Steroid responsiveness in Foxp3⁺ regulatory T cells determines steroid sensitivity during allergic airway inflammation in mice

Abstract

Glucocorticoids are a highly effective first-line treatment option for many inflammatory diseases including asthma. Some patients develop a steroid-resistant condition; yet, the cellular and molecular mechanisms underlying steroid resistance remain largely unknown. Here, we utilized a murine model of steroid-resistant airway inflammation and report that combining systemic dexamethasone and intranasal IL-27 is able to reverse the inflammation. Foxp3⁺ Tregs were required during dexamethasone/IL-27 treatment of steroid-resistant allergic inflammation, and importantly, direct stimulation of Tregs via glucocorticoid or IL-27 receptors was essential. Mechanistically, IL-27 stimulation in Tregs enhanced expression of the agonistic glucocorticoid receptor alpha isoform. Overexpression of inhibitory glucocorticoid receptor beta isoform in Tregs alone was sufficient to elicit steroid-resistance in a steroid-sensitive allergic inflammation model. Taken together our results demonstrate for the first time that Tregs are instrumental during steroid-resistance and that manipulating steroid responsiveness in Tregs may represent the novel strategies to treat steroid refractory asthma.

Introduction

Glucocorticoids are considered the frontline option to effectively treat acute and chronic inflammatory diseases including asthma (1, 2). However, some patients develop a condition refractory to glucocorticoids, known as ‘steroid resistance. Steroid resistance is generally difficult to treat and associated with severe and fatal diseases, requiring high dose steroid treatments to achieve minimal control (3). Since prolonged use of high dose steroids is not recommended due to detrimental side effects, identifying a new strategy to reverse steroid-resistant inflammation is a subject of utmost importance.

Experimental allergic airway inflammation in mice is widely used to investigate the pathogenesis of asthma. Asthmatic inflammation is heterogeneous in nature, and different types of allergic inflammation in the lung can be experimentally induced by differing the modes of sensitization. Antigen sensitization in alum followed by intranasal antigen challenge induces conventional eosinophilic airway inflammation associated with type 2 immunity (4). On the other hand, antigen sensitization in CFA instead induces airway inflammation associated with neutrophils and Th1/Th17 type immunity (5). Eosinophilic inflammation is managed by the treatment with dexamethasone (Dex), a synthetic glucocorticoid, while neutrophilic airway inflammation does not respond to the treatment. Thus, CFA-induced airway inflammation has been proposed as an animal model to investigate steroid-resistant allergic airway inflammation (6). The mechanisms underlying steroid resistance remain largely unknown.

We recently reported that Foxp3⁺ Tregs are indispensable for Dex to attenuate eosinophilic airway inflammation (7). Systemically or locally administered Dex completely loses its therapeutic effects when injected into Treg depleted (DTx injected Foxp3^DTR) mice with ongoing eosinophilic airway inflammation. Dex also failed to attenuate the inflammation in Treg-specific glucocorticoid receptor-deficient mice, suggesting direct signal of Dex in Tregs. We also reported that intranasally administered IL-27, an immune regulatory cytokine previously reported to dampen autoimmune inflammation via Tregs (8), also attenuates eosinophilic airway inflammation by directly acting on Tregs (9).

Here, we report that combining IL-27 with Dex effectively suppresses steroid-resistant airway inflammation in the lung. Interestingly, Dex/IL-27-mediated treatment is also dependent on Tregs, as it fails to attenuate the inflammation when Tregs are depleted. Expression of the receptors for Dex and IL-27 in Treg cells is necessary, suggesting a direct effect of Dex and IL-27 in Tregs. Mechanistically, IL-27 enhances the agonistic GRα expression in Tregs, improving Treg sensitivity to Dex. In support, overexpression of the inhibitory GRβ isoform in Tregs alone is sufficient to convert steroid sensitive inflammation to the resistant form. Taken together, steroid responsiveness in Tregs may be an important determinant of steroid responses and altered GR isoform expression in Tregs may represent a key mechanism underlying steroid resistance. Targeting steroid responsiveness in Tregs may offer a novel strategy to treat steroid-resistant asthma.

