Before investigating the consequences of bacterial airway infection on AAI, we established a murine infection model to first study the immune responses triggered by M. catarrhalis. Two hours after intranasal inoculation, infection was fully established and bacterial titers of the lung dropped massively at day 2 p.i. and were below detection at day 6, revealing clearance of M. catarrhalis in normal animals (Figure 1A). It has to be noted that approx. 50% of wt animals intranasally infected with M. catarrhalis showed moribund symptoms at day 3 p.i. (Figure 2G). Total cell numbers in bronchoaleveolar lavage (BAL) increased 1 day after intranasal inoculation, peaked at day 4, and then decreased. M. catarrhalis infection caused airway inflammation primarily through massive influx of neutrophils and lower numbers of eosinophils and macrophages reaching their maximum at day 4, followed by an increase in lymphocytes which peaked at day 7 p.i. Twelve days after infection, all BAL cells numbers returned to baseline level (Figures 1B,C). Accordingly, early and strong expression of CXCL1 and CXCL10 in the BAL was observed, both known to attract neutrophils and other inflammatory cells (Figure 1D). As M. catarrhalis infection causes airway inflammation (9, 13), we performed a kinetic study of inflammatory cytokines in the BAL. While IL-6, TNF-α and IL-1β peaked at day 1 p.i. and rapidly decreased to low levels by day 4, the amounts of IFN-γ and IL-17 were highest at day 7 p.i., coinciding with the peak of BAL lymphocytes (Figure 1E). Due to the late occurrence of IL-17 in the BAL, we continued to analyze the mRNA expression of this cytokine in lung homogenates. In accordance with BAL data, strong IL-17 mRNA expression was found at day 7, correlating with the rise of IL-17⁺CD4⁺ T cells at this time point (Figures 1F,G). To identify the type of T cells producing pulmonary IL-17 after M. catarrhalis infection, lung lymphocytes were analyzed by flow cytometry. The majority of IL-17 secreting cells in the lung were conventional CD4⁺ T cells (70%) which increased between days 7 and 15 p.i. and only a minor fraction of TCR γ⁺δ⁺ cells (14%) was positive for IL-17 (Figure 1H). We next investigated inflammatory responses in IL-17 deficient mice. Interestingly, despite similar initial bacterial colonization of the lungs, IL-17 KO mice showed increased bacterial clearance as compared to wt animals (Figure 2A). Furthermore, these mice revealed significantly reduced total and differential BAL cell counts (Figures 2B,C) and lung tissues showed a massive reduction of cellular infiltrates as compared to wt animals (Figure 2D). The amounts of IL-6 and TNF-α in BAL and serum of infected IL-17 KO were diminished and mice did not succumb to M. catarrhalis infection wt animals (Figures 2E–G). Interestingly, despite the lack of IL-17, the number of neutrophils in the lung at day 1 p.i. was comparable to wt animals. Similar amounts of the CXCL1, 5, and 10 chemokines in IL-17 KO mice might compensate for early neutrophil attraction, independently in IL-17 (Figure S1 in Supplementary Material). Likewise, neutralization of TNF-α, which is also produced by epithelial cells upon M. catarrhalis infection (Figure S2 in Supplementary Material), resulted in reduction of local and systemic IL-1β, IL-6, and IL-17 and subsequent protection against M. catarrhalis infection (Figures 3A–C). Interestingly, in contrast to IL-17 and TNF-α, we found that neutralization of IL-6 increased local and systemic TNF-α as well as early IL-1β and IL-17 levels in the BAL. Anti-IL-6 treatment resulted in slightly enhanced numbers of animals with moribund symptoms, suggesting that IL-6 exerts an anti-inflammatory function during pulmonary M. catarrhalis infection (Figures 4A–C).

