Dendritic cell function and pathogen-specific T cell immunity are inhibited in mice administered levonorgestrel prior to intranasal Chlamydia trachomatis infection

Abstract

The growing popularity of levonorgestrel (LNG)-releasing intra-uterine systems for long-acting reversible contraception provides strong impetus to define immunomodulatory properties of this exogenous progestin. In initial in vitro studies herein, we found LNG significantly impaired activation of human dendritic cell (DCs) and their capacity to promote allogeneic T cell proliferation. In follow-up studies in a murine model of intranasal Chlamydia trachomatis infection, we analogously found that LNG treatment prior to infection dramatically reduced CD40 expression in DCs isolated from draining lymph nodes at 2 days post infection (dpi). At 12 dpi, we also detected significantly fewer CD4⁺ and CD8⁺ T cells in the lungs of LNG-treated mice. This inhibition of DC activation and T cell expansion in LNG-treated mice also delayed chlamydial clearance and the resolution of pulmonary inflammation. Conversely, administering agonist anti-CD40 monoclonal antibody to LNG-treated mice at 1 dpi restored lung T cell numbers and chlamydial burden at 12 dpi to levels seen in infected controls. Together, these studies reveal that LNG suppresses DC activation and function, and inhibits formation of pathogen-specific T cell immunity. They also highlight the need for studies that define in vivo effects of LNG use on human host response to microbial pathogens.

Figure 1. LNG inhibits human DC activation.

Negatively selected human DCs were incubated for 24 h with indicated LNG concentrations or vehicle alone, then incubated for 24 h with poly I:C (1.5 μg/mL). DCs were stained with a live/dead near-IR dye and a panel of fluorescently-tagged mAbs to identify viable DC populations by flow cytometry (described in Materials and Methods). (a) Representative contour plots of CD40 and CD80 expression by untreated or LNG (4 μM)-treated mDCs stimulated with poly I:C; quadrant numbers denote percent expression. (b–d) mDC expression of (b) CD80, (c) CD86, and (d) CD40 after poly I:C stimulation. Data from 8 independent experiments with results normalized (i.e., by designating vehicle-only cultures as 100% activation) as detailed in Materials and Methods (bars denote means ± SD). Statistical analyses performed using 1-way ANOVA with Dunnett’s multiple comparisons test, *p < 0.05; ***p < 0.001.

Figure 2. LNG inhibits human DC function.

(a–c) Negatively selected human DCs were LNG-treated and poly I:C stimulated as described in Fig. 1, then co-cultured with CTV-labeled naïve allogeneic T cells. Co-cultures were maintained 7 days, then T cells immunostained for flow cytometric analysis of proliferation. (a) Representative contour plots of CD4⁺ T cell proliferation from co-cultures with untreated or LNG (4 μM)-treated DCs (numbers denote percentages of proliferating CD3ε⁺CD4⁺ T cells). (b–c) Proliferation of (b) CD4⁺ and (c) CD8⁺ T cells from co-cultures with untreated or LNG-treated DCs (data from 6 independent experiments normalized and analyzed as detailed in Materials and Methods); (bars indicate means ± SD). (d–f) CTV-labeled naïve allogeneic T cells were incubated with LNG or vehicle overnight, then stimulated with beads coated with anti-CD3 and anti-CD28 antibodies. T cell cultures were maintained 5 days, and cells immunostained for flow cytometric analysis of proliferation. (d) Representative contour plots of T cell proliferation (4 μM LNG); (numbers identify percentages of proliferating CD3ε⁺CD4⁺ T cells). Proliferation of (e) CD4⁺ and (f) CD8⁺ T cells in T cell-only cultures treated with vehicle or LNG (data from 6 independent experiments that were normalized and analyzed as defined in Materials and Methods); (bars denote means ± SD). All statistical analyses were performed using 1-way ANOVA with Dunnett’s multiple comparisons test, ****p < 0.0001.

Figure 3. LNG inhibits human DC response to C. trachomatis stimulation.

Negatively selected human DCs were incubated for 24 h with indicated LNG concentrations or vehicle, then incubated for 24 h with inactivated C. trachomatis (MOI = 0.1 prior to inactivation). To identify live DCs, cells were stained as described in Fig. 1. (a) Representative contour plots displaying CD40 and CD80 expression by untreated and LNG (4 μM)-treated mDCs stimulated with C. trachomatis; quadrant numbers indicate percent expression. (b–d) mDC expression of (b) CD80, (c) CD86, and (d) CD40 induced by C. trachomatis. Data from 8 independent experiments were normalized as described in Materials and Methods (i.e., by designating vehicle-only cultures as 100% proliferation); (bars denote means ± SD). Statistical analyses made using 1-way ANOVA with Dunnett’s multiple comparisons test, *p < 0.05; **p < 0.01; and ***p < 0.001.

