The intracellular progesterone receptor regulates CD4⁺ T cells and T cell-dependent antibody responses

The intracellular progesterone receptor links the reproductive and immune systems via regulation of CD4+ T cells.

Abstract

Pg has distinct immunomodulatory properties involved in poorly understood immune phenomena, including maternal tolerance of the fetus, increased risk of certain infections during pregnancy or after Pg birth control, and pregnancy-associated remission of autoimmune disease. Several potential mechanisms have been identified, including alteration of Th1 and Treg activity, but the precise cellular and molecular targets of Pg immunomodulation in vivo remain obscure, partly because Pg can signal through several different receptor types. One such receptor, the iPR, encoded by the pgr gene, is essential for reproduction in female mice and is expressed in the thymus and CD4⁺ T cells. We hypothesized that iPR regulates CD4⁺ T cell activity and adaptive immune responses in vivo. With the use of iPR KO mice, we demonstrate that iPR specifically suppresses TD antibody responses, primarily by dampening CD4⁺ Teff activity, likely via transcriptional repression of the IFN-γ gene and modulation of other programs regulating CD4⁺ T cells. Our results highlight a novel mechanism linking the endocrine and immune systems, and they offer insight into important but poorly understood phenomena in women's health and autoimmunity.

Introduction

Pg is a reproductive steroid with distinct immunoregulatory properties involved in important but poorly understood phenomena, such as maternal tolerance of the fetus [1], increased susceptibility to infection by select intracellular pathogens during pregnancy [2], increased risk of HIV infection after use of synthetic Pg birth control [3], and pregnancy-related remission of the autoimmune diseases, RA and MS [4]. Potential mechanisms include suppression of Th1 and Th17 activity, expansion of Tregs, thymic involution, and regulation of B cells (reviewed in refs. [4, 5]). However, the precise cellular and molecular targets of Pg immunomodulation in vivo remain obscure, in part because Pg can signal through several distinct receptor types: iPR, a ligand-activated transcription factor encoded by the pgr gene; recently described mPRs; PGRMCs; and at high-physiologic concentrations—the GR [6, 7]. Whereas in vivo immune functions of GR have been studied extensively, very little in this regard is known about the individual Pg receptors. Moreover, when compared with naturally occurring Pg, synthetic progestins and antiprogestins vary considerably in their binding to and activation of iPR, mPR, PGRMCs, and GR [6, 7]. Thus, dissecting the specific immune functions of each Pg receptors is essential for understanding how endogenous Pg and commonly prescribed progestin drugs influence immunity, tolerance, and autoimmunity.

We focused our investigation on iPR, as it is both critical to female reproduction [8] and expressed in the thymus [9] and CD4⁺ T cells [10]. Accordingly, we hypothesized that one in vivo function of iPR is to regulate CD4⁺ T cell activity and adaptive immunity. With the use of iPR KO mice, we demonstrate that iPR specifically suppresses TD antibody responses, primarily by dampening CD4⁺ Teff activity, likely via transcriptional repression of the IFN-γ gene and modulation of other programs involved in T cell help. These results highlight a novel mechanism linking the reproductive and immune systems, and they offer insight into poorly understood but important phenomena in women's health and autoimmunity.

MATERIALS AND METHODS

Mice

iPR KO mice on a mixed 129/B6 background were kindly provided by Dr. John Lydon [8] and housed in the University of Washington animal facilities (Seattle, WA, USA) under SPF conditions. The iPR KO mice were backcrossed nine generations onto inbred B6 mice to create B6.iPR KO mice, which were used in select experiments, and also crossed with B6 mice expressing a TCR Tg specific for H-2^b and an OVA-specific peptide (B6.OT-II mice) [11]. Only adult male and adult female virgin mice were used in experiments, which were performed in compliance with the University of Washington Institutional Animal Care and Use Committee.

