The 2′-3′ and 3′-3′ cGAMP were purchased from InvivoGen (San Diego, CA, USA). HDM extracts (Mite Extrac-Df; LG5339, extract from house dust mites, Dermatophagoides farinae, 1.7 μg endotoxin/mg) were purchased from LSL (Osaka, JAPAN). Amlexanox was purchased from the Tokyo Chemical Industry and Takeda Pharmaceutical Company (Osaka, Japan). Dexamethasone was purchased from Sigma Aldrich (St. Louis, MO, USA).

Six-week-old female C57BL/6J mice were purchased from CLEA Japan Inc. (Osaka, Japan). Sting^−/− mice were purchased from Phoenix Bio (Hiroshima, Japan). Il-33^−/−, Rag2^−/−, and St-2^−/− mice were obtained as previously described (14–16). Tslp^−/− mice were kindly provided by Dr. Ziegler (Immunology Program, Benaroya Research Institute, Seattle, WA, USA). Irf7^−/− mice were provided by the RIKEN BioResource Center through the National Bio-Resource Project of the Ministry of Education, Culture, Sports, Science and Technology (Ibaraki, Japan) (17). Irf3^−/−Irf7^−/− mice were generated by cross-breeding Irf3^−/− mice with Irf7^−/− mice, and Tnf ^−/−Tbk1^−/− mice were obtained as previously described (18). All animal experiments were conducted in accordance with the institutional guidelines and with the approval of the ethics committee for the National Institute of Biomedical Innovation, Health and Nutrition (NIBIOHN, Approval ID: DNA301).

Mice were anesthetized by ketamine and xylazine followed by the intranasal administration of 1 μg HDM, 1 μg cGAMP, or HDM + cGAMP in 30 μl of PBS at day 0, and then challenged with 1 μg HDM four times (at days 7, 9, 11, and 13). Samples [serum, broncho-alveolar lavage fluid (BALF), lymph nodes, and lungs] were collected 24 h after the last HDM challenge.

Stocks of influenza virus (A/PR8/34) were grown in Madin-Darby canine kidney cells. Mice were infected intranasally with 140 plaque-forming units of influenza virus. Stocks of purified RSV (strain A2, ATCC) were grown in HEp-2 cells. Inactivation was performed by heating the virus to 56°C for 30 min. For infection, mice were anesthetized, and an intratracheal instillation with 2.5 × 10⁵ plaque-forming units of RSV was performed.

To obtain suspensions of single lung cells, mouse lungs were removed, minced, and digested in RPMI 1640 with 200 μg/ml DNase (Roche, Penzberg, Germany) and 200 units/ml collagenase (Wako, Osaka, Japan) and then incubated at 37°C for 1 h. The lungs were homogenized with the gentleMACS™ Dissociator (Miltenyi Biotech, Gladbach, Germany), and debris was removed by using a 70-μm separation filter. The cells were washed twice with phosphate-buffered saline (PBS) and then resuspended in Dulbecco's modified Eagle's medium with 10% fetal calf serum and 5 μg/ml of insulin and seeded into 75 cm² culture flasks (Corning, Corning, NY, USA). Medium was replaced after 24 h, and then every 2 days. Adherent cells were used as lung fibroblast cells. These cells were incubated with cGAMP (1 μg/ml or 10 μg/ml) for 6, 24, 48 h, after the cytokine content of the resulting cell lysates were measured. Cell lysates were prepared with three freeze-thaw cycles.

Broncho-alveolar lavage fluid was collected by two lung lavages with 0.7 ml of PBS each via a tracheal cannula. The collected BALF was centrifuged (2,000 × g, 5 min, 4°C) to separate the cells and fluids, and then the cells were counted in a hemocytometer. Next, the cells were incubated with anti-mouse CD16/32 antibody, and then stained with anti-CD11c, anti-CD11b, anti-Siglec F, anti-F4/80, anti-Ly6G, anti-CD3, and anti-B220 antibodies. All labeled antibodies were purchased from BD Biosciences or BioLegend. The samples were analyzed by flow cytometry (LSRII, BD Biosciences, Franklin Lake, NJ, USA). The resulting data were analyzed using FlowJo software.

Single-cell suspensions of bronchial lymph nodes (2 × 10⁵ cells in a 96-well plate) were cultured in RPMI (containing 10% fetal calf serum and 1% penicillin/streptomycin) with or without HDM (10 μg/ml) for 5 days. Enzyme-linked immunosorbent assays (ELISAs) (described below) were used to assess the IL-5 and IL-13 levels in the supernatants.

