The Injectable Contraceptive Medroxyprogesterone Acetate Attenuates Mycobacterium tuberculosis-Specific Host Immunity Through the Glucocorticoid Receptor

Abstract

Background: The effects of the widely used progestin-only injectable contraceptives, medroxyprogesterone acetate (MPA) and norethisterone acetate (NET-A), on host susceptibility to Mycobacterium tuberculosis (Mtb) are unknown.

Methods: We recruited human immunodeficiency virus-uninfected females, not taking any contraceptives, from Cape Town, South Africa, to evaluate the effect of MPA, NET-A, and dexamethasone on Mtb containment in monocyte-derived macrophages co-incubated with purified protein derivative (PPD)-driven peripheral blood-derived effector cells.

Results: MPA (P < .005) and dexamethasone (P < .01), but not NET-A, significantly attenuated Mtb containment in Mtb-infected macrophages co-cultured with PPD-driven effector cells at physiologically relevant concentrations and in a dose-dependent manner. Antagonizing the glucocorticoid receptor with mifepristone (RU486) abrogated the reduction in Mtb containment. In PPD-stimulated peripheral blood mononuclear cells, MPA and dexamethasone, but not NET-A, upregulated (median [interquartile range]) regulatory T cells (5.3% [3.1%-18.2%]; P < .05), reduced CD4+ T-cell interferon-γ (21% [0.5%-28%]; P < .05) and granzyme B production (12.6% [7%-13.5%]; P < .05), and reduced CD8+ perforin activity (2.2% [0.1%-7%]; P < .05). RU486 reversed regulatory T-cell up-regulation and the inhibitory effect on Th1 and granzyme/perforin-related pathways.

Conclusions: MPA, but not NET-A, subverts mycobacterial containment in vitro and downregulates pathways associated with protective CD8+- and CD4+-related host immunity via the glucocorticoid receptor. These data potentially inform the selection and use of injectable contraceptives in tuberculosis-endemic countries.

Keywords: injectable contraceptive; medroxyprogesterone acetate; norethisterone; pathogenesis; tuberculosis.

Figure 1.

Medroxyprogesterone acetate (MPA), but not norethisterone acetate (NET-A), decreases peripheral effector cell–mediated Mycobacterium tuberculosis (Mtb) containment in vitro. Mtb-infected monocyte-derived macrophages (MDMs) were incubated alone (n = 11) or co-cultured with peripheral blood mononuclear cells that were incubated without (n = 10) or with (n = 11) 12 µg/mL purified protein derivative (PPD) alone and in parallel with physiologically representative concentrations of dexamethasone (n = 9), MPA (n = 11), or NET-A (n = 9). A, The number of colony-forming units/mL was determined by counting colonies on Middlebrook H9/OADC agar. B, Percentage of mycobacterial containment in the presence of 12 µg/mL PPD alone or in combination with 100 nM dexamethasone, MPA, or NET-A (error bars represent median and interquartile range). The percentage of containment was calculated relative to the reference control (Mtb-infected MDMs only); the dotted line represents relative levels of containment by the reference control (Mtb-infected MDMs only, no Mtb containment). C, Dose-dependent levels of percentage of mycobacterial containment in the presence of PPD and increasing concentrations of dexamethasone, MPA, and NET-A (a figure greater than –100 indicates proliferation of the organism; error bars represent mean and standard deviation). A nonlinear fit of the dose response was performed for MPA and dexamethasone. Data were analyzed for statistical significance by one-way analysis of variance withDunnett post test or Wilcoxon signed-rank paired t test, where **, *** and **** indicate P < .01, P < .005 and P < .0001, respectively. Statistical trend analysis for each dose response was performed by the Wilcoxon signed-rank paired t test, as further extended by Cuzick [18], and showed a significant trend only for dexamethasone (P < .0001) and MPA (P < .0001). Abbreviations: ANOVA, analysis of variance; CFU, colony-forming units; Dex, dexamethasone; MDM, monocyte-derived macrophage; MPA, medroxyprogesterone acetate; Mtb, Mycobacterium tuberculosis; NET-A, norethisterone acetate; PPD, purified protein derivative.

