Autoantibodies to estrogen receptor α interfere with T lymphocyte homeostasis and are associated with disease activity in systemic lupus erythematosus

Authors


Abstract

Objective

Estrogens influence many physiologic processes and are also implicated in the development or progression of numerous diseases, including autoimmune disorders. Aberrations of lymphocyte homeostasis that lead to the production of multiple pathogenic autoantibodies, including autoantibodies specific to estrogen receptor (ER), have been detected in the peripheral blood of patients with systemic lupus erythematosus (SLE). This study was undertaken to assess the presence of both anti-ERα and anti-ERβ antibodies in sera from patients with SLE, to analyze the effect of these antibodies on peripheral blood T lymphocyte homeostasis, and to evaluate their role as determinants of disease pathogenesis and progression.

Methods

Anti-ER antibody serum immunoreactivity was analyzed by enzyme-linked immunosorbent assay in samples from 86 patients with SLE and 95 healthy donors. Phenotypic and functional analyses were performed by flow cytometry and Western blotting.

Results

Anti-ERα antibodies were present in 45% of the patients with SLE, whereas anti-ERβ antibodies were undetectable. In healthy donors, anti-ERα antibodies induced cell activation and consequent apoptotic cell death in resting lymphocytes as well as proliferation of anti-CD3–stimulated T lymphocytes. A significant association between anti-ERα antibody values and clinical parameters, i.e., the SLE Disease Activity Index and arthritis, was found.

Conclusion

Our data suggest that anti-ERα autoantibodies interfere with T lymphocyte homeostasis and are significantly associated with lupus disease activity.

The involvement of estrogens, which influence many physiologic processes, has been shown in the development or progression of several autoimmune disorders (1–4). There is evidence that 17β-estradiol directly modulates the development and function of immune cells, although the mechanism by which this might occur is not well understood (5, 6). The primary mechanism of 17β-estradiol activity is mediated by transcription activity of the intracellular estrogen receptors (ERs), ERα and ERβ, to produce genomic effects (7). A variety of cellular responses to physiologic concentrations of 17β-estradiol occur rapidly, within seconds to a few minutes. These rapid estrogen-mediated effects (referred to as nongenomic) are triggered through the activation of membrane-associated ER and are independent of transcription pathways and protein synthesis. These receptors are structurally similar to their intracellular counterparts, and it has been hypothesized that after ligand binding they ignite various intracellular signals (8). In particular, previous studies of cultured cell lines pointed to membrane-associated ERα as capable of inducing cell cycle progression and preventing the apoptotic cascade via activation of the MAPK/ERK and phosphatidylinositol 3-kinase pathways. By contrast, membrane-associated ERβ has been demonstrated to contribute to the activation of the apoptotic cascade via the p38/MAPK pathway (9). Recently, the expression of membrane-associated ERα in human T lymphocytes has also been suggested (10).

Systemic lupus erythematosus (SLE) is a multifactorial disease that involves genetic, environmental, and hormonal factors and occurs in women and men at a ratio of 9:1. SLE onset is most common in women of reproductive age, and the basis of this sex predisposition is not fully understood (11). The role of estrogen and of its receptors in the pathogenesis of SLE has been demonstrated by several previous studies (12–16). ERα, rather than ERβ, plays a major role in regulating autoimmunity in (NZB × NZW)F1 mice, in a lupus-like syndrome develops spontaneously, by promoting the production of pathogenic autoantibodies (17). Although the pathogenesis of SLE remains enigmatic, alterations in T lymphocyte homeostasis and the production of multiple pathogenic autoantibodies, including antibodies specific to ER (anti-ER antibodies), have been repeatedly demonstrated in the peripheral blood of patients with SLE (18–20).

Membrane-initiated anti-ER antibody effects on lymphoid cells are very likely but have not yet been investigated. Elucidation of the possible role of anti-ER antibodies in modulating lymphocyte homeostasis, via this unexplored mechanism of membrane-initiated responses, could open up new perspectives in the understanding of the pathogenesis of autoimmune diseases that display a significant sex disparity. Regulation of T cell proliferation and apoptosis is crucial in maintaining immune homeostasis and in avoiding unregulated clonal expansion of autoreactive immune cells (21). Therefore, we investigated the presence of anti-ER antibodies in sera from patients with SLE and analyzed their role in peripheral T lymphocyte homeostasis. In particular, we investigated the ability of anti-ER antibodies to activate membrane-associated ER and to modulate T cell apoptosis and proliferation. We also assessed the possible relationship between the presence of anti-ER antibodies and the clinical features of the disease.

PATIENTS AND METHODS

Patients and biologic samples.