To measure the HDM-specific IgG1 levels, serial dilutions of sera were prepared in 96-well plates coated with 10 μg/ml of HDM. Horseradish peroxidase-conjugated anti-mouse IgG1 (Southern Biotech, Birmingham, AL, USA) was used as the secondary antibody. The total IgE was measured by a mouse IgE ELISA Quantitation kit (Bethyl Laboratories Inc., Montgomery, TX, USA) according to the manufacturer's protocol. Cytokines (IL-5 and IL-13) were quantified by ELISA kits (R&D Systems, Minneapolis, MN, USA) according to the manufacturer's protocols.

The double-stranded (ds)DNA concentrations were measured by using Qbit dsDNA HS Assay kits (InvitroGen, Carlsbad, CA) according to the manufacturer's protocol.

Lung tissues were dissected and fixed in 4% paraformaldehyde overnight. To prepare the paraffin sections, tissues were gradually dehydrated and embedded in paraffin. The sections (5 μm thick) were prepared and stained with hematoxylin and eosin (H&E) or periodic acid-Schiff (PAS) using a Periodic Acid-Schiff kit (Sigma Aldrich) according to the manufacturer's protocol. Representative images were acquired with a ScanScope® AT slide scanner (Leica, Wetzlar, Germany).

Airway hyper-responsiveness was measured by whole body plethysmography (Data Sciences International, New Brighton, MN, USA) (19). Mice received nebulized PBS, followed by increasing concentrations of methacholine (3.125, 6.25, 12.5, 25, and 50 mg/ml) to induce bronchoconstriction.

Data are presented as means alone or means ± SD. Two group comparisons were performed using Mann–Whitney U-tests. One-way analysis of variance with Tukey's multiple comparison tests was used to compare multiple groups. Statistical analyses were performed using Prism6 (Graph-Pad Software), and differences were considered statistically significant when p < 0.05 (indicated in figures as ^*p < 0.05 and ^**p < 0.01).

Viral infections impact the host immune system not only by releasing viral pathogen-associated molecular patterns but also by leading host cells to release DAMPs, both of which have been shown to be involved in the induction and exacerbation of asthma (20). To examine viral infection induce host DNA as a part of DAMPs in infected lung tissue, mice were intratracheally inoculated with RSV, followed by BALF collection. Although RS virus is a RNA virus, dsDNA was detected in the BALF (Figure 1A). In addition, influenza virus infection also induced dsDNA in the BALF (Figure 1B). These results suggest that host dsDNA as DAMPs can be induced by RNA virus airway infections and might be associated with infection-induced lung inflammation and cell death, as well as the severity of these conditions.

Because cGAS binds dsDNA and generates the second messenger cGAMP (9), we utilized cGAMP as an alternative stimulus for DNA. To assess the impact of cGAMP in allergic inflammation in mice, we intranasally sensitized mice with HDM alone, cGAMP alone, or a combination of HDM and cGAMP (HDM + cGAMP), and 7 days later, we challenged the mice with HDM four times every other day (Figure 2A). On day 14, the mice sensitized with HDM + cGAMP had significantly higher serum levels of HDM-specific IgG1 and total IgE (Figure 2B), and HDM-specific Th2 cytokines in the lymph nodes (Figure 2C), than those sensitized with HDM alone or with cGAMP alone. When we examined the numbers of total cells, eosinophils, alveolar macrophages, neutrophil, B cells, and T cells in BALF, significantly higher numbers were observed in the mice sensitized with HDM + cGAMP than in the control mice (Figure 2D). The number of eosinophils recruited into the lung was dependent on the dose of cGAMP, suggesting that these responses were directly associated with the dose of cGAMP during the sensitization phase (Supplementary Figure 1A). Notably, there are two types of cGAMP, based on their structure. Mammalian cGAS produces 2′-3′ cGAMP, while bacterial cGAS produces 3′-3′ cGAMP. However, we obtained similar results using both types of cGAMPs (Supplementary Figure 2).

Histological analyses revealed severe inflammatory cell infiltration and goblet cell hyperplasia in the lungs of mice sensitized with HDM + cGAMP (Figure 2E). In addition, mice sensitized with HDM and cGAMP significantly increased the airway hyper-responsiveness to methacholine in vivo compared to the mice sensitized with HDM alone or treated with cGAMP alone (Figure 2F). Of note, we chose a relatively low dose (1 μg) of HDM for the sensitization because a higher dose of HDM induced inflammatory responses in the absence of cGAMP, while a singular administration of 1 μg of HDM failed to induce inflammatory responses in the absence of cGAMP (Supplementary Figure 1B). The allergic inflammation in the lung induced by repeated intranasal administration of high dose of HDM is known to be dependent on TLR4 (21), but our model likely differs from the conventional HDM-induced allergy model, as the allergic inflammation in our model was independent of TLR4 (Supplementary Figure 3). These results suggest that cGAMP can be one of the aggravating factors for HDM-induced allergic responses in the airway, especially under certain conditions in which DNA and cGAMP are physiologically relevant.

