Estradiol enhances primary antigen-specific CD4 T cell responses and Th1 development in vivo. Essential role of estrogen receptor α expression in hematopoietic cells

Authors

  • Arlette Maret,

    1. INSERM U397, Institut Louis Bugnard, CHU Rangueil, Toulouse, France
    Search for more papers by this author
    • The first three authors contributed equally to this work.

  • Jérôme D. Coudert,

    1. Institut National de la Santé et de la Recherche Médicale (INSERM) U563, Centre de Physiopathologie de Toulouse Purpan, Institut Claude de Préval, Hôpital Purpan, Toulouse, France
    Search for more papers by this author
  • Lucile Garidou,

    1. Institut National de la Santé et de la Recherche Médicale (INSERM) U563, Centre de Physiopathologie de Toulouse Purpan, Institut Claude de Préval, Hôpital Purpan, Toulouse, France
    Search for more papers by this author
  • Gilles Foucras,

    1. Institut National de la Santé et de la Recherche Médicale (INSERM) U563, Centre de Physiopathologie de Toulouse Purpan, Institut Claude de Préval, Hôpital Purpan, Toulouse, France
    Search for more papers by this author
  • Pierre Gourdy,

    1. INSERM U397, Institut Louis Bugnard, CHU Rangueil, Toulouse, France
    Search for more papers by this author
  • Andrée Krust,

    1. Institut de Génétique et de Biologie Moléculaire et Cellulaire (IGBMC), CNRS-INSERM-ULP-Collège de France, Illkirch, France
    Search for more papers by this author
  • Sonia Dupont,

    1. Institut de Génétique et de Biologie Moléculaire et Cellulaire (IGBMC), CNRS-INSERM-ULP-Collège de France, Illkirch, France
    Search for more papers by this author
  • Pierre Chambon,

    1. Institut de Génétique et de Biologie Moléculaire et Cellulaire (IGBMC), CNRS-INSERM-ULP-Collège de France, Illkirch, France
    Search for more papers by this author
  • Philippe Druet,

    1. Institut National de la Santé et de la Recherche Médicale (INSERM) U563, Centre de Physiopathologie de Toulouse Purpan, Institut Claude de Préval, Hôpital Purpan, Toulouse, France
    Search for more papers by this author
  • Francis Bayard,

    1. INSERM U397, Institut Louis Bugnard, CHU Rangueil, Toulouse, France
    Search for more papers by this author
  • Jean-Charles Guéry

    Corresponding author
    1. Institut National de la Santé et de la Recherche Médicale (INSERM) U563, Centre de Physiopathologie de Toulouse Purpan, Institut Claude de Préval, Hôpital Purpan, Toulouse, France
    • INSERM U563, Hôpital Purpan, Place du Dr. Baylac, F-31 059 Toulouse, France Fax: +33-5-6177-9291
    Search for more papers by this author

Abstract

It is widely accepted that females have superior immune responses than males, but the ways by which sex hormones may enhance T cell responses are still poorly understood. In the present study,we analyzed the effect of estrogens on CD4 T cell activation and differentiation after immunization with exogenous antigens. We show that administration of low doses of 17ß-estradiol (E2) to castrated female mice results in a striking increase of antigen-specific CD4 T cell responses and in the selective development of IFN-γ-producing cells. Quantitative assessment of the frequency of T cells bearing a public TCR ß chain CDR3 motif demonstrated that the clonal size of primary antigen-specific CD4 T cells was dramatically increased in immune lymph nodes from E2-treated mice. By usingmice with disrupted estrogen receptor (ER) α or ß genes, we show that ERα, but not ERβ, was necessary for the enhanced E2-driven Th1 cell responsiveness. Furthermore, ERα expressionin hematopoietic cells was essential, since E2 effects on Th1 responses were only observed in mice reconstituted with bone marrow cells from ERα+/+, but not ERα-deficient mice. These results demonstrate that estrogen administration promotes strong antigen-specific Th1 cell responses in a mechanism that requires functional expression of ERα in hematopoietic cells.

Abbreviations:
E2:

17ß-estradiol

HEL:

Hen egg-white lysozyme

ER:

Estrogen receptor

Ovx:

Ovariectomized

1 Introduction

Naive CD4 T cell activation requires the engagement of the TCR with antigenic peptides presented by MHC class II molecules on the surface of professional APC, e.g. dendritic cells 1. Ligand-TCR interactions induce a cascade of signal transduction events resulting in the proliferation and differentiation of CD4 T lymphocytes into effector cells that differ in their cytokine secretion profiles 2. Mouse T helper 1 (Th1) cells produce IL-2, IFN-γ, and TNF-β and are prominent mediators of cell-mediated immunity, inflammatory responses and are implicated in organ-specific autoimmune diseases 3. Th2 cells selectively secrete IL-4, IL-5, IL-10 and IL-13, stimulate humoral immunity and are responsible for allergic responses 4. While it is clear that the choice of differentiation into either Th1 or Th2 cells is strongly influenced by the predominance of IL-12 or IL-4 in the microenvironment 5, the polarization of Th cells can be influenced by other factors 6, including the hormonal status of the host 7.