To study the mechanisms of M. catarrhalis caused exacerbation of AAI, wt C57Bl/6 mice (white bars) were intranasally infected with M. catarrhalis after the second exposure with HDM and analyzed on day 23 as shown in Figure 5A. While HDM exposure induced airway inflammation, additional infection of HDM sensitized wt mice with M. catarrhalis led to hugely increased numbers of leukocytes in the BALF mainly through infiltrating neutrophils, eosinophils, and lymphocytes (Figures 5B,C). This was also evident in lung histology, showing increased airway inflammation, goblet cell hyperplasia, and mucus production in M. catarrhalis infected, HDM-allergic mice (Figures 5D–F). To further study the cytokine pattern of pulmonary CD4⁺ T-cell populations, we compared HDM allergic mice with those that additionally contracted an infection. The latter group revealed significantly increased frequencies of IFN-γ, IL-17A, and IL-5⁺/IL-13⁺ double positive CD4⁺ T cells, although the absolute frequencies of IL-5⁺/IL-13⁺ Th2 cells were low as compared to those of IFN-γ⁺ and IL-17⁺CD4⁺ T cells (Figure 5G; Figure S5A in Supplementary Material). Accordingly, increased amounts of IL-17 and IFN-γ were found in the lungs of HDM allergic wt mice (white bars) concomitantly infected with M. catarrhalis (Figure S5B in Supplementary Material). Of note, cellular infiltrates and cytokine responses measured at day 23 were significantly lower than during acute infection with M. catarrhalis.

We then wondered whether the time of M. catarrhalis infection (prior to, during or after complete HDM sensitization) has an influence on exacerbation of AAI. Similar to infection during HDM sensitization, bacterial infection after established allergic reactions increased total BAL leukocyte counts mainly by neutrophils and lymphocytes. Together with increased lymphocytic infiltration, airway infection enhanced the frequencies of lung IL-17, IFN-γ, and IL-5 secreting CD4⁺ T cells as compared to with the HDM-only group (Figure S3 in Supplementary Material). Furthermore, we wondered whether a previously resolved infection with M. catarrhalis is still able to exacerbate HDM allergic reactions. Similar to infection during or after HDM sensitization, infection prior to allergen exposure triggered airway infiltrations of CD4⁺ T cells secreting increased amounts of IL-17, IFN-γ, and IL-5 (Figure S4 in Supplementary Material).

Since IL-17 has been reported to be associated with mucoepithelial infections and allergic responses (14), we next tested IL-17 deficient mice for the development of HDM-allergic airway responses. Interestingly, wt animals (white bars) and IL-17 KO mice (black bars) revealed similar numbers of neutrophils, eosinophils, and lymphocytes in the BALF (Figures 5B,C). Lung histology revealed slightly reduced pulmonary infiltrates in IL-17 KO mice as compared to with wt mice after HDM exposure (Figure 5D). Accordingly, airway inflammation, goblet cell hyperplasia, and mucus production were marginally diminished in IL-17 KO mice as compared to with controls (Figures 5E,F).

As early antibacterial immunity is mediated by IL-17-controlled influx of neutrophils into infected tissues (15), we next investigated the role of this cytokine to trigger infection-induced exacerbation of AAI. IL-17 KO and wt mice were infected during the second HDM allergen exposure and analyzed at day 23, as shown in Figure 5A. The data demonstrate that in contrast to wt animals, the number of neutrophils, eosinophils, and lymphocytes in the BALF of infected IL-17 KO mice was comparable to the non-infected, HDM group. Furthermore, comparable numbers of mucus secreting cells and similar inflammation scores suggest IL-17 as key cytokine of M. catarrhalis-induced exacerbation of AAI (Figures 5D–F). Yet, M. catarrhalis infection of HDM allergic IL-17 KO mice still enhanced the frequency of pulmonary IFN-γ and IL-5/IL-13 secreting CD4⁺ T cells (Figure 5G; Figures S5A,B in Supplementary Material).

Taken together, these findings show that IL-17 essentially contributes to M. catarrhalis triggered exacerbation of AAI.

It has been shown that neutrophil recruitment during inflammation is mediated by synergistic effects of IL-17 and TNF-α in endothelial cells by enhancing the expression of P- and E-selectin (16). We thus wondered whether neutralization of TNF-α also attenuates infection-induced exacerbation of AAI. Briefly, HDM sensitized mice were treated only once with anti-TNF-α mAb 4 h before intranasal infection with M. catarrhalis and analyzed at day 23, as depicted in Figure 6A. Control groups included mice that received phosphate-buffered saline (PBS) or HDM alone as well as infected, HDM sensitized animals.