Materials and Methods

Mice

C57BL/6 and C57BL/6 Foxp3^DTR mice were purchased from the Jackson Laboratory (Bar Harbor, ME). C57BL/6 Foxp3^GFP mice were obtained from Dr. Vijay Kuchroo and further crossed to CD45.1⁺ congenic mice (10). All mice were maintained in a specific pathogen free facility located in the Lerner Research Institute and Northwestern University. All animal experiments were performed in accordance with protocols approved by the Cleveland Clinic and Northwestern University Institutional Animal Care and Use Committee.

Allergic airway inflammation

To induce steroid-resistant airway inflammation, mice were immunized s.c. with 5μg of cockroach Ag (CA, Greer Laboratory, Lenoir, NC) emulsified in 100μl of CFA containing 5mg/ml H37Ra (BD Difco, Franklin Lakes, NJ). Starting on day 14, the mice were i.n. challenged with 5μg CA in 50μl PBS for 4 consecutive days. Mice were sacrificed 24 hours after the last CA challenge. For steroid-sensitive inflammation, CA mixed in 100μl of alum adjuvant (Aluminum hydroxide, Sigma, St. Louis, MO) were injected i.p. on days 0 and 7, followed by intranasal CA challenge. In some experiments, Tregs were depleted by i.p. injection of 1μg of diphtheria toxin (DTx, Sigma). Bronchoalveolar lavage (BAL) and lung cells were isolated and examined as before (9). Lung histopathology was assessed as before (9).

Flow cytometry

Cells were stained with the following fluorescence-conjugated antibodies: anti-Ly6G (1A8), anti-Siglec-F (1RNM44N), anti-CD4 (RM4–5), anti-Nrp1 (3E12), anti-Foxp3 (FJK16s), anti-CD44 (IM7), anti-CD25(PC61.5), anti-GITR (DTA-1), anti-PD1 (J43), and anti-CTLA4 (UC10–489). Samples were acquired using a FACSFortessa flow cytometer (BD Biosciences, San Jose, CA) and analyzed using a FlowJo software (Tree Star Inc.).

In vitro T cell differentiation and adoptive transfer

Naive CD4⁺Foxp3^–CD44^low T cells were sorted and stimulated with immobilized anti-CD3 (2C11) and soluble anti-CD28 (37.51) mAbs, 100 U/ml rhIL-2 (obtained from the Biometric Research Program, National Cancer Institute), and 5 ng/ml TGFβ (PeproTech, Rocky Hill, NJ). After 3 days, inducible Tregs (CD4⁺ Foxp3⁺) cells were sorted using a FACSAria (BD Biosciences). A total of 2×10⁶ cells was i.v. transferred.

Murine GRβ overexpression plasmid and nucleofection

mGRβ cDNA overexpression in pcDNA3.1⁺ vector was previously reported (11). Naive CD4⁺ T cells were nucleofected with the control pcDNA or GRβ-plasmid using a mouse T cell Nucleofector Kit and Amaxa nucleofector (Amaxa, Koelin, Germany). Transfected cells were activated under the indicated polarizing conditions and treated with Dex (50nM) and harvested after 72 h.

RNA isolation and real-time PCR

Total RNA was isolated from iTreg cells using TRIzol reagent (Thermo Fisher Scientific, MA, USA). cDNA was synthesized using Moloney Murine Leukemia Virus (M-MLV) reverse transcriptase (Promega, Madison, WI). Primers used are shown in Supp Fig 1. qPCR analysis was performed using a QuantStudio 3 Real-Time PCR System (Applied Biosystems, Waltham, MA) using a SYBR green mastermix (Applied Biosystems). The expression was normalized to GAPDH.