It is generally accepted that females develop superior immune responses than males, that may contribute not only to a more effective immune surveillance and increased longevity, but also to a propensity for developing autoimmunity 8, 9. Although it has been suggested that sex steroid hormones, such as estrogens, may contribute to susceptibility or resistance to autoimmune diseases by influencing the development and function of pathogenic T cells, their mechanisms of action on the immune system are still poorly understood. In vitro analysis of the direct effect of estrogens on cytokine secretion by human T cell clones have shown that 17β-estradiol (E2) can modulate both pro- and anti-inflammatory cytokine synthesis by CD4 T cells depending on the dose of hormone 10. Therefore, it has been hypothesized that the response to estrogens might be biphasic, with high levels of estrogens inhibiting cell-mediated immune function by promoting Th2 immunity, whereas low levels increase Th1 immunity and susceptibility to cell-mediated autoimmune diseases 11. However, despite the observations that female mice tended to have increased T cell responses 12 and IFN-γ-production 13, 14 following in vivo antigenic challenge, and that E2 could regulate the IFN-γ promoter 15 and IFN-γ protein secretion 10, 16 by mouse and human T cells in vitro, direct evidence that E2 could enhance primary antigen-specific CD4 T cell expansion and Th1 differentiation through nuclear estrogen receptor (ER) in vivo were still lacking.

In the present study, we investigate the effect of continuous administration of low-dose E2, ranging from cycle substitutive to gestational levels, on the polarization of Th cell responses in mice following protein antigen immunization. Our data clearly establish that E2 dramatically enhances primary antigen-specific CD4 T cell responses and promotes strong Th1 cell development. Furthermore, we provide direct evidence that this effect required functional expression of the nuclear ERα but not ERß subtype in bone marrow derived cells.

2 Results

2.1 Low dose E2 administration enhances primary antigen-specific T cell proliferative responses and IFN-γ production

We have examined the effect of E2 substitution on CD4 T cell responses in mice. Female B6 mice, castrated at 4 weeks of age, were implanted with 0.1-mg E2 pellets. This treatment has been shown to result in the maintenance of a constant serum E2 concentration of 60–100 pg/ml 17 corresponding to plasma levels found during late pregnancy in the mouse 18. At 8 weeks of age, mice were immunized with OVA-CFA and the polarization of the antigen-specific T cell response in the draining lymph nodes tested 8 to 9 days later. The data in Fig. 1A–C show that E2 administration in ovariectomized (Ovx) mice resulted in a dramatic up-regulation of the T cell proliferative response (Fig. 1A) and IFN-γ-production (Fig. 1B) in response to titrated amounts of OVA, as compared to the untreated control group or sham operated mice (not shown). This was associated with a strong increase in IL-2 production (Fig. 1C). The Th2 signature cytokine IL-4 was undetectable in culture supernatants from all groups (not shown). Since CFA is a strong Th1-polarizing adjuvant, we next evaluated whether E2 administration could enhance Th1 responsiveness in mice primed with antigen in a more neutral adjuvant such as IFA. As illustrated in Fig. 1D–F, the antigen-specific T cell proliferative response, as well as the production of the type 1 cytokine IFN-γ and IL-2, were dramatically exacerbated in E2-treated mice. Again, IL-4 was undetectable in all culture supernatants (not shown) probably as a consequence of the failure of the B6 strain to mount strong Th2 responses. Since immunization of antigen in IFA can induce measurable Th2 responses in BALB/c mice 19, we next evaluated the effect of E2 on Th cell development in this strain. As shown in Fig. 2A, E2 administration in BALB/c mice resulted in an enhanced IFN-γ production by antigen-specific T cells. Interestingly, this increased Th1 response was associated with a concomitant decrease in IL-4 synthesis (Fig. 2B). Therefore, continuous E2 exposure in vivo selectively enhances Th1 cell priming.

We next tested whether the enhancing effect of E2 was dependent on the dose of hormone administered in vivo. The data in Fig. 3 show that mice implanted with E2 pellets exhibited a dose-dependent and significant increase in antigen-specific Th1 responses as compared to untreated castrated mice. Strong enhancement of Th1 cell priming was also observed with mice treated with 0.25-mg E2 pellets (not shown). Interestingly, treatment of mice with a cycle substitutive dose of E2 (0.01-mg pellets), that has been shown to induce substantial increase in uterus weight 17, was still able to induce a moderate but reproducible up regulation of IFN-γ synthesis in the majority of E2-treated mice in both experiments.