Differential cell counts of BAL cells revealed that predominantly neutrophils were reduced by anti-TNF-α treatment, while the numbers of macrophages were increased when compared with HDM/infected controls. The increase of BAL macrophages after TNF-α neutralization seems to compensate for the low number of neutrophils in order to eliminate M. catarrhalis. Interestingly, the number of eosinophils or lymphocytes were not significantly reduced after TNF-α treatment as shown in Figures 6B,C. Lung histology confirmed that anti-TNF-α treatment vastly reduced infection-triggered inflammatory infiltrates in HDM allergic animals which was also reflected by the decreased number of PAS⁺ cells and the low-inflammation score (Figures 6D–F). We next wondered whether TNF-α neutralization also affects pulmonary T-cell activation. Anti-TNF-α treatment attenuated the activation of IFN-γ⁺ and IL-17⁺ lymphocytes but had no influence on the low frequency of IL-5/IL-13 secreting, pulmonary CD4⁺ T cells (Figure 6G; Figure S6 in Supplementary Material).

In summary, the experiments revealed that a single application of anti-TNF-α mAb is sufficient to markedly attenuate M. catarrhalis-induced exacerbation of AAI.

Female 8–10-weeks-old C57BL/6J mice were purchased from Charles River Laboratories, Germany. Female IL-17AF^−/− mice on C57BL/6J background were kindly provided by Prof. Dr. Immo Prinz (Hannover, Germany) and housed in a pathogen-free facility with individually ventilated cages. All experiments were approved by the Regierunspräsidium Gießen (Protocol no. MR 20/6 Nr. 123/2012).

A patient isolate of M. catarrhalis was stored at −80°C in glycerol. Bacteria from frozen stocks were grown aerobically in Columbia blood agar plates (BBL; Becton Dickinson, MD, USA) with 5% sheep blood at 37°C overnight, washed off the plate and resuspended in sterile PBS for infection. Anesthetized mice were inoculated intranasally (i.n.) with 2 × 10⁸ CFU M. catarrhalis resuspended in 50-µL PBS, before during or after HDM challenge.

Lungs were aseptically removed and homogenized in 1 mL of sterile PBS.

Serial dilutions of BAL fluid and lung homogenates were prepared in sterile PBS, plated on Columbia blood agar plates (BBL; Becton Dickinson, MD USA) with 5% sheep blood and incubated overnight at 37°C. Colonies were enumerated and bacterial numbers per lung calculated.

8–10-week-old C57Bl/6 mice were slightly anesthetized (Ketamin/2%Rompun, i.p.) and weekly challenged by intranasal (i.n.) administration of 100-µg HDM extract (GREER, Lenoir, USA) in 50-µL PBS on days 0, 7, 14, and 21. Two days after last HDM challenges mice were sacrificed and blood, bronchoalveolar lavage (BAL) fluids, lungs were collected for further analysis.

To assess the role of IL-6 and TNF-α during the infection with M. catarrhalis, neutralizing anti-IL-6 (clone MP5-20F3) or anti-TNF-α (clone MP6-XT22) antibodies were used. Rat IgG1 isotype control was applied for treatment of control mice. Four hours before M. catarrhalis infection, mice were treated intraperitoneally once with 100 µg of neutralizing antibodies, respectively. All used antibodies were generated and purified by Manuela Staeber (Max Planck Institute for Infection Biology, Berlin, Germany).

Bronchoaleveolar lavage was performed 48 h after the last challenge using 1-mL cold PBS supplemented with a protease inhibitor cocktail (Roche, Mannheim, Germany). An automated Casy TT cell counter (Schaerfe System, Reutlingen, Germany) was used to determine the total leukocyte cell counts. Cells were centrifuged and the cell-free supernatant was stored at −20°C until cytokines were measured by Elisa. For differential cell counts, Cytospin preparations were fixed and then stained with Diff-Quick (Merz &Dade AG, Dudingen, Switzerland). Macrophages, lymphocytes, eosinophils, and neutrophils were identified by standard morphologic criteria and 300 cells were counted per cytospin. The measurement of cytokines IFN-γ, IL-17, and chemokines (CXCL1, CXCL5, and CXCL10) was performed by Cytometric Bead Array (BD, Heidelberg, Germany) according to manufacturer’s protocol.

Directly after BAL, lungs were fixed with 10% formalin via the trachea, removed, and stored in 10% formalin. Lung tissues were embedded into paraffin. Tissues were cut in 3-µm sections and stained using hematoxylin and eosin (HE) as well as periodic acid Schiff (PAS). Goblet hyperplasia was assessed by cell counting and expressed as the number of goblet cells per 100-µm basement membrane, according to Foster et al. (2000).