We analyzed sera from 86 consecutive patients with SLE (70 women and 16 men with a median age of 40 years [range 19–71 years]) diagnosed according to the American College of Rheumatology (ACR) revised criteria (22). Disease activity categories were defined on the basis of the SLE Disease Activity Index (SLEDAI) scores (23), where a SLEDAI score of 0 = no activity, SLEDAI scores of 1–5 = mild activity, SLEDAI scores of 6–10 = moderate activity, SLEDAI scores of 11–19 = high activity, and a SLEDAI score of 20 = very high activity. Eighty-three percent of the patients with SLE were receiving glucocorticoids, 48% were receiving hydroxychloroquine, and 37% were receiving immunosuppressive drugs (azathioprine, cyclophosphamide, cyclosporin A, methotrexate, or mycophenolate mofetil); 13% of the patients were not treated.

In addition, we analyzed sera from 30 consecutive patients with a diagnosis of rheumatoid arthritis (RA) according to the criteria of the ACR (24) and from 22 consecutive patients with a diagnosis of Behçet's disease (BD) according to the International Study Group for BD criteria (25), as disease controls. All of the patients attended the Rheumatology Division of “Sapienza” University. Informed consent was obtained from each patient, and the local ethics committee approved the study. The control group consisted of 95 age- and sex-matched healthy donors (86 women and 9 men with a median age of 38 years [range 21–66 years]). Sera were obtained by standard methods and stored at −20°C until used. Serum samples from 12 of the 86 SLE patients were tested at 2 time points, with an interval of 1 year between tests. For phenotypic and functional analyses, blood samples were obtained from 10 randomly selected healthy donors.

Enzyme-linked immunosorbent assay (ELISA).

An ELISA was developed essentially as previously described (26). Briefly, polystyrene plates (MaxiSorp; Nunc) were coated with the antigen (2 μg/well ERα and ERβ) (Sigma) in 0.05M NaHCO3 buffer, pH 9.5, and incubated overnight at 4°C. Plates were blocked with 100 μl/well of 3% milk for 1 hour at 37°C. Human sera were diluted in phosphate buffered saline (PBS)–Tween (PBST) and 1% milk (1:100 for total IgG and 1:50 for IgM and IgA), 100 μl per well. Peroxidase-conjugated goat anti-human IgG (Bio-Rad), anti-human IgA (Sigma), and anti-human IgM (ICN Biomedicals) were diluted in PBST containing 1% milk (1:3,000, 1:3,000, and 1:500, respectively) and incubated for 1 hour at room temperature. O-phenylenediamine dihydrochloride (Sigma) was used as a substrate, and the optical density (OD) was measured at 490 nm. Three SD above the mean OD reading in the healthy donors was considered the cutoff level for positive reactions. All assays were performed in quadruplicate. Data are presented as the mean OD corrected for background (wells without coated antigen). The results of unknown samples on the plate were accepted if internal controls (2 serum samples, 1 positive and 1 negative) had an absorbance reading between 10% less than and 10% greater than the mean of previous readings. A standard curve, using anti-ERα antibody concentrations ranging from 0.01 to 0.1 mg/ml, was constructed to quantify the serum antibody concentration.

To evaluate the antibody specificity, the serum pool obtained from patients with SLE who were positive for anti-ERα antibodies was diluted 1:100 in PBST and incubated overnight at room temperature in the presence of 40 μg/ml of ERα (Sigma), according to the method described by Huang and colleagues (27). As a negative control, the serum pool was preincubated with 40 μg/ml of bovine serum albumin (BSA; Sigma).

Purification of specific autoantibodies from patient sera.

Antibody purification was performed as previously described (26). ERα or BSA (50 μg; Sigma) was spotted onto a nitrocellulose filter and incubated with sera from 10 SLE patients that had OD >0.5 by ELISA for anti-ERα antibody purification and with sera from 10 healthy donors for anti-BSA antibody purification (negative control). After washing with PBST, the antibodies were eluted with 100 mM glycine, pH 2.5, and mixed for 10 minutes. The eluted antibodies were immediately neutralized with 1M Tris HCl, pH 8, and dialyzed against PBS. Endotoxin contamination of antibodies was determined by the quantitative chromogenic Limulus amebocyte cell lysate assay (QCL-1000; BioWhittaker). Antibodies from a preparation of intravenous immunoglobulin (IVIG) precipitated by saturated ammonium sulfate solution were used as an additional control. For each set of functional experiments described below, anti-ERα antibodies eluted from sera from at least 5 distinct patients were used, and we did not find any variation in antibody activity from patient to patient. Purified anti-ERα antibodies and anti-BSA antibodies were analyzed by indirect immunofluorescence on Crithidia luciliae (serum dilution 1:10) to evaluate possible cross-reaction with double-stranded DNA (dsDNA) antibodies.

Cell cultures and treatments for functional studies.

Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Hypaque density-gradient centrifugation (Lympholyte-H; Cedarlane). Separation of untouched T cells from PBMCs was performed by immunomagnetic-based depletion of non-T cells using the Pan T Cell Isolation Kit II (Miltenyi Biotec). Cells were cultured in RPMI 1640 medium without phenol red (Gibco BRL) supplemented with 10% charcoal-stripped fetal bovine serum (EuroClone), 2 mM glutamine (Sigma), and 50 μg/ml gentamycin (Sigma), and treated with 50 μg/ml of human anti-ERα antibodies, IVIG, or anti-BSA antibodies. We selected this antibody concentration on the basis of the standard curve, described above, showing a mean ± SD anti-ERα antibody concentration of 50 ± 20 μg/ml in sera from SLE patients. For ERK activation, the dose-response curve was constructed by incubating T lymphocytes with serial dilutions of anti-ERα antibodies (5–50 μg/ml) for 30 minutes. For cell proliferation, PBMCs were labeled with 10 μM carboxyfluorescein succinimidyl ester (CFSE) according to the recommendations of the manufacturer (Molecular Probes). Cells were then treated as described above in the presence or absence of plate-bound anti-CD3 monoclonal antibody (mAb) (clone UCHT1; 4 μg/ml) (R&D Systems) for 72 hours. The human MCF-7 (ER-positive) and MDA-MB-231 (ER-negative) breast cancer cell lines were cultured under standard conditions (28).

Flow cytometric analysis.

Cell surface and intracellular phenotyping were performed with combinations of mAb conjugated with fluorescein isothiocyanate (FITC), phycoerythrin, PerCP, or allophycocyanin (APC), as previously described (29). Conjugated mAb against human CD3, CD4, CD8, CD95, HLA–DR, and control mouse IgG1 (all from BD Immunocytometry Systems) and Bcl-2 (Dako) were used. CD95 ligand (CD95L, CD178) was detected using anti-CD95L mAb (BD Immunocytometry Systems) and FITC-conjugated goat anti-mouse IgG as a secondary antibody (BD Immunocytometry Systems). Human anti-ERα antibodies (1 μg per 1 × 106 cells) purified from SLE patient sera and the appropriate FITC-conjugated secondary antibody (Pierce) were used for staining membrane-associated ERα and intracellular ERα. Intracellular staining was performed after fixation with 4% paraformaldehyde and permeabilization with fluorescence-activated cell sorting permeabilizing solution (BD Immunocytometry Systems) (29). Equal amounts of IVIG or anti-BSA antibodies were used as negative controls. At 4°C, preincubation with 1 μM membrane-impermeant 17β-estradiol–BSA or an equal amount of BSA alone was also used to verify the specificity of staining.

Apoptosis was quantified using an FITC-conjugated annexin V and propidium iodide apoptosis detection kit according to the recommendations of the manufacturer (Marine Biological Laboratory). Reported data are referred to as annexin V–positive apoptotic cells. The levels of cell proliferation were quantified by monitoring the sequential loss of green fluorescence intensity of the CFSE-labeled cells in lymphocyte subsets using FITC-conjugated anti-CD3, APC-conjugated anti-CD4, and PerCP-conjugated anti-CD8 mAb (BD Immunocytometry Systems) as previously described (10). Dead cells were excluded with Sytox blue (Invitrogen) staining. Acquisition was performed on FACSCalibur and FACSAria flow cytometers (BD Immunocytometry Systems), and 50,000 events per sample were run. Data were analyzed using CellQuest Pro (BD Immunocytometry Systems) and FlowJo, version 7.2.5 (Tree Star) software. Untreated cell cultures on day 3 were considered as the starting point for the proliferation profile analyses. The division index (i.e., the average number of cell divisions that a cell in the original population has undergone), and the percent divided (i.e., the percentage of the cells in the original sample that divided) were determined for each stimulated sample.

Transmission electron microscopy (TEM) and pre-embedding immunogold labeling.

Purified T cells were incubated with human anti-ERα antibodies purified from SLE patient sera or IVIG for 1 hour at 4°C, and TEM was performed as previously described (30). Sections were observed with a Philips 208 electron microscope at 80 kV.

Protein membrane purification.

Cell surface membrane proteins were purified from T lymphocytes using the Pierce Cell Surface Protein Isolation Kit, according to the recommendations of the manufacturer and as previously described (10).

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting.

T lymphocytes were lysed in radioimmunoprecipitation assay buffer (100 mM Tris HCl, pH 8, 150 mM NaCl, 1% Triton X-100, 1 mM MgCl, and 100 mM sodium orthovanadate) in the presence of Complete protease inhibitor mixture (Sigma). Protein content was determined by the Bradford assay (Bio-Rad). The level of p-ERK was analyzed by Western blotting as previously described (10). Lymphocyte lysate (25 μg) was loaded in SDS-PAGE, and Western blotting was performed using anti–p-ERK-1/2 mAb (1:1,000 dilution; Cell Signaling Technology). To ensure the presence of equal amounts of protein, the membranes were reprobed with a mouse anti–β-tubulin mAb (Amersham). Protein expression was quantified by densitometric analysis of the autoradiograms (GS-700 Imaging Densitometer; Bio-Rad). Cell surface proteins (25 μg/lane) were run in 10% SDS-PAGE, and Western blotting was performed using human anti-ERα antibodies or anti-BSA antibodies. Peroxidase-conjugated goat anti-human IgG (Bio-Rad) was used as the secondary antibody, and the reaction was developed using 3,3′-diaminobenzidine (Sigma).