Figure 1.

 Continuous E2 administration enhances Th1 responsiveness. After bilateral ovariectomy (Ovx), 4-week-old B6 mice were implanted with E2 pellets (0.1 mg/60 days). Control mice were ovariectomized but received no pellet. At 8 weeks of age mice were immunized s.c. with 50 μg of OVA in CFA (A–C) or IFA (D–F). Draining lymph node cells (LNC) were harvested 9 days later and cultured (4×105 cells/well) in HL-1 synthetic medium for 72 h with the indicated doses of OVA. Proliferation was assessed by pulsing the cells with 1 μCi [3H]dThd during the last 8 h of culture (A, D). For cytokine production analysis, LNC (6×105 cells/well) were cultured in HL-1 medium with the indicated concentrations of OVA (B and E) or with 100 μg/ml of OVA (C and F). IFN-γ (B, E) and IL-2 (C, F) secretion was tested by ELISA in 72-h or 20-h culture supernatants, respectively. Data are expressed as mean ± SD of three mice per group. Results are from one representative experiment out of four performed. (*;p<0.05).

Figure 2.

 E2 selectively promotes Th1 cell priming in BALB/c mice. Castrated BALB/c mice were treated with E2 pellets (0.05 mg, 60-day release) and immunized with 50 μg of OVA in IFA. LNC were stimulated in vitro with the indicated concentrations of OVA. IFN-γ (A) and IL-4 (B) production were assessed by ELISA in 72-h culture supernatants. Results are expressed as mean ± SD of eight to ten mice per group. Data were pooled from two experiments. (*;*;*;p<0.001).

Figure 3.

Dose response effect of E2 on Th1 cell priming. Castrated B6 mice were treated with E2 pellets (60-day release) containing various amounts of hormone (from 0.01 to 0.1 mg) and immunized with 50 μg of OVA in CFA. CD8-depleted LNC were stimulated in vitro with 30 μg/ml of OVA and IFN-γ production assessed by ELISA in 72-h culture supernatants. Results are expressed as IFN-γ synthesis (ng/ml) from three to four mice per group. Two experiments are shown out of three performed.

2.2 E2 enhances antigen-specific T cell expansion in situ

To evaluate whether the enhancing effect of E2 on Th1 responsiveness is an in vivo acquired characteristic of Th lymphocytes, ex vivo isolated CD4 T cells from control or E2-treated mice were stimulated in vitro with antigen and BM-DC from normal B6 mice. In agreement with our previous experiments, CD4 T cells from E2-treated mice exhibited a stronger antigen-specific proliferative response (Fig. 4A) and IFN-γ-production (Fig. 4B) as compared to CD4 T cells from control mice. Moreover, 100-times less antigen was required to achieve the same degree of proliferation or cytokine production when T lymphocytes were obtained from estrogen-treated mice (Fig. 4). Thus, the E2-dependent increase in Th1 cell priming, did not appear to be influenced by origin of the APC involved in the recall T cell response in vitro, and most likely reflects an augmentation of the frequency of antigen-experienced Th1 cells primed in vivo.

To directly address this point, we analyzed the HEL-specific T cell response in B10.D2 mice, in which HEL-specific T cell responses are mainly directed against a subdominant (HEL12–25/IAd) and a dominant (HEL107–116/IEd) epitope of the protein 20, 21. The T cell response against the immunodominant epitope HEL107–116 involves a public BV8S2-J1S5 repertoire, found in all HEL-immunized mice, with a conserved eight-amino acid CDR3 motif GTGNNQAP 22. To estimate the frequency of HEL-reactive CD4 T cells bearing this particular CDR3 motif we designed a real time quantitative PCR approach 23. With this methodology, we could measure the in situ clonal expansion of the BV8S2-J1S5 CD4 T cell clonotype (that recognizes the immunodominant HEL107–116/IEd epitope) in a complex population of T cells 23, 24. Ovx control and E2-treated B10.D2 mice were immunized with HEL in IFA. cDNA prepared from purified CD4 T cells, at day 9 after immunization, were used for public rearrangement quantification. The data in Fig. 5A show that the frequency of T cells expressing the rearranged BV8S2-J1S5 public chain was up-regulated tenfold in E2-treated mice. In vitro antigenic challenge of the same LNC populations showed increased proliferative responses of T cells from E2-treated mice in response to both native protein antigen and immunodominant epitopes of the protein (not shown) and was associated with a strong production of IFN-γ (Fig. 5B) but not IL-4 (not shown). Expansion of antigen-specific CD4 T cell in situ was further demonstrated by analyzing the CDR3 size distribution of rearranged BV8S2-J1S5 gene segments on ex vivo isolated CD4 T cells. As shown in Fig. 5C, the CDR3 size distribution among all BV8S2 rearranged TCR gave a set of seven peaks with a Gaussian profile in both control and E2-treated mice. Conversely, a major peak with a CDR3 size of eight amino acids was observed in BV8S2-J1S5 rearranged TCR in E2-treated but not in castrated mice. Taken together, our data show that induction of primary T cell responses under continuous E2 administration is characterized by a strong increase in the frequency of antigen-specific memory/effector Th1 cells in vivo.

