The Innate Lymphoid System Is a Critical Player in the Manifestation of Mucoinflammatory Airway Disease in Mice

In the healthy respiratory tract, the inhaled biotic or abiotic materials and host-derived end products are effectively cleared by a coordinated effort of the airway mucociliary clearance (MCC) system, resident immune cells, and secreted antimicrobial peptides (1). Cystic fibrosis, a mucoinflammatory lung disease, is characterized by airway surface liquid (ASL) layer depletion, excessive mucus production, mucus obstruction, impaired MCC, and airway inflammation (2). As a consequence of impaired MCC, cystic fibrosis patients are vulnerable to bacterial infections and pulmonary exacerbations because of viral infections (3). Our understanding of the molecular and cellular pathways leading to excessive mucus production in mucoinflammatory disease is incomplete, and a possibility of therapeutic inhibition of these pathways to ameliorate impaired MCC remains untested in experimental animal models.

The Scnn1b transgenic (Scnn1b-Tg⁺) mouse model of mucoinflammatory lung disease overexpresses the sodium channel nonvoltage gated 1, β subunit (Scnn1b) transgene in club cell–specific protein expressing epithelial cells. Scnn1b transgene overexpression produces ASL layer depletion, impaired MCC, excessive mucus accumulation, and airway inflammation (4–7). As a result of mucus obstruction, Tg⁺ neonates contract airspace bacterial infection in early postnatal life (6). The Tg⁺ neonatal mice also exhibit robust Th2 responses with marked mucous cell metaplasia (MCM) and mucus obstruction (7, 8). Our recent data indicate that the deficiency of IL-33, a regulator of Th2 immunity, in Tg⁺ juveniles resulted in a remarkable attenuation of MCM but with no obvious amelioration of mucus obstruction (9). Because mucus production has also been associated with Th1 inflammatory responses in the lungs (10), it remains unclear whether non-Th2 adaptive and innate lymphoid cells (ILCs) are involved in mucus production and/or obstruction in Tg⁺ mice.

In this study, we have tested the hypotheses that innate lymphoid and adaptive immune cells are essential for the manifestation of MCM and mucus secretory responses in Tg⁺ mice. ILCs (i.e., ILC1, ILC2, ILC3, and NK cells) represent the amnestic equivalence of adaptive immune cells Th1, Th2, Th17, and cytotoxic T cells, respectively (11–14). The availability of IL-2 receptor subunit γ (Il2rg) and recombination activating gene 1 (Rag1) knockout strains allow complete ablation of innate lymphoid (15) and adaptive immune systems, respectively (16). To investigate the contribution of, and interactions between, the innate lymphoid and adaptive immune systems in mucoinflammatory lung disease in Tg⁺ mice, ILC-deficient (Il2rg knockout mice [Il2rg^KO]), adaptive immune cell–deficient (Rag1 knockout mice [Rag1^KO]), and innate lymphoid/adaptive immune cell–deficient (Il2rg^KO/Rag1^KO) mice were generated on wild-type (WT) (absence of Scnn1b transgene) and Tg⁺ (presence of Scnn1b transgene) backgrounds. Designated mucoinflammatory end points, including airspace bacterial burden, airway inflammation, mucus obstruction, MCM, and gene expression, were examined. The results from this study provide novel insights into the independent and combined contribution of the innate lymphoid and adaptive immune systems in the pathogenesis of mucoinflammatory lung disease in Tg⁺ mice.