Immunizations

Mice were immunized i.p. with one of the following: 25 μg of the hapten DNP conjugated to KLH (DNP-KLH; United States Biological, Swampscott, MA, USA) and adsorbed to 4 mg alum (Pierce, Rockford, IL, USA); 10 μg of the hapten NP conjugated to OVA (NP-OVA; Biosearch Technologies, Novato, CA, USA) and adsorbed to 4 mg alum; or 10 μg DNP conjugated to Ficoll (DNP-Ficoll; Biosearch Technologies) alone or in some experiments, adsorbed to 4 mg alum. Mice were bled at 0–21 days after immunizations.

Determination of serum and culture supernatant Ig levels

Total serum Ig levels were determined by ELISA using goat anti-mouse Ig capture antibodies (SouthernBiotech, Birmingham, AL, USA), followed by HRP-conjugated goat anti-mouse IgM, IgA, IgE, and IgG subclass detection antibodies (SouthernBiotech). Depending on genetic background, IgG_2a (129/B6) or IgG2_c (B6) was assessed. Serum antihapten Ig levels were determined by ELISA, using BSA conjugated to NP or DNP as capture molecules and the HRP-conjugated antibodies mentioned above.

Cell isolation and cell culture

Freshly isolated spleens were treated with Liberase Blendzyme 2, per the manufacturer's instructions (Roche Applied Science, Indianapolis, IN, USA), or minced; red cells were lysed with hypertonic solution (BioLegend, San Diego, CA, USA). Splenic CD4⁺ T cells and CD19⁺ B cells were isolated using positive-selection magnetic columns (Stemcell Technologies, Vancouver, Canada); CD4⁺ T cells were 85–90% pure and CD19⁺ B cells, 90–95% pure. T cells were stimulated with plate-bound hamster anti-mouse CD3 (10 μg/ml; Clone 2C11) and graded doses of soluble hamster anti-mouse CD28 (Clone 37N), both provided by Dr. Jeff Ledbetter, University of Washington, in the presence of Pg (10⁻⁸ M; Sigma-Aldrich, St. Louis, MO, USA) or vehicle (ethanol). B cells were stimulated with various combinations and doses of rat anti-mouse CD40 (Clone 1C10, provided by E.A.C.), mouse rIL-4 (R&D Systems, Minneapolis, MN, USA), and Escherichia coli LPS (Sigma-Aldrich). All cell culture experiments used RPMI-1640 growth medium (HyClone, Thermo Scientific, Logan, UT, USA), supplemented with 10% charcoal-stripped FBS (Sigma-Aldrich).

Flow cytometry

For immunophenotyping, we used fluorochorome-conjugated mAb specific for CD4, CD8, CD11c, CD19, B220, and CD86 (BD Biosciences, San Jose, CA, USA); CD11b, CXCR5, PD-1, GL7, and Vα2 TCR (eBioscience, San Diego, Ca, USA); IgG₁, CD3ϵ, CD25, CD28, CD44, CD45.2, CD69, and I-A^b (BioLegend); and mPDCA1 (Miltenyi Biotec, Germany). FITC PNA was from Vector Labs (Burlingame, CA, USA). Flow cytometric data were acquired with FACScan Aria and Canto machines (BD Biosciences) and analyzed further with FlowJo software (TreeStar, Ashland, OR, USA).

Adoptive transfer of OT-II spleen CD4⁺ T cells

Splenic CD4⁺ T cells were purified from iPR⁺ (pooled WT and iPR^+/−) or iPR KO × OT-II Tg mice (CD45.2⁺). After assessment for purity and OT-II-specific TCR expression (∼60% in all samples), 1 million CD4⁺ Vα2⁺ spleen cells were injected into age- and sex-matched CD45.2⁻ B6 recipients. Twenty-four hours later, recipient mice were immunized with NP-OVA + alum.