Figure 4.

E2-dependent increase in antigen-specific CD4 T cell proliferation and IFN-γ-production. Castrated B6 female mice were treated with E2 pellets (0.1 mg/60 days) and immunized s.c. with 50 μg of OVA in IFA as in Fig. 1. Draining LNC were harvested 9 days later. Purified CD4+ T cells (2×105 cells/well) were stimulated with 1 to 100 μg/ml of OVA in the presence of 30×103 C57BL/6 BM-DC. Proliferative response (A) and IFN-γ production (B) in 72-h culture supernatants were measured as in Fig. 1. Data are expressed as the mean of three to five mice per group ± SD. Results are from one representative experiment out of four performed. (*;p <0.05).

Figure 5.

Continuous E2 administration increases antigen-specific CD4 T cell expansion in immune lymph nodes. After bilateral ovariectomy, 4-week-old B10.D2 mice were implanted with E2 pellets (0.1 mg/60 days). Control mice were sham operated but received no pellet. At 8 weeks of age mice were immunized s.c. with 14 nmoles (200 μg) of HEL in IFA. Draining lymph node were harvested 9 days later and cDNA were prepared from purified CD4 T cells. The clonal size of HEL107–116/IEd-reactive CD4 T cells that expressed a public BV8S2-J1S5 CDR3 motifs was quantified by real time quantitative PCR as described in Sect. 4 (A). For cytokine production analysis (B), CD8-depleted LNC (5×105 cells/well) were cultured with HEL protein or peptides as indicated. IFN-γ secretion in 72-h culture supernatants was tested by ELISA. Results are expressed as mean of three individual mouse ± SD. Values are from one representative experiment out of three performed. Panel (C) shows the CDR3 size distribution analysis of rearranged BV8S2-J1S5 gene segments in CD4 T cells from Ovx or Ovx/E2-treated mice obtained as in (A). BV8S2-BC PCR products were subjected to run-off reactions with BC- or BJ1S5-specific fluorescent primers as indicated and analyzed by Immunoscope. Data are from one representative mouse out of three to four mice analyzed in each group.

2.3 The E2-driven increase in Th1 cell development in vivo requires the functional expression of ERα but not ERβ in bone marrow-derived cells

To evaluate whether the E2-driven enhancement on Th1 cell priming was mediated through classical nuclear ER, we tested the effect of continuous E2 delivery on the T cell response induced in mice with a disrupted ERα or ERβ gene 25. Data illustrated in Fig. 6 show that E2 administration induced a strong expansion of antigen-specific IFN-γ-producing CD4 T cells in both ERß–/– (Fig. 6D) and ERβ+/+ wild-type (Fig. 6C) littermates. In striking contrast, no increase in Th1 responsiveness was seen in ERα-deficient mice supplemented with E2 (Fig. 6B). As expected, antigen-specific Th1 cell development was strongly up-regulated in E2-treated ERα+/+ littermates (Fig. 6A). Thus, our data demonstrate that the effect of E2 on the enhancement of T cell priming and differentiation toward type-1 effector lymphocytes in vivo requires the functional expression of ERα but not ERβ.