Scnn1b-Tg⁺ mice (Tg(Scgb1a1-Scnn1b)6608Bouc/J), Il2rg^KO (B6.129S4-Il2rg^tm1Wjl/J), and Rag1^KO (B6.129S7-Rag1^tm1Mom/J) were obtained from The Jackson Laboratory (Bar Harbor, ME). These three strains were cross-bred to generate ILC-deficient (Il2rg^KO), adaptive immune cell–deficient (Rag1^KO), and combined innate lymphoid and adaptive immune cell–deficient (Il2rg^KO/Rag1^KO) WT and Tg⁺ mice. To generate littermates of various genotypic combinations, double-heterozygous (Il2rg^HET/Rag1^HET) Scnn1b-Tg⁺ females were mated with double-knockout (Il2rg^KO/Rag1^KO) WT males. In reciprocal crossing, double-heterozygous (Il2rg^HET/Rag1^HET) Tg⁺ males were mated with double-knockout (Il2rg^KO/Rag1^KO) WT females. Of note, the Il2rg gene is located only on X chromosomes; although the Il2rg-sufficient males are hemizygous at Il2rg locus, for the sake of simplicity, we will address both Il2rg-sufficient males and Il2rg-heterozygous females as Il2rg^HET. Genotyping for all mice was performed by PCR for the determination of Scnn1b-Tg⁺, Il2rg, and Rag1 alleles (data not shown). Mice were maintained in individually ventilated, hot-washed cages on a 12-h dark/light cycle and were supplied regular diet and water ad libitum. Mice were housed in a specific pathogen-free facility at the Division of Laboratory Animal Medicine at Louisiana State University (Baton Rouge, LA). All animal use procedures were approved by the Institutional Animal Care and Use Committee of Louisiana State University.

Heart and lungs were removed en bloc from 10-d-old WT and Tg⁺ littermates. Lungs were perfused by flushing 5 ml of PBS through the right ventricle of the heart. Perfused lungs were transferred to C-tubes (Miltenyi Biotec, Gladbach, Germany) and digested in DMEM/F12 medium containing collagenase type IV (Worthington Biochemical, Lakewood, NJ) and DNAase (Sigma-Aldrich, St. Louis, MO). Lung tissues were homogenized using a gentleMACS Dissociator (Miltenyi Biotec) followed by incubation on a rocker at 37°C. Cell suspensions were filtered through a 70-μm nylon cell strainer followed by centrifugation at 300 × g for 10 min. Cells were resuspended in 10 ml of DMEM/F12 medium and counted using TC20 automated cell counter (Bio-Rad Laboratories, Hercules, CA) with a gating size of 6–22 nm.

For Th stimulation, ∼2 million cells were incubated in DMEM/F12 medium containing a leukocyte activation mixture (BD Biosciences, San Jose, CA), a protein transport inhibitor containing monensin (BD Biosciences), and 10% FBS at 37°C for 3 h. Cells were collected and permeabilized/fixed using Fixation/Permeabilization buffer and 1× Permeabilization/Wash Buffer (BD Biosciences). Cells were suspended in 200 μl of PBS with 3% FBS.

Unstimulated cells were centrifuged at 300 × g for 5 min, and the cell pellets were resuspended in 2 ml of PBS. Cells were fixed in 4% paraformaldehyde solution (Electron Microscopy Sciences, Hatfield, PA) followed by a wash and final resuspension in 200 μl of PBS with 3% FBS. Single cell suspensions were stained with various fluorochrome-conjugated anti-mouse mAbs for cellular phenotyping as follows: Lin (CD3, B220, CD11b, TER-119, Gr-1, CD11c, NK1.1, FceR1α, CD8a, and CD4), CD45, CD90, IL-33R, CD278, CD3, CD4, CD25, Foxp3, NK1.1, CD127, RORrt, Eomes, IFN-γ, IL-4, and IL-17. Data were acquired using a BD LSRII flow cytometer (BD Biosciences) and analyzed using FlowJo software (v9.2).

Neonates (postnatal day [PND] 6–7) and juveniles (PND 22) were anesthetized, and aseptic bronchoalveolar lavage (BAL) fluid was harvested as described previously (17). Unlavaged left lung lobes were stored in 10% neutral buffered formalin for histopathological analyses. Cell-free BAL fluid (BALF) was stored at −80°C for cytokine analyses. Right lung lobes were snap frozen and stored at −80°C for gene expression analyses.

Total and differential cell counts were measured in BALF, as described previously (8).

Cell-free BALF was assayed for mouse IL-4, IL-5, IL-10, IL-13, keratinocyte chemoattractant (KC; CXCL1), MIP-2 (CXCL2), G-CSF, MCP-1 (CCL2), IP-10 (CXCL10), and TNF-α concentrations using a Luminex xMAP-based assay (MCYTOMAG-70K), as per the manufacturer’s instructions (EMD Millipore, Billerica, MA).