RNA isolation and Q-PCR

Cell pellets were frozen in liquid nitrogen and stored at −80°C. Total RNA was isolated using RNeasy Mini Kits (Qiagen, Valencia, CA, USA). After enzymatic degradation of DNA, RT was performed using the SuperScript III First-Strand synthesis system (Invitrogen, Carlsbad, CA, USA). From cDNA isolates, specific sequences were amplified using Absolute Blue Q-PCR SYBR Green ROX Mix (Thermo Scientific). The results were normalized to 18S RNA levels in each sample. Forward/reverse primer sequences used were: ifng (GCTTTGCAGCTCTTCCTCAT/GTCACCATCCTTTTGCCAGT), il2 (ACTTGCCCAAGCAGG CCACA/GGACCTCTGCGGCATGTTCTGG), il4 (GACGCCATGCACGGAGATGGA/GCGAAGCACCTTGGAAGCCCT), and tbx21 (GGTGTCTGGGAAGCTGAGAG/CCACATCCACAAACATCCTG). Efficiency for each primer pair was optimized and ranged from 80% to 95%.

Statistical analysis

Student's t-tests, ANOVA with Bonferroni post-tests, and Wilcoxon signed rank tests were performed using Prism 5 software (GraphPad, San Diego, CA, USA).

RESULTS AND DISCUSSION

iPR selectively suppresses antibody responses to a TD antigen

To assess if loss of iPR impacted immune homeostasis, we compared the numbers of splenic subsets and serum Ig levels in adult iPR⁺ (WT or iPR^+/−) and iPR KO mice. Compared with iPR⁺ mice, iPR KO mice had normal numbers of major spleen subsets (Fig. 1A). Consistent with this, we observed no differences in total levels of serum Ig classes and IgG subclasses (Fig. 1B). As Pg is known to suppress Th1 and enhance Th2 responses [4], we also compared ratios of Th1- to Th2-related Ig classes and IgG subclasses in individual mice but found no differences (Fig. 1C). Subanalysis by sex of spleen cell subsets and Ig levels/ratios did not appreciably change the results (data not shown). Thus, loss of iPR in adult SPF mice does not impact spleen cell homeostasis or levels of circulating Ig.

Figure 1. Spleen cell subset numbers and serum Ig levels in iPR KO mice.

Spleen cell subset abundances (A), serum Ig levels (B), and serum Ig ratios (C) in unimmunized age- and sex-matched iPR⁺ and iPR KO mice. Numbers of spleen B cells (CD19⁺B220⁺), CD4⁺ and CD8⁺ T cells (CD3⁺), NK cells (NK1.1⁺), myeloid DCs (mDCs; CD11c^hi/mPDCA1⁻), and pDCs (B220⁺mPDCA1⁺) were determined by flow cytometry. Serum Ig levels were determined by ELISA. n.s. = not significant: P > 0.05 by Bonferroni post-test of two-way ANOVA.

To assess impacts of iPR on CD4⁺ T cell and B cell responses, we immunized WT and iPR KO mice and measured antibody responses. Compared with WT mice, iPR KO mice generated significantly more hapten-specific IgG after immunization with the TD antigen, DNP-KLH (Fig. 2A and B). As antibody responses varied between experiments, we also assessed overall effects by combining results and performing an aggregate analysis of all four TD immunization experiments (Fig. 2B). This analysis showed that iPR KO mice consistently made significantly more antigen-specific IgG₁ and IgG_2a antibodies but not more antigen-specific IgM. This was evident in female or male mice (data not shown), indicating that iPR, although critical for female reproduction [8], regulates antibody responses in both sexes and outside of pregnancy. This is not entirely surprising, as iPR is expressed in tissues of both sexes [12]; basal Pg levels between male and female rodents [13] and humans [14] do not differ greatly; and male iPR KO mice show behavioral abnormalities [15]. Thus, low-physiologic concentrations of Pg may be sufficient to activate iPR and suppress TD antibody responses in vivo.

Figure 2. Loss of iPR differentially impacts primary TD and TI-2 antibody responses.

Serum anti-DNP responses in groups of age- and sex-matched mice immunized with DNP-KLH + alum (TD; A and B) or DNP-Ficoll (TI-2; C and D). (A and C) Representative of four experiments each. (B and D) Aggregate analysis of all TD or TI-2 experiments, where values normalized to maximal WT response. Shown are means ± se for individual mice. *P < 0.05; **P < 0.01, Student's t-test (A and C) or Wilcoxon signed rank test (B and D).