To determine whether ERα expression was required in hematopoietic cells, bone marrow chimeras were generated by injecting bone marrow cells from control or ERα-deficient mice into 6-week-old irradiated B6 Rag2–/– recipients (ERα–/– → B6 Rag2–/–). After 3 to 4 weeks of treatment with E2 (0.1-mg pellets), mice were immunized with OVA-CFA and the polarization of OVA-specific T cell response was evaluated in vitro. Since it has been previously shown that E2 administration may perturb T cell homeostasis by inducing thymic and splenic atrophy 26, 27, we first determined the effect of E2 treatment on lymphocyte populations in immune lymph node cells from ERα+/+ and ERα–/– chimeras treated or not with E2. The data in Table 1 show that the number of lymphoid cells recruited in the lymph node was significantly decreased by E2 treatment in both type of chimeras and mainly affected CD4 and CD8 T lymphocytes. Similar results were obtained in normal B6 mice treated with E2 (not shown). Thus, the diminished T cell numbers observed in E2-treated animals did not required ERα expression in hematopoietic cells. In striking contrast, while the number of T cells in the draining lymph nodes were similar in both group of chimeras treated with E2, only mice reconstituted with bone marrow cells from wild-type, but not ERα–/– mice, exhibited a dramatic increase in antigen-specific T cell proliferation and IFN-γ production (Fig. 7A and B). This E2-driven enhancement of Th1 cell responsiveness was further confirmed by intracellular cytokine staining of antigen-stimulated CD4 T cells. As illustrated in Fig. 8A, the frequency of IFN-γ-producing CD4 T cells was increased in E2-treated ERα+/+ → B6 Rag2–/– chimeras but not in mice reconstituted with ERα-deficient bone marrow cells (Fig. 8B). The frequency of IL-4-producing T lymphocytes was very low (<1%) in all groups. The CD4 T cell recovery was increased fourfold in antigen-stimulated LNC from ERα+/+ →B6 Rag2–/– chimeric mice treated with E2, as compared to control Ovx animals (not shown), and was associated with a strong expansion of IFN-γ-producing CD4 T cells (Fig. 8C). Increased Th1 cell development was further confirmed by analyzing IL-18R expression on CD4 T cells. As shown in Fig. 8D, E2-treatment resulted in a stronger expansion of CD44highIL-18Rpos CD4 T cells in ERα+/+ → B6 Rag2–/– mice, but not in B6 Rag2–/– mice reconstituted with ERα–/– bone marrow cells. These data demonstrate that the enhancing effect of E2 on antigen-specific priming of memory/effector Th1 cells required the functional expression of ERα on bone-marrow derived cells in vivo.

Figure 6.

Endogenous expression of ER α but not β is required for the up-regulation of Th1 cell responsiveness induced by E2. Castrated B6 mice with disrupted ERα (B) or ERβ (D) genes treated or not with E2 (0.1 mg/60 days) were immunized with OVA in IFA. ERα+/+ (A) or ERβ+/+ (C) littermates were used as control. Antigen-specific IFN-γ production was tested as in Fig. 1. Data are expressed as mean ± SD of three to five mice per group. Results are from one representative experiment out of two performed. (*p<0.05).

Table 1. E2 induces a decreased recruitment of lymph node T cells in E2-treated mice in a mechanism that does not require ERα expression in hematopoietic cells
Bone marrow cells fromB6 Rag2–/– recipientsa)Lymph node cell numberAbsolute number of lymphocytes (×10–6)
 (×10–6)CD4+CD8+B220+
  1. a) Ovariectomized B6 Rag2–/– mice were irradiated and reconstituted with bone marrow cells from ERα+/+ or ERα–/– mice. Chimeric mice were then treated or not with E2 and immunized with OVA-CFA as described in Fig. 7. The absolute number of CD4+, CD8+ T lymphocytes and B lymphocytes in draining lymph node cells were determined by flow cytometric analysis. Results are expressed as mean ± SD of four to five mice per group.

ERα+/+Ovx119.4 ± 29.148.9 ± 8.937.7 ± 6.923.8 ± 13.4
ERα+/+Ovx + E2 77.6 ± 21.9 (p < 0.02)28.5 ± 3.2 (p < 0.01)22.9 ± 5.1 (p < 0.01)20.7 ± 8.3 
ERα–/–Ovx117.4 ± 23.847.9 ± 9.435.6 ± 8.024.0 ± 4.1 
ERα–/–Ovx + E2 93.8 ± 8.5 (p < 0.02)33.8 ± 4.6 (p < 0.01)23.3 ± 2.8 (p < 0.01)24.7 ± 3.1 
Figure 7.

The E2-mediated enhancement of Th1 priming required ERα expression in BM-DC. Recipient mice were castrated at 4 weeks of age. Bone marrow chimeras were generated by injecting bone marrow cells (10×106 cells /mouse i.v.) from control (A) or ERα-deficient (B) B6 mice into 6-week-old irradiated (400 rad) Rag2–/– B6 recipients. At 4 to 5 weeks after bone marrow reconstitution, mice (four to five per group) were implanted or not with 0.1-mg E2 pellets and immunized 3 to 4 weeks later with OVA-CFA. Antigen-specific T cell proliferative response and IFN-γ production were tested as described in Fig. 1. Results are from one representative experiment out of two performed. (*;p<0.05; *;*;p<0.01).

Figure 8.