IL-33, IL-25, and TSLP are known to be pro-allergic cytokines, and they are released from various cells, including epithelial cells (22). Notably, IL-33 and ST2/IL1RL1 (IL-33 receptor) genes are considered to be asthma disease susceptibility genes and have been identified in genome-wide association studies (23). During lung inflammation, IL-33 is mainly released from pulmonary epithelial cells to elicit the recruitment and activation of immune cells that induce Th2 cytokines. These responses seem to be associated with the onset and disease pathologies of allergy (24). When we stimulated mouse lung fibroblast cells in vitro with cGAMP, the cell-associated IL-33, but not the IL-33 in the supernatant (data not shown), was highly upregulated within 6 h (Figure 3A). Given that cGAMP induces IL-33, we next investigated the role of IL-33 in cGAMP-adjuvanted, HDM-induced allergic inflammation using Il-33^−/− mice. Antibody responses and eosinophil recruitment into the lung were significantly decreased in Il-33^−/− mice compared with those in Il-33^+/− mice (Figures 3B,C). To further confirm the importance of IL-33, we carried out similar experiments using St2^−/− mice, which lack the IL-33 receptor, and found that, similar to Il-33^−/− mice, the antibody responses and eosinophil infiltration were significantly lower in St2^−/− mice than those in St2^+/− mice (Figures 3D,E). Although TSLP is known to be an important molecule in allergic inflammation, it is not involved in the antibody responses and inflammatory cell recruitment into the lung (Figures 3F,G). These results suggest that IL-33 and its receptor, ST2, are required for cGAMP-adjuvanted, HDM-induced allergic inflammation in mouse lungs.

IL-33 is a potent activator of group 2 innate lymphoid cells (ILC2) in the lung (25). Allergic inflammation induced by intranasal IL-33 administration is dependent on the activation of ILC2s but not on that of T or B cells, and IL-33 induced by papain stimulates strong airway eosinophilia in Rag2^−/− mice (26, 27). However, in our experimental system, the numbers of total cells and eosinophils in the BALF were significantly reduced in Rag2^−/− mice compared with those in Rag2^+/− mice (Figure 4A). This suggests that, unlike allergic inflammation induced by papain exposure, cGAMP-adjuvanted, HDM-induced allergic inflammation is mediated by adaptive T and/or B cells. To further confirm the role of IL-33 in acquired immunity in this allergic inflammation model, we next examined the HDM-specific cytokine production from the draining lymph nodes. As expected, cells from Il-33^−/− mice displayed lower levels of Th2 cytokines in response to HDM than cells from Il-33^+/− mice (Figure 4B), suggesting that acquired immune cells are necessary for cGAMP-adjuvanted, HDM-induced allergic inflammation in the lung.

Because cGAMP is a ligand for STING and induces type-I interferon through TBK1 and IRF3 activation, we examined the importance of this signaling pathway for cGAMP-adjuvanted, HDM-induced allergic responses using Sting^−/−, Tbk1^−/−, and Irf3^−/−Irf7^−/− mice. As expected, we observed that the amount of eosinophil recruitment into the lungs was significantly lower in all of the tested gene-deficient mice than that in the control mice (Figures 5A–C). Moreover, the intracellular IL-33 level in lung fibroblast cells stimulated by cGAMP was not induced in cells from Irf3^−/−Irf7^−/− mice (Figure 5D).

Corticosteroids are key drugs for clinical asthma treatment (28), so we tested if corticosteroids were effective in our cGAMP-adjuvanted, HDM-induced allergic inflammation model. Dexamethasone (DEX) treatment during the elicitation phase decreased the number of eosinophils in the BALF compared with PBS treatment. Furthermore, histological analyses showed reduced goblet cell hyperplasia in the lung after DEX treatment compared with that after PBS treatment (Supplementary Figures 4A–C). Molecular-targeted therapy, such as anti-cytokine antibodies and anti-IgE antibodies, have recently been proposed for the treatment of allergic diseases (29, 30). Our results suggest that the STING-TBK1-IRF3 axis might be an important therapeutic target for cGAMP-induced allergic inflammation. Amlexanox is a small molecule and an active pharmaceutical ingredient of a drug used to treat asthma and allergic rhinitis in Japan. This drug functions as a TBK1 inhibitor (31). Therefore, we tested amlexanox as a molecular-targeted therapy drug in our model. Intranasal amlexanox administration during both the sensitization and elicitation phases significantly decreased inflammatory cell recruitment into the lung as well as the serum total IgE and HDM-specific IgG1 responses (Figures 5E–G). Similar results were observed in mice treated with amlexanox by daily gavage (Supplementary Figure 5). These results suggest that a TBK1 inhibitor can improve cGAMP-adjuvanted, HDM-induced asthma.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.