In contrast, loss of iPR did not affect antibody responses to a TI-2 antigen, DNP-Ficoll (Fig. 2C and D). As illustrated in Fig. 2C, compared with controls, iPR KO mice immunized with DNP-Ficoll made similar amounts of hapten-specific IgM and IgG₁ and slightly less IgG₃. Aggregate analysis of four independent experiments (Fig. 2D) showed a similar pattern, and again, the differences in IgM and IgG₃ responses did not reach statistical significance. Again, the sex of mice did not influence the TI-2 antibody responses (data not shown). Together, our results indicate that iPR selectively inhibits antigen-specific TD antibody responses but not TI-2 antibody responses.

iPR KO B cells are deficient in IgG production in vitro

Our initial results with TD responses suggested that iPR was functioning at the level of CD4⁺ T cells, TD signaling in B cells, or responses to the alum adjuvant we used. The elevated TD IgG antibody responses and the relatively normal IgM antibody responses in iPR KO mice (Fig. 2A and B) also suggested that the iPR might act directly on B cell Ig CSR. Pauklin and Petersen-Mahrt [16] showed that high physiologic concentrations of Pg, via iPR, can suppress transcription of the activation-induced deaminase gene and limit CSR in mouse B cells. Accordingly, we tested if loss of iPR results in increased T cell-driven B cell IgG production by isolating splenic CD19⁺ B cells from WT and iPR KO mice and comparing their production of IgG after stimulation with anti-CD40 or LPS (Fig. 3A and B); iPR KO B cells produced similar or slightly reduced levels of IgM, IgG₁, and IgG_2c compared with WT B cells. Differences in IgG₁ levels correlated well with numbers of IgG₁⁺ cells in culture (Fig. 3B), implying that fewer iPR KO B cells were undergoing CSR as opposed to simply secreting less IgG₁/cell. Thus, under low Pg concentrations, as occurs in nonpregnant mice and in the cultures conditions we used, iPR may function differently than it does under high Pg concentrations [16]. Consistent with this idea, Vermeulen et al. [17] showed that the iPR agonist medroxyprogesterone enhanced IgG₁ release from B cells maximally at a low physiologic concentration (0.1–1.0 nM). Finally, the iPR KO mice do not appear to have abnormal responses to alum adjuvant, as peritoneal inflammatory responses to alum, critical for its adjuvant effects [18], differed little between genotypes (see Fig. 5B and C). Therefore, we decided to examine whether CD4⁺ T cells from WT and iPR KO mice differed.

Figure 3. Loss of iPR differentially impacts B cell and CD4⁺ T cells.

(A) Ig production by freshly isolated WT and iPR KO splenic B cells cultured for 7 days with IL-4 plus indicated doses of anti-CD40 or LPS; representative of four experiments. (B) Their expression of surface IgG₁, and culture IgG₁ levels after treatment with IL-4 plus anti-CD40 (2 ng/ml) or LPS (2 μg/ml). (C) IFN-γ and IL-4 production by, and survival of, freshly isolated WT and iPR KO splenic CD4⁺ T cells, cultured 5 days with anti-CD3 and indicated doses of anti-CD28; representative of four independent experiments. (D) Surface marker expression on WT (shaded histograms) and iPR KO (lines) CD4⁺ T cells before culture; representative of three experiments. (E) Select gene mRNA expression in WT and iPR KO CD4⁺ T cells cultured for indicated times in the presence of anti-CD3 and anti-CD28 (2.5 μg/ml). (F) Effects of Pg treatment on gene expression in same cultures (2-day time-point). (E and F) Representative of at least two experiments. Shown in all are means ± se of experimental replicates. *P < 0.05; **P < 0.01; ***P < 0.001, Bonferroni post-test of two-way ANOVA.

Figure 5. Spleen GC B cell and Tfh cell responses after TD immunization and peritoneal responses to alum injection in iPR KO mice.

