺ cells.

Surprisingly, we show in this study that when monocytes underwent differentiation into macrophages, the apoptotic effect of estrogen was abolished; moreover, there was further proliferation. Although other intracellular mechanisms may contribute to macrophage resistance to estrogen-induced apoptosis (49), we suggest that a main factor to this effect is related to the different type of ER expressed by these cells. Perlman et al. (16) demonstrated that inhibition of Fas-FasL interaction results in the inability to delete monocytes and the development of autoimmune diseases. Our in vitro studies further confirmed that estrogen might regulate FasL expression and thus affect monocyte and macrophage homeostasis. But then, the main question is how, and when?

Many of the reported effects of estrogen on normal immune responses and autoimmune diseases have appeared contradictory, until it was realized that estrogen has a biphasic dose effect and is receptor isoform-dependent. (50). The recent identification of a second ER has opened new areas of investigation and has provided new insights into the understanding of hormonal effects (51).

Recently, it was shown that estrogen regulates FasL expression in breast and ovarian cancer cells, suggesting that a possible estrogenic effect on monocyte survival may also be mediated through the regulation of FasL expression (32, 52). However, this effect could be indirect through the expression of cytokines or growth factors. Because estrogen is known to act through its receptor and binding to specific regions at the promoter region of target genes, the presence of these elements represent an important indication for a direct effect. In this present study, we found two candidates at the FasL promoter: the classical ERE and the newly ER-related region AP-1 motif. Using different constructs, we were able to define the active regions responsible for the estrogenic effect. Surprisingly, the regions are important in the context of the ER isoform. Although the ERE is the main motif for both ERα and ERβ, the AP-1 region is important only in the presence of ERα. This is relevant in view of the different effects found with estrogen in monocytes and monocyte-derived macrophages, and the type of ER expressed by these cells. Monocytes express only ERβ while monocyte-derived macrophages express mainly ERα. Because the AP-1 motif is associated with the Jun/Fos transcriptional factors which respond to cell survival and proliferation, the interaction of ERα with this region secures the survival of the cell.

Differences in structure and distribution between ERα and ERβ suggest different biological roles of the two receptor isoforms. Neoplastic transformation is associated with changes in the type of receptor expressed and thus the response to estrogen (38). In a recently published study using the ER knockout mouse, it was shown that ERα is mandatory for full development of the thymus and ERβ is required for estrogen-mediated thymic cortex atrophy and thymocyte phenotype shift seen in the female mouse (53).

In summary, the results presented in this study demonstrate the molecular mechanism by which estrogen may regulate immune cell, particularly myeloid cell, homeostasis. Furthermore, these data exemplify how complex is the effect of estrogen on the immune system and the importance, when we study estrogen action, of taking into consideration the status of cell differentiation and receptor isoform.

We thank Dr. Yehudith Zaltsman for her assistance with the sequence of the ER.