Suppression assay

Purified naive CD4 T cells were labeled with 1.25 μM CFSE (Molecular Probe). For in vitro suppression assay, CFSE-labeled responder CD4 T cells were mixed with iTregs at ratios 1:1–1:16. Cells were plated in a U-bottomed 96-well plate with Dynabeads® Mouse T-Activator CD3/CD28 for T-Cell Expansion and Activation (Life Technologies). CFSE dilution was analyzed after 72 h by flow cytometry.

Cytometric bead array (CBA)

BAL fluids were collected from mice induced for allergic airway inflammation and treated with Dex and/or IL-27. Cytokines were quantified using a CBA kit (BD Biosciences) according to the manufacturer’s instructions.

Statistics

Statistical significance was determined by the Mann-Whitney (Two-tailed) or Kruskal-Wallis test using the Prism 7 software (GraphPad, La Jolla, CA). A p value of <0.05 was considered statistically significant.

Results and Discussion

Steroid-resistant allergic airway inflammation is attenuated by Dex with IL-27 supplementation

Allergen sensitization in CFA elicits neutrophilic airway inflammation that is refractory to steroid treatment (6, 12, 13). Mice sensitized by subcutaneous immunization with cockroach antigen (CA) in CFA were i.n. challenged with CA in PBS as in Supp Fig 2A. At sacrifice, neutrophils constituted ~30% of the bronchoalveolar lavage (BAL) cells (Fig 1A), which is in stark contrast to a conventional eosinophilic airway inflammation where neutrophils are only 1~2% (5). Systemic Dex treatment had no impact on attenuating cell infiltration in the airway and lung (Fig 1B and 1C). Inflammatory cytokines secreted in the airway were rather increased by Dex treatment (Fig 1D). Examination of the lung pathology and airway resistance further confirmed Dex resistance (Fig 1E and data not shown). Intranasally administered IL-27 during allergen challenge also downregulates eosinophilic airway inflammation (9). Analogous to ineffective treatment effects of Dex, intranasal IL-27 was unable to manage Dex-resistant airway inflammation (Fig 1B–1E).

Figure 1. Combining IL-27 and Dex effectively inhibits steroid resistant allergic airway inflammation via Tregs.

(A and B) Groups of mice were sensitized with CA in CFA and challenged with CA as described in Materials and Methods. Dex and/or IL-27 treatment was carried out as shown in Supp Fig 1A. The proportion and absolute numbers of inflammatory cells in the bronchoalveolar lavage (BAL) and lung were determined. (C) Cytokine expressing CD4 T cells in the lung. (D) Cytokines in the BAL fluid were determined by cytokine bead array. (E) H&E staining and histology score of the lung tissues are shown (original magnification ×10). (F) Foxp3^DTR mice were sensitized challenged as above. DTx was injected 1 day before and on the day of first Ag challenge. Mice were sacrificed 24 hours after the last challenge. Each symbol represents individually tested animal. Data shown are the mean ± SD of 2 independent experiments (n = 6–8). *, p< 0.05, **, p< 0.01, ***, p< 0.001, ****, p<0.0001 as determined by Kruskal-Wallis nonparametric test.

Interestingly, combining intranasal IL-27 with systemic Dex significantly dampened the inflammation. Total numbers of infiltrating cells, especially cytokine-producing lung CD4 T cells, were drastically reduced by the combined treatment (Fig 1B and 1C), although the proportion of cytokine expressing CD4 T cells was relatively comparable (Supp Fig 2B). BAL cytokines were substantially reduced (Fig 1D). Histopathology and lung function analysis further confirmed the effectiveness of Dex/IL-27 treatment (Fig 1E and data not shown). Thus, steroid-resistant allergic airway inflammation can be managed by supplementing IL-27 to Dex treatment.

