Steady-state estradiol triggers a unique innate immune response to allergen resulting in increased airway resistance

Abstract

Rationale: Asthma is a chronic airway condition that occurs more often in women than men during reproductive years. Population studies have collectively shown that long-term use of oral contraceptives decreased the onset of asthma in women of reproductive age. In the current study, we hypothesized that steady-state levels of estrogen would reduce airway inflammation and airway hyperresponsiveness to methacholine challenge.

Methods: Ovariectomized BALB/c mice (Ovx) were implanted with subcutaneous hormone pellets (estrogen, OVX-E2) that deliver consistent levels of estrogen [68 ± 2 pg/mL], or placebo pellets (OVX-Placebo), followed by ovalbumin sensitization and challenge. In conjunction with methacholine challenge, immune phenotyping was performed to correlate inflammatory proteins and immune populations with better or worse pulmonary outcomes measured by invasive pulmonary mechanics techniques.

Results: Histologic analysis showed an increase in total cell infiltration and mucus staining around the airways leading to an increased inflammatory score in ovarectomized (OVX) animals with steady-state estrogen pellets (OVX-E2-OVA) as compared to other groups including female-sham operated (F-INTACT-OVA) and OVX implanted with a placebo pellet (OVX-Pl-OVA). Airway resistance (Rrs) and lung elastance (Ers) were increased in OVX-E2-OVA in comparison to F-INTACT-OVA following aerosolized intratracheal methacholine challenges. Immune phenotyping revealed that steady-state estrogen reduced CD3+ T cells, CD19+ B cells, ILC2 and eosinophils in the BAL across all experiments. While these commonly described allergic cells were reduced in the BAL, or airways, we found no changes in neutrophils, CD3+ T cells or CD19+ B cells in the remaining lung tissue. Similarly, inflammatory cytokines (IL-5 and IL-13) were also decreased in OVX-E2-OVA-treated animals in comparison to Female-INTACT-OVA mice in the BAL, but in the lung tissue IL-5, IL-13 and IL-33 were comparable in OVX-E2-OVA and F-INTACT OVA mice. ILC2 were sorted from the lungs and stimulated with exogenous IL-33. These ILC2 had reduced cytokine and chemokine expression when they were isolated from OVX-E2-OVA animals, indicating that steady-state estrogen suppresses IL-33-mediated activation of ILC2.

Conclusions: Therapeutically targeting estrogen receptors may have a limiting effect on eosinophils, ILC2 and potentially other immune populations that may improve asthma symptoms in those females that experience perimenstrual worsening of asthma, with the caveat, that long-term use of estrogens or hormone receptor modulators may be detrimental to the lung microenvironment over time.

Fig. 1

Airway and vascular cell infiltrates are increased in the airways when estrogen is given at steady-state levels. A An overview our experimental protocol is shown where female BALB/c mice were ovarectomized at 3 weeks of age (OVX) and implanted with subcutaneous estrogen (17β-E2, 0.1 mg) or placebo pellets. Sham-operated females (F-INTACT) were included as controls. OVA-treated animals were sensitized to chicken egg albumin by intraperitoneal injections with OVA and aluminum hydroxide (100 µL). Saline treated animals were included as controls. B–D Levels of estrogen were determined by ELISA in lung tissue, BAL and serum. One-way ANOVA were used to determine statistical differences followed by Tukey’s post-test to determine between groups differences. Statistical significance was assigned when p-value was less than 0.05; *p < 0.05, **p < 0.01; ***p < 0.001; ****p < 0.0001

Fig. 2

Airway resistance is increased in ovarectomized animals treated with estradiol in comparison to female INTACT, OVA-treated animals. After five daily airway challenges with i.n. ovalbumin, lungs were inflated then excised followed by paraffin embedding and H&E staining. A–I Histological scoring was performed by a blinded pathologist on two sections per animal, 3–4 animals per group. The data are shown as mean ± SEM. Results represent 2 independent experiments. One-way ANOVA were used to determine statistical differences followed by Tukey’s post-test to determine between groups statistical effects. Statistical significance was assigned when p-value was less than 0.05; *p < 0.05, **p < 0.01; ***p < 0.001; ****p < 0.0001

Fig. 3

Airway resistance is increased in ovarectomized animals treated with estradiol in comparison to female INTACT, OVA-treated animals. After completion of the experimental protocol outline in Fig. 1 animals were subjected to methacholine challenge (12.5 mg/mL) using the flexivent system (Scireq). A Resistance (Rrs, baseline subtracted) and B elastance (Ers; baseline subtracted) were measured following methacholine challenge. C Pressure–volume loops were determined and normalized to body weight. The data are shown as means ± SEM. Results represent 3 independent experiments, N = 4 = 7/group. Mixed-ANOVAs were used to determine statistical differences using multiple variate selection followed by Sidak’s post-test (C), while a one-way ANOVA was used to determined statistical differences; as before a Tukey’s post-test was used to determine between group differences. Statistical significance was assigned when p-value was less than 0.05; *p < 0.05, ****p < 0.0001