(A) Groups of age- and sex-matched WT and iPR KO mice immunized with DNP-KLH + alum and at indicated time-points, numbers of spleen B220⁺PNA⁺GL-7⁺ GC B cells and CD4⁺CXCR5⁺PD-1⁺ Tfh cells determined by flow cytometry; n = 4/time-point/genotype; representative of two independent experiments. Age-matched WT or iPR KO mice injected with alum i.p. (4 mg), and 18 h later, numbers of peritoneal lavage cell subsets were determined by flow cytometry using the following scheme: A, macrophages (CD11b^hi); B, neutrophils (B1, CD11b^int/CD86⁻) and monocytes (B2, CD11b^int/CD86⁺); C, lymphocytes (CD11b^lo/CD11c⁻); and D, DCs (CD11b^lo/CD11c⁺). (B) Representative FACS plots from WT mice. FSC, Forward-scatter. (C) Total numbers of peritoneal lavage cells by subset; shown are means ± sd of pooled data from two independent experiments; n = 4 (alum-injected). P > 0.05, Student's t-test.

Increased IFN-γ expression in iPR KO CD4⁺ T cells

Freshly isolated splenic CD4⁺ T cells from WT and iPR KO mice showed no differences in their expression of activation (CD69), memory (CD44), or costimulatory (CD28) markers (Fig. 3D). However, when compared with controls, iPR KO CD4⁺ T cells produced significantly more IFN-γ after anti-CD3/CD28 costimulation (Fig. 3C). These differences could not be attributed to differential cell survival and appeared to be specific to IFN-γ, as IL-4 levels in the same cultures differed little between genotypes (Fig. 3C). The increased IFN-γ production by iPR KO CD4⁺ T cells parallels the increased Th1-associated IgG_2a antibody responses after TD immunization in iPR KO mice (Fig. 2B). These results suggest that the ability of Pg to suppress Th1 responses may be mediated through iPR. As iPR is a transcriptional regulator, we tested whether IFN-γ was dysregulated at the level of mRNA expression. Consistent with IFN-γ protein levels, ifng mRNA levels were significantly higher in iPR KO CD4⁺ T cells after stimulation, whereas IL-4 (il4), IL-2 (il2), and T-bet (tbx21) mRNA levels were not significantly different (Fig. 3E). Thus, iPR may preferentially target ifng for transcriptional repression rather than interfering with more proximal mechanisms of Th1 differentiation. Consistent with this idea, the addition of physiologic concentrations of Pg (10⁻⁸ M) to CD4⁺ T cell cultures significantly impaired ifng mRNA induction in WT cells but not in iPR KO cells (Fig. 3F). Under these conditions, Pg treatment had no significant effects on il4, il2, or tbx21 expression. Thus, Pg can selectively repress ifng gene expression in CD4⁺ T cells via iPR. This mechanism likely contributes to suppression of Th1 effector activity and cellular immune responses during high Pg states such as pregnancy or use of Pg birth control. Interestingly, T cell-specific ablation of the GR gene in mice also results in increased ifng mRNA and IFN-γ protein induction [19], suggesting that GR and iPR may target ifng for transcriptional repression. Furthermore, repression of ifng by iPR could relate to impaired HIV defenses after Pg birth control or suppression of Th1-associated autoimmune diseases, such as RA and MS during pregnancy, a high Pg state. However, repression of ifng alone does not account for why loss of iPR also leads to increased Th2-related IgG₁ antibody responses (Fig. 2A and B), which are normally suppressed by IFN-γ and enhanced by IL-4.