Dex/IL-27 treatment of steroid-resistant inflammation is Treg dependent

We recently reported that Foxp3⁺ Tregs are essential mediators of IL-27- or Dex-mediated treatment of eosinophilic allergic airway inflammation (7, 9). To test whether Tregs are necessary during Dex/IL-27-mediated control of Dex-resistant inflammation, Foxp3^DTR mice sensitized as above were treated with the diphtheria toxin (DTx) one day prior to allergen challenge. >95% of Tregs were depleted (Supp Fig 2C). Consistent with eosinophilic inflammation, Dex/IL-27 induced attenuation was completely abrogated when Tregs were depleted. Dex/IL-27 treatment failed to diminish inflammatory cell infiltration in the BAL following Treg depletion (Supp Fig 2D). Dex/IL-27’s ability to reduce inflammatory cell accumulation in the lung was abrogated in the absence of Tregs (Fig 1F and Supp Fig 2E). Therefore, Tregs are core mediators of combined treatment of Dex and IL-27 in steroid-resistant allergic inflammation in the lung.

Direct signaling through the glucocorticoid receptor or IL-27Rα in Tregs is necessary for Dex/IL-27-mediated treatment of inflammation

Treg depletion drives a systemic inflammatory response (14), and the loss of steroid sensitivity could be due to systemic loss of Tregs or severe local inflammation in the airways. It is thus important to test whether the loss of steroid responses is not due to severe inflammatory conditions and whether the ability of Dex/IL-27 to suppress the inflammation operates via Tregs. We utilized Treg-specific IL-27Ra^−/− (Treg^ΔIl27ra) and GR^−/− (Treg^ΔGR) mouse models where Tregs remain intact. The lack of IL-27Rα or GR in Tregs resulted in a complete loss of the treatment effects. Inflammatory cell infiltration in the BAL was diminished only in Treg^WT mice but not in Treg^ΔIl27ra or in Treg^ΔGR mice (Fig 2A). Likewise, accumulation of cytokine-producing CD4 T cells in the lung was greatly reduced in Treg^WT mice, while the reduction was not found in Treg^ΔIl27ra or in Treg^ΔGR mice (Fig 2B). Histopathologic examination further supported the findings (Fig 2C). Therefore, direct signal through IL-27 or glucocorticoid receptors in Tregs is necessary to reverse Dex-resistant inflammation.

Figure 2. The lack of IL-27Rα and of GR expression in Tregs abolishes the treatment effects of Dex/IL-27.

Treg-specific Il27ra^−/− (Treg^ΔIl27ra), Treg-specific GR^−/− (Treg^ΔGR), and wild type control mice were sensitized, challenged, and treated with Dex/IL-27 as described in Fig 1. Mice were sacrificed 24 hours after the last challenge. (A) Inflammatory cell infiltration in the BAL and lung was determined. (B) Lung cells were stimulated ex vivo and intracellular cytokine expression was determined. (C) H&E staining and histology score of the lung tissues are shown (original magnification ×10). Each symbol represents individually tested animal. Data shown are the mean ± SD of 2 independent experiments (n = 6–8). *, p< 0.05, **, p< 0.01, ***, p< 0.001, as determined by Mann-Whitney nonparametric test.