Statistical analysis.

Statistical analyses were performed using the 2-tailed Mann-Whitney U test, Spearman's rank correlation analysis, and the chi-square test. P values less than 0.05 were considered significant. Flow cytometry data were evaluated using the Kolmogorov-Smirnov test (31) according to the CellQuest Pro software guide (BD Immunocytometry Systems), and a D:s ratio ≥15 was accepted as significant in the experimental condition used. The D:s value reflects the index similarity for 2 histograms, such that a higher D:s value indicates a less similar distribution. The sample size of the cross-sectional study was estimated to be able to assess a difference in the proportion of subjects with antibodies between patients with SLEDAI scores of ≤5 (no or mild disease activity) and patients with SLEDAI scores of >5 (moderate or high disease activity), given the following conditions: 1) a prevalence of patients with SLEDAI scores of >5 of 0.25, corresponding to an odds = 0.33 (i.e., 1 patient with SLEDAI scores of >5 for every 3 patients with SLEDAI scores of ≤5, based on our clinical cohort); 2) a prevalence of anti-ER antibodies in patients with SLEDAI scores of ≤5 of 0.35, based on the results of a pilot study performed previously (Colasanti T, et al: unpublished observations); 3) a minimum expected difference that we wanted to be able to assess between the 2 groups of 0.35 (corresponding to a doubling in the prevalence of anti-ER antibodies in patients with SLEDAI scores of >5 versus patients with SLEDAI scores of ≤5); 4) a significance level of alpha = 0.05 by chi-square test; and 5) a power of 1 − β = 0.80. Under these conditions, a total sample size of 82 patients was required, consisting of 61 patients with SLEDAI scores of ≤5 and 21 patients with SLEDAI scores of >5. To take into account possible differences from the expected values in the prevalence of anti-ER antibodies and in the ratio of patients with SLEDAI scores of >5 to those with SLEDAI scores of ≤5, a total of 86 patients were considered.

RESULTS

Serum anti-ERα antibodies in patients with SLE.

We analyzed serum IgG immunoreactivity to ERα and ERβ by ELISA. We detected anti-ERα antibodies in 39 (45%) of the 86 patients with SLE, whereas no anti-ERα antibodies were found in sera from healthy donors (Figure 1A). In order to assess the specificity of antibody detection in patients with SLE, sera from patients with RA and from patients with BD were also analyzed. One (3.3%) of the 30 patients with RA and 1 (4.5%) of the 22 patients with BD were positive for anti-ERα antibodies (Figure 1A). Preabsorption of the pooled positive sera (from 39 patients with SLE) with ERα completely inhibited the antibody immunoreactivity, thus confirming the specificity of the ELISA (data not shown). We did not detect IgG specific to ERβ in sera from any of the groups of patients studied or in sera from healthy donors, nor did we detect anti-ERα IgA or IgM in sera from patients with SLE or in sera from healthy donors (data not shown).

Figure 1.

Serum anti–estrogen receptor α (anti-ERα) antibodies (Abs) in patients and healthy donors (HD). A, Anti-ERα antibodies in patients with systemic lupus erythematosus (SLE), patients with rheumatoid arthritis (RA), patients with Behçet's disease (BD), and sex- and age-matched healthy donors. Samples were considered positive if the OD at 490 nm was higher than the cutoff value of an OD at 490 nm of 0.2 (broken line). The cutoff value was defined as 3 SD above the mean OD at 490 nm in healthy donors. Circles represent individual samples. P < 0.0001, mean OD at 490 nm in patients with SLE versus all other groups. B, Anti-ERα antibodies in patients with SLE Disease Activity Index (SLEDAI) scores of ≤5 and patients with SLEDAI scores of >5. Data are presented as box plots, where the boxes represent the 25th to 75th percentiles, the lines within the boxes represent the median, and the whiskers represent the 10th and 90th percentiles. A significant difference in the presence of serum anti-ERα antibodies between patients with moderate or high disease activity and those with no or mild disease activity was found. C, Correlation between changes in SLEDAI scores and changes in anti-ERα antibody levels in 12 of the patients with SLE.