Figure 2.

Medroxyprogesterone acetate (MPA) decreases peripheral effector Mycobacterium tuberculosis (Mtb) containment through the glucocorticoid receptor (GR). Mtb-infected monocyte-derived macrophages (MDMs) were cultured alone (n = 6) or were co-cultured with untreated effector cells (n = 6) or effector cells treated with 12 µg/mL purified protein derivative (PPD) alone (n = 6); PPD and 100 nM dexamethasone (n = 6); PPD and 100 nM MPA (n = 6); PPD and 100 nM norethisterone acetate (NET-A) (n = 6); PPD and 1 µM RU486 (a GR antagonist) (n = 6); PPD, dexamethasone, and RU486 (n = 5); PPD, MPA, and RU486 (n = 6); or PPD, NET-A, and RU486 (n = 6). The relative number of colony-forming units/mL (A) was determined as indicated previously. The percentage of mycobacterial containment (B) was determined as indicated previously to the reference control (Mtb-infected MDMs only). Error bars represent median and interquartile range. Data were analyzed for statistical significance by one-way analysis of variance with Dunnett post test or Wilcoxon signed-rank paired t test, where * and ** indicate P < .05 and P < .01, respectively. Abbreviations: ANOVA, analysis of variance; CFU, colony-forming units; Dex, dexamethasone; MDM, monocyte-derived macrophages; MPA, medroxyprogesterone acetate; Mtb, Mycobacterium tuberculosis; NET-A, norethisterone acetate; PPD, purified protein derivative.

Figure 3.

Medroxyprogesterone acetate (MPA), but not norethisterone acetate (NET-A), increases regulatory T-cell (Treg) expansion in vitro, and this is attenuated by RU486. Effector cells were generated by incubating peripheral blood mononuclear cells with (n = 7) or without (n = 7) 12 µg/mL purified protein derivative (PPD) or in combination with 100 nM dexamethasone (n = 6), MPA (n = 7), NET-A (n = 6), and/or 1 µM RU486 (n = 6) for 7 days at 37°C. The levels of Tregs were determined by flow cytometry. Tregs were classified as CD4⁺, CD25⁺, and FOXP3⁺. Dead cells were excluded from the scatterplots prior to analysis, and negative gates were set using fluorescence minus one controls. Only the single cellular populations were analyzed. Error bars represent median and interquartile range. Data were analyzed for statistical significance by one-way analysis of variance with Dunnett post test, where * indicates P < .05. Abbreviations: Dex, dexamethasone; MPA, medroxyprogesterone acetate; NET-A, norethisterone acetate; PPD, purified protein derivative; Treg, regulatory T cell.

Figure 4.

Medroxyprogesterone acetate (MPA), but not norethisterone acetate (NET-A), reduces interferon gamma (IFN-γ) expression and cytotoxic T-lymphocyte (CTL) induction in CD8⁺ and CD4⁺ T cells. Peripheral blood mononuclear cells were treated with (n = 6) or without (n = 6) 12 µg/mL purified protein derivative (PPD) or in combination with 100 nM dexamethasone (n = 6), MPA (n = 6), NET-A (n = 6), and/or 1 µM RU486 (n = 6) for 7 days at 37°C. IFN-γ expression and CTL induction were assessed in the CD8⁺ and CD4⁺ T cells using flow cytometry. Error bars represent median and interquartile range. Data were analyzed for statistical significance by one-way analysis of variance with Dunnett post test, where * and ** indicate P < .05 and P < .01, respectively. Abbreviations: Dex, dexamethasone; IFN-γ, interferon gamma; MPA, medroxyprogesterone acetate; NET-A, norethisterone acetate; PPD, purified protein derivative.