IL-27 stimulation regulates GR expression

We next examined the molecular mechanisms by which IL-27 reverses steroid resistance via Tregs. Through an alternative splicing mechanism, GR mRNA generates two distinct isoforms, GRα and GRβ (15). GRβ does not bind the ligand, acting as an endogenous negative regulator of glucocorticoid signaling, and increased GRβ expression was proposed as a mechanism underlying steroid resistance (16). In support, lung Tregs isolated from Dex-sensitive inflammation expressed higher GRa/GRb ratio than those from Dex-resistant inflammation, while Foxp3 expression was comparable between the groups (Fig 3A). IL-27 stimulation significantly increased the GRα but not GRβ mRNA expression (Fig 3B), possibly rendering Tregs to be more responsive to Dex. Expression of the glucocorticoid-induced leucine zipper (Gilz), a Dex-responsive gene, remained unchanged by IL-27 stimulation (Fig 3B). To test whether GRβ alters Treg function and Dex responsiveness, we overexpressed GRβ in Tregs. GRβ overexpression did not alter surface expression of Treg associated markers, including GITR, CD25, PD1, Nrp1, CTLA4, and Foxp3, regardless of Dex stimulation (data not shown). We then stimulated GRβ overexpressing Tregs with Dex and determined the expression of Dex responsive genes, Fkbp5 and Gilz. Dex increased the expression in control Tregs; however, such increase was significantly diminished in GRβ overexpressing Tregs (Fig 3C). Dex pre-stimulation in Tregs greatly enhanced Treg suppression in vitro (pcDNA+Dex); however, Dex-induced elevation of Treg suppression was lost when GRβ is overexpressed (Fig 3D). Notably, GRβ overexpression itself did not affect the basal level Treg suppression, as control and GRβ expressing Tregs were equally suppressive without Dex stimulation (Fig 3D), suggesting that GRβ-dependent effects in Treg function occurs during steroid treatment.

Figure 3. Overexpression of GRβ Inhibits the immunosuppressive function of Tregs by Dex.

(A) Lung infiltrating Tregs were FACS sorted from mice sensitized with CA in CFA- or alum-adjuvant and challenged with CA for 4 consecutive days. Foxp3, GRa, and GRb mRNA expression was determined by qPCR analysis. (B) In vitro generated Foxp3⁺ Tregs (iTregs) were restimulated with or without IL27 for 72 h. Gra, Grb, and Gilz mRNA expression was determined by qPCR analysis. (C) Naive CD4⁺ T cells isolated from Foxp3^GFP mice were nucleofected with pcDNA (control vector) or GRβ expressing vector, and then activated in vitro under iTreg cell conditions with or without Dex for 72 h. Expression of the Fkbp5 and Gilz was determined by qPCR. (D) iTregs generated were cocultured with CFSE labeled naïve CD4 T cells at different ratios and stimulated using mouse T-activator CD3/CD28 dynabeads. Three days after stimulation, CD4 T cell proliferation was measured. Suppression was determined based on control CD4 T cell proliferation without Treg coculture. Data shown are the mean ± SD of 2–3 independent experiments. *, p< 0.05, **, p< 0.01, ***, p< 0.001 as determined by Mann-Whitney or by Kruskal-Wallis nonparametric test.

GRβ expression in Tregs determines in vivo steroid responsiveness

To investigate the role of GRβ in Treg function in vivo, we compared Treg’s ability to control allergic inflammation with or without GRβ overexpression. To this aim, we adoptively transferred Tregs into Treg-depleted mice (Supp Fig 3A). Foxp3^DTR sensitized with CA in alum to induce steroid-sensitive airway inflammation were treated with DTx to deplete endogenous Tregs. The mice then received control or GRβ-overexpressing CD45.1⁺ iTregs and further treated with Dex during CA challenge. Dex treatment attenuated eosinophil infiltration in the BAL in control but not in GRβ-overexpressing Treg recipients (Supp Fig 3B). Effector CD4 T cell accumulation in the lung was similarly diminished in control but not in GRβ-overexpressing Treg recipients (Fig 4A). Histopathologic examination supported the finding that the recipients of GRβ overexpressing Tregs displayed Dex resistance (Fig 4B). Dex’s inability to reduce inflammation was not due to defects in Treg survival, trafficking, or stability. The total numbers of transferred Tregs and those Tregs that have lost Foxp3 expression (ex-Tregs) were comparable between the groups (Fig 4C and Supp Fig 3C). Moreover, expression of Treg associated markers, such as PD-1, GITR, CD25, and Foxp3, was not affected by GRβ overexpression (Supp Fig 3D). Therefore, these results demonstrate that Dex responsiveness in Tregs may play a key role in steroid sensitivity. Lastly, iTregs adoptively transferred into mice induced for steroid resistant inflammation were unable to reverse steroid responses (Fig 4D and Supp Fig 4). Since adoptively transferred Tregs are fully able to mediate steroid responsiveness and to attenuate inflammation (7), these results strongly suggest that a Treg-extrinsic factor(s) appears to be involved in rendering Tregs refractory to steroid stimulation.