Figure 4. LNG reduces DC activation elicited by Chlamydia infection.

(a) LNG or matching placebo pellets were implanted into Balb/cJ female mice, and peripheral blood collected 5 days later to measure serum LNG concentrations. (b) Serum LNG levels; symbols represent values from individual mice and horizontal lines indicate means. (c–d) Other mice implanted with pellets as described in (a) were i.n. infected with 10⁵ IFU of live C. trachomatis. Mice were euthanized at 2 dpi, and DLNs collected to assess DC activation by flow cytometry. Representative contour plots show (c) CD40 and CD80 expression in live CD11c^hiMHC-II⁺CD8α^-CD11b⁺ cells (mDCs); numbers denote percent expression. (d) mDC expression of CD40 and CD80 from placebo- or LNG-pelleted mice from two independent experiments with 6 mice per condition; bars designate mean values ± S.D. Statistical analyses performed with 2-tailed unpaired Student t tests.

Figure 5. LNG impairs T cell expansion and pathogen-specific T cell effector function.

Mice were administered LNG or matching placebo pellet 5 days before i.n infection with 10⁵ IFU of live C. trachomatis. Other LNG-pelleted mice received agonist anti-CD40 mAb injection 1 day after infection. (a) Mice were euthanized at 12 dpi, and lungs processed into single-cell suspensions to enumerate CD45⁺, CD90.2⁺, CD4⁺, and CD8⁺ cells by flow cytometry. (b-c) In separate studies, groups of mice were treated and infected identically as described in (a). Mice were euthanized at 12 dpi, and lungs processed into single-cell suspensions to quantify GrzB levels inside CD8⁺ T cells and define Chlamydia-specific CD4⁺ and CD8⁺ T cell effector function. For the latter, BMDCs were obtained from untreated, uninfected syngeneic mice as defined in Materials and Methods, and stimulated overnight with C. trachomatis (MOI = 0.1). The next day, Chlamydia-infected mice (i.e., at 12 dpi) were euthanized, and single-cell suspensions of lung tissue incubated 24 h with the Chlamydia-activated BMDCs to quantify T cell secretion of IFN-γ, TNF, and IL-17 by flow cytometry. (b) Representative contour plots display production of IFN-γ, TNF, and IL-17 by CD4⁺ T cells and GzmB levels inside CD8⁺ T cells. (c) Between group-differences in levels of GzmB (CD8⁺ T cells) or intracellular accumulation of IFN-γ, TNF, and IL-17 (CD4⁺ and CD8⁺ T cells) (bars indicate mean values ± S.D). Data in (a) and (c) are from 2 independent experiments with 6 mice per group. Statistical analyses were completed using 1-way ANOVA with Dunnett’s multiple comparisons test.

Figure 6. LNG diminishes C. trachomatis clearance from pulmonary tissue.

Groups of mice were administered LNG or matching pellets 5 d prior to i.n infection with 10⁵ IFU of live C. trachomatis. Other groups of LNG-pelleted mice received agonist anti-CD40 mAb 1 day after chlamydial infection or CD4⁺ cell-depleting mAb 1 d prior and every other day for study duration. Animals were euthanized at 6 dpi or 12 dpi, and both lung lobes excised to quantify chlamydial DNA by RT-qPCR; (bars indicate mean ± C.I); (data from 2 independent experiments with 6 animals per group). Chlamydia levels at each dpi were compared using the 2-tailed unpaired Student t test.

Figure 7. Greater chlamydial burden in LNG-treated mice elicits increased mononuclear cell inflammation.

In separate studies, groups of mice were treated and infected identically as described in Fig. 6. Mice were euthanized at 12 dpi, and pulmonary tissue processed to define inflammation by histology. (a) Representative results from hematoxylin and eosin, IHC, and IF staining depicts the increased mononuclear cell inflammation seen in the lungs of LNG-treated- and CD4⁺ cell-depleted mice; IF staining: CD11b⁺ (green), DAPI (blue); black and white bars denote 200 μm and 20 μm, respectively. (b) Quantification of pulmonary CD45⁺ inflammatory aggregates identified by IHC staining (5 high-power fields examined per section); (bars denote mean ± S.D.). Data are from 2 independent experiments with 6 mice per group. Statistical analyses were performed using 1-way ANOVA with Dunnett’s multiple comparisons test.