Fig. 4

Total immune cell infiltrate is reduced in BAL fluid from OVX-E2-OVA-treated animals compared to F-INTACT, OVA-treated. Animals were treated as in Fig. 1. Cytospins were prepared from BAL fluids collected immediately following methacholine challenges. Total BAL return volume and numbers of total cells (B) were recorded and applied to cytospin slides at 300×g for 10 min. Slides were stained with Giemsa and 2–4 fields were assessed per slide. 200 cells per field were counted as C macrophages, D neutrophils, E eosinophils or F lymphocytes. A Percentages of each cell population are shown. B–F Are the counts of cells determined by multiplying the percentages by the total number of cells recovered in each lavage. Data are representative of 3 independent experiments with 3–8 animals per group. One-way ANOVA was used to determine statistical differences across all groups; Tukey’s post-test were used to determine statistical differences between groups. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 5

Allergic inflammatory cytokines in OVX-E2-OVA-treated animals in BAL and lung tissue compared to female intact mice treated with airway allergen. Animals were prepared as in Fig. 1. Animals were euthanized by overdosing with ketamine (200 µg/kg). Bronchoalveolar lavage fluid was collected and cleared by centrifugation (400×g) prior to applying to ELISA plates for analysis. A, B IL-5, C, D IL-13, E, F IL-33, G, H CCL3, I, J CCL22, K, L CCL12, M, N CCL11, O, P MPO, and Q, R KC are shown as pg/mL. Statistical significance was determined as in previous studies using a one-way ANOVA followed by Tukey’s post-test which are reported as p < 0.05 and considered significant in all studies. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 6

Circulating cytokines and chemokines are selectively altered with steady-state estrogen treatment. Groups of mice were prepared as in Fig. 1. Circulating cytokines were determined by ELISA in serum that was diluted 1:5 for all groups. A IL-5 and B IL-13 cytokines were detected, C the neutrophil activation marker, MPO, and D–F circulating chemokines D CCL22, E CCL3 and F KC are shown. The data are shown as mean ± SEM. Data are representative of 4 independent experiments. Statistical significance was determined as before with a p < 0.05 (indicated by *), **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 7

Estrogen treatment reduces eosinophils and neutrophils in the BAL of OVX-OVA-challenged mice. Animals were prepared as in Fig. 1. A Flow cytometry gating strategy for eosinophils defined as live, singlet, CD11b⁺Siglec-F⁺CD11c⁻ cells is shown, while neutrophils are defined as FSC^loSSC^hiCD11b+CD11c+Siglec-F^lo. B Total CD45+ BAL cells, C eosinophil and D neutrophils per mL of lavage fluid are shown for all groups. E Total CD45+ cells detected in the remaining lung tissue F eosinophils and G neutrophils were quantified. Bar graphs show the data as mean ± SEM (n = 5–6 mice/group). The data represent 4 independent experiments. Statistical significance was determined as in Fig. 3. * Indicates p < 0.05; **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 8

Steady-state estrogen reduces airway and lung T cells, but CD19+ B cells are preserved and comparable to hormonally intact, female controls. Groups of animals were prepared as in previous figures. Bronchoalveolar lavage fluid was collected immediately following euthanasia. A–D By flow cytometry total CD3+ and CD19+, T and B cells, respectively, were determined following treatment with 5 consecutive daily i.n. OVA administration. A, C Counts of T cells and B cells per mL of returned lavage fluid and B, D counts of T and B cells acquired from remaining lung tissue are shown. E OVA-specific IgE was also measure in serum that was diluted 1:5 by ELISA. Results are representative of two separate experiments with 4–6 animals included per group. As before, statistical significance was determined using one-way ANOVA with Tukey’s post-tests. As in previous figures * indicates p < 0.05; **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 9

In vivo estrogen treatment reduces ILC2 numbers and dampens the IL-33 responses of ILC2. A Flow cytometry gating strategy of ILC2 defined as live, singlet cells lacking lineage markers (LIN: CD11b, CD11c, CD3, CD19, NK1.1, FCeR1, Ter-1) and expressing KLRG-1 and CD127, as well as CD25, ICOS and IL-33R (data not shown for confirmatory markers). B, C Counts of ILC2 in the B BAL and C lungs are shown. D Viability was assessed for ILC2 in bronchoalveolar lavage (BAL) fluid and lungs during flow cytometry analysis. The data are shown as mean ± SEM and are representative of 4 independent experiments. E–H Cytokine expression by lung ILC2 cultured with IL-2, IL-7 and IL-33 for 3 days were determined by ELISA. E IL-5, F IL-13, G CCL3, and H CCL22 are displayed as pg/cell based on ILC2 count determined at the beginning of culture. Results are representative of 3 separate experiments with 4–6 animals included per group. As before, statistical significance was determined using one-way ANOVA with Tukey’s post-test as in previous figures. * Indicates p < 0.05; **p < 0.01, ***p < 0.001, ****p < 0.0001