In order to ascertain whether the presence of anti-ERα antibodies plays a role in the clinical course of the disease, we then divided the patients with SLE into subgroups according to the presence of serum anti-ERα antibodies and examined the epidemiologic and clinical parameters in each group (Table 1). We found a significant association between the presence of serum anti-ERα antibodies and disease activity (P = 0.02). Consistent with this observation, Spearman's rank analysis showed a significant positive correlation between anti-ERα antibody level and SLEDAI score (r = 0.3, P = 0.03) (data not shown). When the patients were divided into groups according to SLEDAI scores (23), we observed a significant difference in the presence of serum anti-ERα antibodies between patients with moderate or high disease activity and those with no or mild disease activity (P = 0.0005) (Figure 1B). We also found a significant association between the presence of serum anti-ERα antibodies and arthritis in SLE patients (P = 0.003) (Table 1). To assess whether levels of anti-ERα antibodies varied according to disease activity, anti-ERα antibody levels were determined in serum samples obtained longitudinally from 12 of the 86 patients with SLE. Interestingly, we observed a positive correlation between changes in SLEDAI scores and changes in antibody values (r = 0.82, P = 0.006) (Figure 1C). In contrast, there was no difference in anti-ERα antibody positivity among patients on different treatment regimens (data not shown).

Table 1. Epidemiologic and clinical characteristics of the 86 SLE patients according to the presence or absence of serum anti-ERα antibodies*
 Anti-ERα IgG–positive patients (n = 39)Anti-ERα IgG–negative patients (n = 47)
  • *

    Except where indicated otherwise, values are the number (%) of patients. Statistical analyses were performed using the Mann-Whitney test for age, disease duration, and Systemic Lupus Erythematosus Disease Activity Index (SLEDAI). The chi-square test was used for all other parameters. Anti-ERα = anti–estrogen receptor α; NPSLE = neuropsychiatric SLE; APS = antiphospholipid syndrome.

  • P = 0.02 versus anti-ERα IgG–positive patients.

  • P = 0.003 versus anti-ERα IgG–positive patients.

Age, median (range) years39 (26–66)41 (19–71)
Sex, no. of men/no. of women8/318/39
Disease duration, median (range) years7 (0.6–23.6)7.7 (0–22.5)
SLEDAI score, median (range)5 (0–18)2 (0–14)
Skin lesions19 (49)24 (51)
Arthritis32 (82)20 (42)
Kidney involvement11 (28)12 (25)
Serositis15 (38)10 (21)
NPSLE15 (38)15 (31)
Secondary APS11 (28)16 (34)

Because the SLEDAI takes into account the presence of anti-dsDNA autoantibodies, we also tested whether anti-ERα antibodies cross-reacted with dsDNA. We found that purified autoantibodies did not react with dsDNA, suggesting that no cross-reaction occurred (results not shown).

Activation of ERK signaling by anti-ERα antibodies.

To determine whether antibodies to the membrane-associated ERα that is expressed on peripheral T lymphocytes exert an activating signal effect on ERK, we used Western blotting to examine ERK activation after treatment with anti-ERα antibodies. ERK phosphorylation peaked at 30 minutes and was diminished after 60 minutes of treatment with anti-ERα antibodies. Treatment of cells with IVIG or anti-BSA antibodies did not have any effect on ERK activation (Figure 2A). Incubation of T lymphocytes with serial dilutions of anti-ERα antibodies resulted in a dose-response curve, supporting the notion of a specific effect of anti-ERα antibody binding (Figure 2B).

Figure 2.

Activation of ERK signaling by anti-ERα antibodies purified from SLE patient sera. A, Western blot analysis of p–ERK-1/2 in untreated, anti-ERα antibody–treated, intravenous immunoglobulin (IVIG)–treated, and anti–bovine serum albumin (anti-BSA) antibody–treated T lymphocytes. Anti-ERα antibodies, IVIG, or anti-BSA antibodies were added to the cells for 15, 30, or 60 minutes. To ensure the presence of equal amounts of protein, the membranes were reprobed with a mouse anti–β-tubulin monoclonal antibody. Blot is representative of those from 10 experiments. Bars show the mean ± SD increase in the ratio of p-ERK to β-tubulin after exposure to anti-ERα antibodies (n = 10 independent experiments). ∗ = P < 0.05 versus untreated, IVIG-treated, or anti-BSA antibody–treated cells at 30 minutes. B, Dose-response curve for treatment with anti-ERα antibodies. Cells were treated with 5 μg/ml, 10 μg/ml, or 50 μg/ml of anti-ERα antibodies for 30 minutes. Blot is representative of those from 5 experiments. Bars show the mean ± SD increase in the ratio of p-ERK to β-tubulin after exposure to increasing doses of anti-ERα antibodies (n = 5 independent experiments). ∗ = P < 0.05. See Figure 1 for other definitions.

nduction of T lymphocyte apoptosis and activation by anti-ERα antibodies.