The amount of HDM-specific IgG1 and IgG2c antibodies was measured by the ELISA. Maxi-Sorb plates were coated with 50-µg/mL HDM extract in bicarbonate buffer (pH 9.6) overnight at 4°C. Subsequently, coated wells were blocked with 1% w/V BSA in PBS for 2 h at RT. After washing, serum samples (1:100) were incubated overnight at 4°C, washed, and incubated with the corresponding biotin-labeled anti-immunoglobulin antibody overnight at 4°C. Plates were washed and incubated with streptavidin-peroxidase for 30 min at RT. After development with substrate, the reaction was stopped using H2SO4 and ODs were read at 450 nm.

TNF-α, IL-6, IL-1β, IL-5, IL-13, IFN-γ, and IL-17 were measured in cell-free lavage fluid by Cytometric Bead array (CBA; BD BIO-sciences; San Diego, CA, USA) as described by the manufacturer.

Expression of IFN-γ and IL-17 mRNA was determined by quantitative real-time polymerase chain reaction (RT-PCR) analysis using LightCycler technology (Roche). RNA was extracted from lung tissues according to manufacturer’s instructions. cDNA synthesis was performed after removal of contaminating genomic DNA by DNase treatment using Superscript reverse transcriptase (Invitrogen, Karlsruhe, Germany) as described by manufacturer. Quantitative PCR (QPCR) was performed by using the QuantiTect SYBR Green PCR kit (QIAgen) and primers specific for IFN-γ (forward: 5′ GCT TTG CAG CTC TTC CTC AT 3′ and reverse 5′ GCA GGA TTT TCA TGT CAC CA 3′), IL-17 (forward: 5′ AAG GCA GCA GCA ATC ATC CC 3′ and reverse 5′ GGG TCT TCA TTG CGG TGG AG 3′) Results were expressed as x-fold expression in relation to the house keeping gene L-32.

Lungs were cut into 2–5 mm pieces and incubated for 30 min at 37°C in the incubation medium (IM) (RPMI-1640 with L Glutamine and NaHCO3, 10% FCS, 16 non-essential amino acids (PAA, Pasching, Austria), 100 mg/mL streptomycin, 120 mg/mL penicillin) supplemented with 1-mg/mL Collagenase D (Roche, Basel, Switzerland) and 20 µg/mL DNase I (Roche, Basel, Switzerland). The pre-digested lungs were minced through 100-mm nylon cell strainer (BD Falcon, NJ, USA) diluted with IM and centrifuged at 1,700 rpm for 5 min. Cell pellets were resuspended in erythrocytes lysis buffer (8 g/L NH₄Cl; 1 g/L KHCO₃; 37.2 mg/L EDTA) and incubated at room temperature for 3–5 min. Cells were washed, centrifuged, and resuspended in IM supplemented with 50 ng/mL PMA, 750 ng/mL ionomycin, and 10 mg/mL brefeldin A for 4 h. Extracellular staining was performed in PBS with 1% FCS and presence Fc blocking antibody (clone 2.4G2, BD, Heidelberg, Germany) and fluorophore-labeled antibodies (eBioscience, San Diego, CA, USA, if not specified otherwise). The following antibodies for extracellular T-cell staining was performed: anti-CD4-V450 (RM4-5), anti-CD8α-V500 (53-6.7), and anti-CD25-PE-Cy7 (PC61.5) (BD, Heidelberg, Germany). After extracellular staining, T cells were fixed with FoxP3-Fixation Kit (eBioscience, San Diego, CA, USA) and permeabilized with 0.3% Saponin, 1% FCS in PBS followed by an intracellular staining with the following fluorophore-labeled antibodies: IFN-γ-PerCP-Cy5.5 (XMG1.2, Biolegend, San Diego, CA, USA); IL-17-AlexaFluor 700 and FoxP3-AlexaFluor 700 (FJK-16 s); IL-5-PE (TRFK5); IL-13-AlexaFluor 647 [(eBio13A) (eBioscience, San Diego, CA, USA)]. Every staining included negative and isotype controls. Fluorescence signals were acquired by flow cytometry (FACSAria III; BD, Heidelberg, Germany) and analyzed using FACSDivaTM software.

Data are expressed as ± SEM. We performed one-way analysis of variance (ANOVA) statistical analysis followed by Bonferroni HSD test using GraphPad Prism 5 software. The values of P < 0.05 were considered to be significant.