Lack of Th1 cell expansion in the absence of ERα expression in hematopoietic cells. (A and B) Immune LNC (2×106 cells/ ml) from chimeric mice treated as described in Fig. 7 were cultured with OVA (10 μg/ml ) in HL-1 medium in 24-well plates (1 ml/ well). After 4 days of culture, leaving cells were collected after Ficoll centrifugation. Purified CD4 T cells were then stimulated with PMA/ionomycin, surface-labeled with anti-CD4 mAb and then fixed. Cells were then stained intracellularly for IFN-γ (FL-1) and IL-4 (FL-2) production. Data are expressed as mean ± SD of four to five mice per group. Th1 cell expansion was assessed by quantifying the number of IFN-γ-producing CD4 T cells (C), or by measuring the frequency of CD44high/IL-18R+ T cells (D). Data are expressed as absolute number of T cells with the indicated phenotype per input of 106 CD4 T cells (mean ± SD of four to five mice per group). The frequencies of CD4 T lymphocytes in LNC from the different groups were: ERα+/+ Ovx (41.4±3.7); ERα+/+ Ovx/E2 (33.6±2.1); ERα–/– Ovx (46.6±1.7) and ERα–/– Ovx/E2 (35.9±2.2).

3 Discussion

Although it has long been recognized that gender differences affect adaptive immune responses and susceptibility to Th1-dependent autoimmune diseases, sex-linked factors able to influence activation, expansion and/or differentiation of Th cells to the type 1 pathway remained to be clearly identified. Data in the present report demonstrate that, at physiological levels, the female hormoneE2 dramatically enhances CD4 T cell expansion and promotes Th1 cell development in vivo. Several lines of evidence support this conclusion. The frequency of IFN-γ, but not IL-4-producing CD4 T cells was selectively increased in E2-treated mice. Enhanced Th1 cell expansion was further confirmed by analyzing the frequency of CD4 T cells that had up-regulated IL-18R that is selectively expressed on Th1, but not Th2 cells 28, 29. Moreover, continuous administration of E2 induced a massive expansion of antigen-specific type-1 memory/effector CD4 T cells independently of the type of adjuvant used for immunization (CFA or IFA). This effect of E2 was observed for physiological levels of hormone. Administration of 0.05 to 0.1-mg pellets (secreting levels of E2 corresponding to physiological serum level <100 pg/ml found during late pregnancy in mice 18) resulted in a strong expansion of antigen-specific Th1 cells. Even treatment of castrated female mice with 0.01-mg pellets that secrete cycle substitutive level of E2 30 had measurable effect on Th1 cell priming. Thus, although the protocol we have used for E2 delivery do not allow normal fluctuations in estrogens levels which normally occur in mice with ovaries, we could demonstrate a reproducible effect of E2 on the enhancement of primary antigen-specific CD4 T cell responses, for doses ranging from cycle substitutive to late pregnancy levels. This is in agreement with previous experiments by others showing enhanced T cell proliferative responses 12 and IFN-γ-production 13, 14 in female versus male mice following antigen immunization. Finally, we directly showed that the frequency of HEL107–116/IEd-specific T cells bearing the public BV8S2-J1S5 rearrangement was strongly increased in immune lymph node CD4 T cells from E2-treated B10.D2 mice primed with HEL-IFA. The strong antigen-specific Th cell priming that we observed in E2-treated mice may reflect an overall increase in the frequency of the primary T cell repertoire, and/or could be due to the selective expansion of T cell clones expressing preferred CDR3 loop and high affinity TCR. Alternatively, this condition of immunization may have led to the presentation of a greater variety of T cell determinants, including cryptic epitopes, by lymph node DC, resulting in the recruitment of T cells with broader specificity. The latter possibility is unlikely since the T cell response to various HEL epitopes in E2-treated B10.D2 mice remained mainly focused to the same dominant HEL determinants. Thus, the increased T cell responsiveness induced by E2 was not due to a diversification of the immune response against subdominant or cryptic epitopes. Rather, it seemed to involve the same T cell repertoire as the one selected in primary T cells from control untreated mice, although we cannot exclude that selective expansion of T cell clones with preferred CDR3 motifs had occurred in E2-treated mice.

Estrogens exert their biological effect through nuclear hormone receptors, which interact with promoter elements to regulate gene transcription. Two different ER have been described ERα 31, 32 and ERβ 33. ER have been found to be expressed in human lymphoid cells 34, CD8 T cells 35, 36 and macrophages 36, 37. Likewise, ERα gene expression has been also reported in CD4 T cells and macrophages in mice 38. However, it has been recently shown that E2 may target both mouse CD4 and CD8 T cells through putative plasma membrane receptors rather than through the classical nuclear ER, resulting in cytosolic Ca2+ influx 39. Whatever the differences in tissue distribution and estrogen-signaling pathway between classical ERs and putative plasma membrane receptors, our data show that, as for many others physiological actions of E2 40, ERα is essential to mediate the E2-driven enhancement of primary antigen-specific CD4 T cell expansion and Th1 development in vivo.Indeed, up-regulation of Th1 cell responsiveness following E2 delivery was abrogated in ERα-deficient mice, but not in mice with disrupted ERβ genes. Furthermore, using bone marrow chimeras, we have clearly established that ERα expression in hematopoietic cells was required for the enhancing effect of E2 on primary antigen-specific CD4 T cell responses. Current experiments are in progress to determine which populations of immunocompetent cells is specifically targeted by E2 in vivo.