The ERK pathway, which is rapidly activated by the membrane-associated ERα–17β-estradiol complex, has been demonstrated to play a critical role in 17β-estradiol action as a cell survival agent (32). In this context, we addressed the role of anti-ERα antibodies, evaluating their effects on lymphocyte apoptosis, activation marker expression, and proliferation. Parallel experiments were carried out with IVIG; no differences were observed between IVIG-treated and untreated cells in any of the cell types tested. As a first step, we evaluated the ability of anti-ERα antibodies to modulate the susceptibility of human resting T cells to undergo apoptosis. We observed that anti-ERα antibody treatment induced a significant increase in the percentage of apoptotic T lymphocytes, as compared to no treatment or IVIG treatment, after 48 hours of culture (a 1.4-fold increase) and after 72 hours of culture (a 1.6-fold increase) (Figure 3A). A more detailed analysis of the lymphocyte population confirmed, within both the CD4 and CD8 cell subsets, significant percentages of apoptotic cells after 48 hours of culture, with no significant difference between CD4+ and CD8+ T cells (for CD4+ T cells, mean ± SD 12 ± 2% of cells after 48 hours of treatment versus 8 ± 3% of untreated cells [P = 0.03]; for CD8+ T cells, mean ± SD 13 ± 3% of cells after 48 hours of treatment versus 7 ± 2% of untreated cells [P = 0.01]) (data not shown).

Figure 3.

Induction of apoptosis and changes in CD95, CD95L, and HLA–DR expression by anti-ERα antibodies on T lymphocytes. A, Percentage of annexin V (AV)–positive T cells. Resting lymphocytes were cultured for 24, 48, or 72 hours alone or in the presence of anti-ERα antibodies or intravenous immunoglobulin (IVIG). Bars show the mean ± SD (n = 10 independent experiments). ∗ = P < 0.05. B and C, Percentage of CD95+ cells (B) and HLA–DR+ cells (C) within the CD4+ and CD8+ populations after 48 hours of culture alone or in the presence of anti-ERα antibodies or IVIG. Bars show the mean ± SD (n = 10 independent experiments). ∗ = P < 0.05. D, Representative flow cytometry histogram plots showing the fluorescence intensity of fluorescein isothiocyanate (FITC)–conjugated anti-CD95L monoclonal antibody in untreated T cells (gray lines) and anti-ERα antibody–treated T cells (black lines) after 48 hours of culture. Isotype control staining is represented by the dotted line. See Figure 1 for other definitions.

We next evaluated the expression of relevant surface markers associated with cell activation and cell fate (i.e., HLA–DR, CD95, and CD95L molecules) on CD4+ and CD8+ T lymphocytes after 48 hours of anti-ERα antibody treatment. As shown in Figures 3B and C, anti-ERα antibodies significantly up-regulated CD95 and HLA–DR surface expression on both CD4+ T cells (P = 0.0001 for CD95; P = 0.01 for HLA–DR) and CD8+ T cells (P < 0.0001 for CD95; P = 0.002 for HLA–DR). Interestingly, under this experimental condition, the CD95L molecule, which has been shown to deliver a death signal through its receptor CD95/Fas (33), was also up-regulated at the cell surface by anti-ERα antibodies on both CD4+ and CD8+ T cells (for CD4+ cells, mean ± SD mean fluorescence intensity [MFI] 9.2 ± 0.5 versus 5.9 ± 0.2 in untreated cells, D:s >15; for CD8+ cells, mean ± SD MFI 9.5 ± 0.3 versus 5.8 ± 0.4 in untreated cells, D:s >15) (Figure 3D).

In this set of experiments, the expression of Bcl-2, a key molecule in the inhibition of the apoptotic cell death pathway (34), was also assessed. Flow cytometric quantification of the Bcl-2 molecule showed a slight down-regulation of Bcl-2 expression after 48 hours of anti-ERα antibody treatment in T lymphocytes (for CD4+ cells, mean ± SD MFI 57 ± 3 versus 69 ± 4 in untreated cells, D:s <15; for CD8+ cells, mean ± SD MFI 55 ± 1 versus 66 ± 2 in untreated cells, D:s <15) (data not shown). These results suggest that anti-ERα antibodies induced apoptosis of resting T lymphocytes mainly by the death receptor (i.e., CD95/CD95L) apoptotic pathway.

nduction of the proliferation of activated T lymphocytes by anti-ERα antibodies.

Next, we evaluated T lymphocyte proliferation after staining resting or anti-CD3–activated PBMCs with CFSE, a cytoplasmic dye whose fluorescence intensity decreases as it is partitioned among daughter cells. Resting CD4+ and CD8+ T cells did not show any increase in cell proliferation when treated with anti-ERα antibodies for up to 72 hours (data not shown). In contrast, anti-ERα antibodies significantly increased CD4+ and CD8+ T lymphocyte proliferation in response to anti-CD3 mAb (Table 2). Treatment with IVIG did not have any effect on lymphocyte function.