Figure 4. GRβ overexpression in Tregs induces steroid resistance in vivo.

B6 Foxp3^DTR mice were induced for steroid-responsive eosinophilic airway inflammation. DTx was injected at CA challenge, and vehicle or Dex was intraperitoneally injected on the first and third days of CA challenge (Supp Fig 4A). The mice received 2×10⁶ in vitro generated GRβ overexpressing or control FACS sorted CD45.1⁺ Foxp3 (GFP)⁺ iTreg cells 2 days before antigen challenge. (A) Cytokine expressing CD4 T cell accumulation in the lung was calculated by intracellular cytokine staining. (B) H&E staining and histology score of the lung tissues are shown (original magnification ×10). (C) Foxp3 expression of iTregs was determined. (D) In vitro generated Foxp3⁺ Tregs were sorted and adoptively transferred into recipients induced for neutrophilic inflammation one day prior to CA challenge (2 × 10⁶ cells per mouse). Mice were also treated Dex as above. Each symbol represents individually tested mouse from three independent experiments (n = 6–12). *, p< 0.05, **, p< 0.01, ***, p< 0.001, ****, p< 0.0001 as determined by Kruskal-Wallis nonparametric test.

In the present study we found that IL-27 provision reverses steroid resistance and that the presence of Tregs is instrumental for Dex/IL-27 to attenuate the inflammation. The fact that GR or IL-27Rα deficiency in Tregs abrogates the Dex/IL-27’s ability to control inflammation supports the importance of direct Dex and IL-27 signaling in Tregs. Interestingly, IL-27 enhances GRα expression in Tregs, tipping the balance of GRα/GRβ ratio and rendering Tregs to better respond to Dex. Indeed, Dex stimulation enhances Treg suppressive function in vitro, and GRβ overexpression abolishes Dex’s ability to do so. Moreover, GRβ overexpression in Tregs alone is sufficient to convert steroid-sensitive inflammation to a steroid-resistant form, strongly suggesting that Tregs, precisely Treg steroid responsiveness, may determine steroid sensitivity.

The precise mechanisms by which GRβ expression is induced remain to be determined. As inflammatory cytokines such as IL-17 and IL-23 increase the GRβ expression (17, 18), it is possible that GRβ expression is increased under CFA-sensitized neutrophilic inflammatory responses, eliciting steroid resistance. Although our finding demonstrates that increased GRβ expression in Tregs is sufficient to nullify steroid sensitivity in vivo, it is possible that different cell types may be involved in mediating steroid-mediated anti-inflammatory effects during different types of inflammation (19). Therefore, factors and mechanisms involved in steroid resistance may be non-overlapping and possibly distinct depending on the type of inflammation and of cells. The finding that IL-27 enhances GRα expression in Tregs is interesting and it demands further investigation. Wenzel and colleagues reported that IL-27 expression, in combination with a type 2 immunity signature, may identify severe asthma phenotype (20). Of note, IL-27 level in BAL cells from the tested severe asthmatic patients is highly heterogeneous (20). It is thus possible that those severe asthmatic patients with lower IL-27 level might be associated with steroid resistant endotype in asthma. Identifying the mechanisms underlying IL-27-dependent regulation of steroid responses via Tregs could thus be critical for the development of novel therapies to effectively treat steroid resistant inflammation.

Key points.

• Supplementing IL-27 during Dex treatment attenuates steroid-resistant inflammation.

• IL-27 and Dex directly act on Tregs to reverse steroid-resistant inflammation.

• IL-27 upregulates agonistic glucocorticoid receptor expression in Tregs.

Abbreviations used:

Treg

regulatory T cell

Dex

dexamethasone

DTx

diphtheria toxin

CFA

complete Freund’s adjuvant