Despite the strong enhancing effect of E2 on antigen-specific primary CD4 T cell responses, we observed a significant decrease (30–40%) in the absolute number of T lymphocytes of both CD4 and CD8 subsets in immune lymph nodes. This effect was also observed in the spleen, although it was less pronounced (10–26%, data not shown). Interestingly, using bone marrow chimeras that selectively lacked ERα expression in hematopoietic cells, we show that, while the E2-driven enhancing effect on Th1 responses was abolished, the diminished recruitment of T lymphocytes was unchanged. Thus, the inhibitory effect of E2 on T cell homeostasis and/or tissue recruitment was not dependent on ERα expression in bone marrow derived cells. Several mechanisms could explain this effect. Estrogens have been shown to regulate the expression of endothelial cell adhesion molecules and to alter leukocytes binding to cultured endothelial cells 41. E2 could therefore alter T lymphocytes homing to lymphoid organs. Indeed, high dose E2 administration have been shown to reduce lymphocyte numbers in both spleen and thymus 26. Alternatively, E2 could modulate the central mechanism of positive and/or negative selection in the thymus. In support of this, it has been shown that ERs are expressed at near uterine levels in thymic epithelial cells in the mouse 42 and that ERα expression on non-hematopoietic cells is necessary in thymic development and E2-induced thymic alterations 27.

The E2-driven increase in Th1 responsiveness described in the present study identifies an important physiological mechanism by which sex-linked factors can affect T cell activation, expansion and differentiation in vivo. Understanding the cellular and molecular basis that drive the selective expansion of Th1 cells under E2 supplementation may provide important new insights into the mechanisms by which estrogens affect immunity and susceptibility to autoimmune diseases, and might be useful to design improved strategies for vaccination.

4 Materials and methods

4.1 Mice, antigens and immunization

C57BL/6 (B6) and BALB/c mice were purchased from the Centre d'Elevage R. Janvier (Le Genest St Isle, France) and maintained in our animal facilities under pathogen-free conditions. Rag2-deficient B6 mice were obtained from the Centre National de la Recherche Scientifique (Centre de Développement des Techniques Avancées, Orléans, France). B10.D2 mice were obtained from Harlan (UK). 129/Sv-B6 mice with a disrupted estrogen receptor (ER) α or β gene were obtained as previously described 25 and back-crossed on B6 background at least for five generations. One- to three-month-old female mice were used in all experiments. For estrogen hormone administration, 3-mm pellets (Innovative Research of America, Sarasota, FL) containing varying amounts of E2 were implanted subcutaneously (s.c.) on the animal back at 4 weeks of age after bilateral ovariectomy. These pellets provide continuous controlled release of a constant level of hormone over a period of 60 days. Four weeks later mice were subsequently immunized in the hind footpads with protein antigen in complete (CFA) or incomplete (IFA) Freund's adjuvant and the polarization of CD4 T cell response was tested after in vitro restimulation with the antigen as described 19. Hen egg-white lysozyme (HEL) and OVA were obtained from Sigma (St Louis, MO). HEL peptides (purity>85%) were purchased from Neosystem (Strasbourg, France).

4.2 Generation of ERα+/+ or ERα–/– bone marrow chimeras

At 4 weeks of age, B6 Rag 2–/– mice were bilaterally ovariectomized. One week later, they were sub-lethally irradiated (400 rad) and i.v. injected the day after with 10×106 bone marrow cells from either ERα+/+ or ERα–/– mice. After 4 additional weeks, treated mice were implanted subcutaneously with E2 pellets and immunized as indicated above.