iPR KO CD4⁺ T cells induce stronger antigen-specific Ig responses in vivo

To further investigate which features of the abnormal TD responses in iPR KO mice might be intrinsic to CD4⁺ T cells, we generated B6.iPR KO mice bearing the OT-II Tg. Purified, splenic OT-II T cells from iPR⁺ or iPR KO donors were injected into WT hosts, and serum hapten-specific antibody was measured after immunization with NP-OVA (Fig. 4A). WT and iPR KO OT-II cells had similar expression of CD69 and CD44 before transfer (data not shown). However, mice receiving iPR KO OT-II cells generated significantly higher anti-NP IgM, IgG₁, and IgG_2c responses compared with those receiving iPR⁺ OT-II cells (Fig. 4A). These results indicate that iPR can suppress TD antibody responses via modulation of CD4⁺ T cells. Importantly, the similarity in enhancement of antibody responses in global iPR KO mice (Fig. 2A and B) and WT mice receiving iPR KO OT-II cells (Fig. 4A) suggests that CD4⁺ T cells are a chief target of iPR action in regulating TD antibody responses. The fact that the number of WT and iPR KO OT-II cells recovered from spleens did not differ appreciably (Fig. 4B) further suggests that iPR regulates CD4⁺ Teff activity rather than proliferation or survival. The increased IgG_2c antibody responses in mice receiving iPR KO OT-II cells could be a result of the enhanced production of IFN-γ by iPR KO OT-II cells (Fig. 3E and F). On the other hand, the increased IgG₁ and IgM antibody responses in mice receiving iPR KO OT-II cells, evident to some degree in iPR KO mice (Fig. 2A and B), suggest iPR regulates additional CD4⁺ T cell functions involved in generating antibody responses. Differentiation of naïve CD4⁺ T cells into Tfh cells is required for generation of normal TD IgM and IgG responses [20]. However, Tfh differentiation after TD immunization does not appear to be suppressed by iPR, as iPR KO mice immunized with DNP-KLH + alum showed normal increases in numbers of total splenic CD4⁺CXCR5⁺PD-1⁺ Tfh cells, despite having marginally elevated numbers of PNA⁺GL-7⁺ GC B cells (Fig. 5A). Thus, it will be of interest to test if loss of iPR in Tfh cells results in increased expression of molecules important for GC B induction, such as CD40 ligand, signal lymphocyte activation molecule-associated protein, OX40, ICOS, or IL-21 [20]. Given recent observations that Pg treatment in mice can expand systemic Treg populations [10], possibly via iPR in CD4⁺ T cells [21], it may be that iPR signaling in CD4⁺ T cells also enhances expansion of a Treg subset—Tfr cells—which limits Tfh induction/activity and in turn, GC B cell responses [22]. In other words, iPR in CD4⁺ T cells may suppress TD antibody responses by altering the balance of Tfh versus Tfr responses after TD immunization. Finally, pgr mRNA is expressed in a variety of immune cells, including macrophages and DCs (data not shown), so it is possible that iPR may signal in cell types other than CD4⁺ T cells or B cells to regulate TD antibody responses. Together, our results indicate that iPR regulates the magnitude and quality of TD antibody responses, primarily through control of CD4⁺ Teffs, which likely involves transcriptional repression of ifng and regulation of other programs involved in T cell help.

Figure 4. iPR regulates TD antibody responses via CD4⁺ T cell-intrinsic mechanisms.

(A) Serum NP-specific Ig responses in WT mice receiving WT or iPR KO OT-II cells and immunized with NP-OVA + alum. Shown are serum anti-NP levels in individual recipients at indicated time postimmunization (p.i.). (B) Numbers and percentages of CD4⁺Vα2⁺CD45.2⁺ OT-II cells recovered from spleens of recipients at Day 14 postimmunization. *P < 0.05; **P < 0.01; ***P < 0.001, two-way ANOVA. P > 0.05, paired Student's t-test; representative of three experiments.

To our knowledge, this is the first report of an in vivo immune function for the pgr gene product, iPR. Very recently, Lee et al. [21] used iPR KO CD4⁺ T cells to demonstrate that high physiologic Pg concentrations enhance and stabilize in vitro Treg induction via iPR-dependent mechanisms. Thus, iPR may control TD antibody responses via modulation of CD4⁺ Tregs and Teffs. iPR effects in CD4⁺ T cells may also contribute to suppression of Th1 and Th17 activity and expansion of Treg populations during normal pregnancy or after Pg treatment [4]. iPR actions in other immune cell types, particularly DCs and macrophages, are the subject of ongoing investigations. Additional studies comparing cell-specific iPR functions in low and high Pg states will provide further mechanistic insight into important immune phenomena associated with pregnancy and use of Pg birth control.

AUTHORSHIP

A.H.W. and G.C.H. contributed substantially to the conception, design, and execution of all experiments. E.A.C. contributed substantially to the conception, design, and execution (through collaborative sharing of resources) of all experiments.