Table 2. Anti-ERα antibody–induced proliferation of CD4+ and CD8+ T cells*
TreatmentDivision index% divided
CD4+CD8+CD4+CD8+
  • *

    Values are the mean ± SD (n = 10 independent experiments). Peripheral blood mononuclear cells were cultured for 72 hours with the indicated treatments. The division index refers to the average number of cell divisions that a cell in the original population has undergone. The percent divided refers to the percentage of the cells in the original sample that divided. Anti-ERα = anti–estrogen receptor α; mAb = monoclonal antibody.

  • P < 0.001 versus untreated and intravenous immunoglobulin (IVIG)–treated cells.

None000.11 ± 0.030.05 ± 0.05
Anti-CD3 mAb0.10 ± 0.020.4 ± 0.0215 ± 130 ± 1.5
Anti-CD3 mAb and anti-ERα antibodies0.55 ± 0.010.75 ± 0.0150 ± 456 ± 2
Anti-CD3 mAb and IVIG0.16 ± 0.010.4 ± 0.0219 ± 1.534 ± 2

nti-ERα antibodies recognize membrane-associated ERα on lymphocytes.

To test the ability of anti-ERα antibodies purified from the sera of SLE patients to recognize membrane-associated ERα on human peripheral T lymphocytes, flow cytometric analysis was performed. As shown in Figure 4A, positive ERα signals were detected in both the CD4+ and CD8+ T cell subsets. The mean ± SD MFI ratios (i.e., the MFI obtained using specific antibodies divided by the MFI obtained using IVIG) were 2.6 ± 0.2 and 2.4 ± 0.1 for CD4+ and CD8+ T cells, respectively (D:s ratio ≥15 in both cases). Preincubation of T lymphocytes with 17β-estradiol–BSA significantly decreased membrane-associated ERα staining, confirming antibody specificity (Figure 4A). Staining with anti-BSA antibodies or BSA alone did not differ from that obtained with IVIG (results not shown). The breast cancer cell lines MDA-MB-231 and MCF-7 were used as negative and positive controls, respectively, for antibody binding (Figure 4B).

Figure 4.

Anti-ERα antibodies from SLE patient sera recognize cell surface membrane–associated ERα expression. A, Representative flow cytometry histogram plots showing the fluorescence intensity of fluorescein isothiocyanate (FITC)–labeled anti-ERα antibodies compared to that of intravenous immunoglobulin in peripheral T lymphocytes (A) and in human MCF-7 (ER-positive) and MDA-MB-231 (ER-negative) cancer cell lines (B). Dotted lines represent control staining; solid lines represent anti-ERα–labeled cells. Preincubation with 17β-estradiol–bovine serum albumin (BSA) significantly decreased membrane-associated ERα staining (gray line in A). C, Binding of anti-ERα antibodies to membrane-associated ERα on T lymphocytes as detected by transmission electron microscopy. Note the presence of distributed gold particles on the external leaflet of the plasma membrane (arrows). Inset, Higher-magnification view of a labeled microvillous structure. D, Purified cell membrane proteins analyzed by Western blotting with anti-BSA antibodies (lane 1) and anti-ERα antibodies (lane 2). MW markers are shown on the left. See Figure 1 for other definitions.

Binding of the anti-ERα antibody to membrane-associated ERα on T lymphocytes was also evaluated by TEM, by means of pre-embedding immunogold labeling. A well-defined presence of gold particles, indicating the expression of membrane-associated ERα–specific epitopes, was observed (Figure 4C). No aspecific labeling was detected in the presence of IVIG (results not shown). Finally, we tested anti-ERα antibody immunoreactivity by Western blotting using cell surface membrane proteins as antigen. A single band of 46 kd (membrane-associated ERα) was detected (Figure 4D), further supporting the specificity of the antibody binding with the 46-kd membrane isoform of ERα. The anti-ERα antibodies bound the cytoplasmic form of ERα but only after lymphocyte permeabilization as detected by flow cytometry (data not shown).

DISCUSSION

In this study, we analyzed the presence of autoantibodies specific to ERα and/or ERβ in sera from patients with SLE. We observed that anti-ERα antibodies were present in 45% of the patients with SLE, whereas anti-ERβ antibodies were not present in any of the patients with SLE. Analysis of the effects of anti-ERα antibodies on T lymphocyte homeostasis indicated that these antibodies induced cell activation and consequent apoptotic cell death in resting lymphocytes. Conversely, they increased the proliferation of anti-CD3–stimulated T cells. Importantly, a significant association between anti-ERα antibody levels and clinical parameters (i.e., SLEDAI and arthritis) was found.

Low levels of naturally occurring antibodies to ER have previously been detected in human sera (35). These antibodies behaved like potent estrogens in cell cultures, and their biologic activity was neutralized by ICI 164,384, a pure steroid antagonist of 17β-estradiol (36, 37). Additionally, Kelly and Vertosick (20) observed that anti-ER antibody levels were significantly higher in sera from patients with SLE than in sera from age- and sex-matched controls.