4.3 T cell assays

For cytokine production analysis, popliteal lymph node cells (LNC), depleted or not of CD8+ cells, were cultured at 5 or 6×105 cells/well in 96-well culture plates (Costar, Cambridge, MA) in synthetic HL-1 medium (HYCOR, Irvine, CA) supplemented with 2 mM L-glutamine (Gibco, Cergy Pontoise, France) and 50 μg/ml gentamicin (Sigma) with the indicated antigen concentrations. Cultures were incubated for 3 days in a humidified atmosphere of 6% CO2 in air. LNC were depleted of CD8 cells by sequential incubation with KT1.5 mAb 43 culture SN and M-450 anti-rat IgG Dynabeads (Dynal, Oslo, Norway), followed by magnetic cell separation as described 19. CD4 T cells were isolated by the same way using a cocktail of following mAb: anti-CD8 KT1.5, anti-B220 RA3 (TIB 146, ATCC, Rockville, MD), anti-MHC II M5/114 ( TIB 120, ATCC) and anti-CD11b Mac-1 (TIB 128, ATCC). The purity of CD4 T cells wasabove 95% by flow cytometric analysis. CD4+ T (2×105 cells/well) isolated as described below were stimulated with syngeneic bone marrow-derived dendritic cells (BM-DC, 3×103 cells/well) and antigen in HL-1 medium. BM-DC were prepared from C57BL/6 bone marrow cells essentially as described elsewhere 24. Supernatants were collected between 20 to 72 h for cytokine analysis. IFN–γ, IL-2 and IL-4 were quantified by sandwich ELISA as described 19. For T cell proliferation assays, cell cultures were pulsed for 8 h with 1 Ci [3H]dThd (40 Ci/nmol, the Radiochemical Centre, Amersham, GB) before harvesting on glass fiber filter. Incorporation of [3H]dThd was measured by direct counting using an automated β-plate counter (MatrixTM 9600, Packard, Meriden, CT).

4.4 Flow cytometric analysis

For the intracellular analysis of cytokine synthesis, LNC were resuspended at 106/ml and stimulated with PMA (Sigma, 50 ng/ml) plus ionomycin (Sigma, 0.5 μg/ml) for 4 h, in the presence of brefeldin A (Sigma) at a concentration of 10 μg/ml during the last 2 h. Cells were then harvested, washed in the presence of brefeldin A and stained using biotinylated-GK1.5 anti-CD4 mAb (TIB 207, ATCC) and PE-conjugated Pgp1 anti-CD44 (PharMingen), followed by Streptavidin-CyChrome (PharMingen). Labeled cells were then fixed with 2% paraformaldehyde (Fluka Chemie AG, Buchs, Switzerland). Intracytoplasmic staining for IL-4 and IFN-γ was performed as described 19. For IL-18R staining, cells were incubated with anti-IL-18R goat IgG (R&D Systems Europe Ltd., Abingdon, GB) followed by biotinylated anti-sheep/goat Ig antibodies (Amersham France SA, Les Ulis, France) and streptavidin-CyChrome (PharMingen). Cells were also labeled with anti-CD4-FITC and anti-CD44-PE mAb as above. Analysis was performed on CD44high CD4 T cells. The frequency of lymphocyte populations in immune LNC was determined by using anti-CD4-FITC, anti-CD8-PE and anti-B220-biotin mAb (PharMingen). Data were collected on 20,000 CD4+ cells on a XL Coulter cytometer (Coultronics, France), and analyzed using the CellQuest software (Becton Dickinson, MountainView, CA).

4.5 TCR-β repertoire analysis and public clone quantification

Total RNA was extracted from purified CD4 T cells using the RNeasy system (QIAGEN S.A., Courtaboeuf, France) and cDNA was synthesized using a oligo(dT)17 primer and M-MLV reverse transcriptase (Boehringer Mannheim, Mannheim, Germany). The CDR3 size distribution of BV8S2-BC rearranged TCR was performed by Immunoscope as described 19. The frequency of HEL107–116/IEd-CD4+ T cells bearing a public clonotypic rearrangement was performed by real time quantitative PCR as described 23, 24.

4.6 Statistical analysis

Statistical significance of differences between groups of continuous variables were analyzed using the Mann-Whitney U test.

Acknowledgements

The skillful technical assistance of M. Calise and S. Pilipenko is gratefully acknowledged. We wish to thank J. van Meerwijk and E. Joly for critical reading of the manuscript. This work was supported by Institut National de la Santé et de la Recherche Médicale, and by grants from Association pour la Recherche sur la Sclérose en Plaques, AssociationFrançaise contre les Myopahties, Association pour la Recherche contre le Cancer, European Community (QLG1-CT2001–01918) and Theramex Laboratories. The work at IGBMC was supported by a grant from the Association pour la Recherche sur le Cancer and the Fondation pour la Recherche Médicale. L. G. is supported by a fellowship from Ligue Française contre la Sclérose en Plaques. G. F. is on leave of absence from Ecole Nationale Vétérinaire de Toulouse, France.

Footnotes

  1. 1

    WILEY-VCH

  2. 2

    WILEY-VCH

  3. 3

    WILEY-VCH

  4. 4

    WILEY-VCH

  5. 5

    WILEY-VCH

  6. 6

    WILEY-VCH

  7. 7

    WILEY-VCH

  8. 8

    WILEY-VCH

Ancillary