We found that affinity-purified anti-ERα antibodies from SLE patient sera recognized membrane-associated ERα on both CD4+ and CD8+ T lymphocytes. The inhibition of membrane-associated ERα staining observed after preincubation of T lymphocytes with 17β-estradiol–BSA excluded the possibility that anti-ERα antibodies cross-reacted with nonspecific membrane proteins. To date, the precise location of membrane-associated ER is a matter of debate, i.e., it has not been determined whether the membrane-associated ER binding site is on the inner or outer leaflet of the plasma membrane. Several investigators, using a variety of ER-specific antibodies in intact, nonpermeabilized cells, have hypothesized that the expression of ER occurs at the external leaflet of the cell membrane (38–45). Conversely, other research groups have pointed to the cytoplasmic tail of membrane-associated ER as a key functional domain (8). However, in our experience, we cannot exclude either of these hypotheses, i.e., the existence of an extracellular binding domain or the possibility that anti-ERα antibodies enter the cells and bind ER localized at the intracellular side of the plasma membrane. In fact, since the first description of anti-RNP antibody entrance into a subset of T lymphocytes, several reports have described similar findings for other autoantibodies in different systems (46).

Emerging data support clear roles for membrane-associated ERα in modulating many aspects of cell biology (8). However, the role of membrane-associated ERα in regulating immune functions is still scarcely explored. We recently demonstrated that 17β-estradiol–BSA, a membrane-impermeant form of 17β-estradiol, determined ERK signaling activation and induced proliferation of preactivated T lymphocytes (10). In the present study, we observed that anti-ERα antibodies behave as true agonists based on their ability to activate ERK signaling. This result is intriguing in light of the finding that the ERK/MAPK pathway plays a major role in the selection, differentiation, and maturation of T cells and is therefore profoundly involved in the induction and maintenance of T cell anergy and self-tolerance (47). Moreover, it has been suggested that the persistent activation of the ERK pathway may result in cell apoptosis, which is dysregulated in active SLE (48). Anti-ERα antibodies, playing a role in the dynamic activation of the ERK pathway, may consequently modulate lymphocyte apoptosis.

The balance between cell apoptosis and peripheral proliferation plays a fundamental role in the maintenance of immune system homeostasis (21, 49). In this study, we demonstrated that anti-ERα antibodies can induce apoptosis of resting T lymphocytes per se, possibly by an up-regulation of CD95/CD95L cell surface expression. However, we cannot exclude the possibility that anti-ERα antibodies might also induce complement-mediated cytotoxicity.

The proapoptotic effect of anti-ERα antibodies on T lymphocytes may contribute to the release of nuclear material in the circulation that can represent an important source of autoantigens if not removed in time. In this regard, an impaired clearance of dying cells was found in SLE patients, in whom accumulation of nuclear autoantigens may stimulate the immune system to produce autoantibodies (50, 51). Our data also showed a significantly increased proliferation of anti-CD3–activated T lymphocytes after treatment with anti-ERα antibodies, suggesting an additional pathogenetic role of anti-ERα antibodies that might contribute to the autoreactive T lymphocyte expansion. A large body of evidence has suggested that the preponderance of SLE in women is due, at least in part, to estrogen that, acting through the ER, enhances T cell activation (13, 14).

In this study, we extended current knowledge about the role of ERα in the pathogenesis of SLE, providing evidence that anti-ERα antibodies may interfere with T cell activity and have a direct clinical effect. In fact, we observed a significant association between anti-ERα antibody positivity and the SLEDAI score. After Bonferroni correction, the difference between anti-ERα antibody–positive patients and anti-ERα antibody–negative patients with regard to the SLEDAI score was no longer significant. However, the association between anti-ERα antibody positivity and the SLEDAI score is supported by the significant difference in the anti-ERα antibody titers observed between patients with low SLEDAI scores and patients with high SLEDAI scores, the significant correlation between the SLEDAI score and anti-ERα antibody titer, and the significant correlation between changes in SLEDAI scores and changes in anti-ERα antibody levels. Thus, the association between the SLEDAI score and anti-ERα antibodies seems to be reliable. In contrast, the significant difference between anti-ERα antibody–positive patients and anti-ERα antibody–negative patients with regard to the presence of arthritis held true even after Bonferroni correction, suggesting a potential role of anti-ERα antibodies in this clinical manifestation, and more generally, their usage as additional biomarkers for the monitoring of SLE.

In conclusion, we believe that our findings may contribute to the understanding of the complex pathogenetic mechanism underlying SLE. Further longitudinal and cross-sectional studies are, however, needed in order to translate these findings into clinical practice.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Malorni had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Conti, Franconi, Malorni, Pierdominici, Ortona.

Acquisition of data. Colasanti, Maselli, Sanchez, Tinari.

Analysis and interpretation of data. Alessandri, Barbati, Vacirca, Chiarotti, Giovannetti, Valesini.

Acknowledgements

We thank Dr. Angelo Gallina for technical advice.

Ancillary