For the determination of bacterial burden in airspaces, aseptically harvested BAL was serially diluted onto Columbia Blood Agar plates (Hardy Diagnostics, Santa Maria, CA) and incubated anaerobically in a candle jar for 48 h at 37°C as previously described (6, 17). CFUs were enumerated, and morphologically distinct colonies were amplified via restreaking onto Columbia Blood Agar plates. Bacterial samples were heat inactivated for 3 min at 100°C, and genomic DNA was extracted as previously described (8).

Bacterial genomic DNA was extracted as previously described (8). Genomic DNA was processed and analyzed by sequencing hypervariable regions of the 16S rRNA gene by using the following primer pairs: 27 forward, 5′-AGAGTTTGATCMTGGCTCAG-3′ and 1492 reverse, 5′-TACGGYTACCTTGTTACGACTT-3′ and 518 forward, 5′-CAGCAGCCGCGGTAATACG-3′ and 800 reverse, 5′-TACCAGGGTATCTAATCC-3′ using LongAMP Taq 2× Master Mix (New England BioLabs, Ipswich, MA). Sequencing was performed on Applied Biosystems 3720XL DNA sequencer (Psomagen, Rockville, MA). Genus or species of bacterial isolates were identified based on sequence homology using the 16S rRNA sequence database at National Center for Biotechnology Information using the Basic Local Alignment Search Tool.

Unlavaged left lung lobes were formalin fixed, paraffin embedded, sectioned, and processed for histopathological analyses. Histological slides stained with H&E were used to assess cellular and morphological alterations. Alcian blue–periodic acid Schiff (AB-PAS)–stained sections were used to determine intracellular as well as extracellular accumulation of mucopolysaccharide substances. Using Nikon camera software, MCM analyses were performed by quantifying the number of mucous cells per millimeter of basement membrane, under 40× objective of ECLIPSE Ci-L microscope with DS-Fi2 camera attachment (Nikon, Melville, NY). To compare other histopathological parameters, a previously described histological semiquantitative grading strategy was employed by a blinded board–certified anatomic pathologist (18).

The degree of mucus obstruction was accessed by quantifying the percentage of obstruction of large airways. Briefly, photomicrographs of a lung section containing airway lumens with maximum obstruction were captured under 20× objective of ECLIPSE Ci-L microscope with DS-Fi2 camera attachment (Nikon). Using ImageJ software, the captured images were processed to determine the percentage of total luminal area obliterated by the AB-PAS–stained mucopolysaccharide content (19).

Formalin-fixed, paraffin-embedded lung sections were used for immunohistochemical localization of MUC5AC and MUC5B. Five-micrometers sections were deparaffinized with CitriSolv (2 × 5 min each) and were rehydrated with graded ethanol (100, 95, 70, and 30% distilled water; 3 min each). Ag retrieval was performed using a citrate buffer–based, heat-induced Ag-retrieval method (heating slides in 10 mM sodium citrate solution [with 0.05% Tween 20; pH 6] at 95–100°C for 30 min, followed by cooling to room temperature). Quenching for endogenous peroxidases was performed with 3% hydrogen peroxide in deionized water for 10 min at room temperature, followed by a PBS wash. After blocking with 3% goat serum for 30 min, sections were incubated for 2 h at room temperature with rabbit polyclonal MUC5B primary Abs (UNC223; University of North Carolina, Chapel Hill, NC), MUC5AC primary Abs (UNC294; University of North Carolina), and isotype control Abs (Rabbit IgG; Jackson ImmunoResearch Laboratories, West Grove, PA). The sections were then rinsed in PBS (2 × 5 min each) and processed using VECTASTAIN Elite ABC HRP Kit (Vector Laboratories, Burlingame, CA), followed by chromogenic substrate conversion to insoluble colored precipitate using ImmPACT NovaRED HRP substrate Kit (Vector Laboratories). After a rinse in tap water, sections were counterstained with Gill’s Hematoxylin I, rinsed in deionized water, dehydrated with graded alcohol solutions, and coverslipped with VectaMount Mounting Media (H-5000; Vector Laboratories). Mounted slides were imaged by light microscopy (Nikon Ci-L microscope).

Differentially stained cytospins were used for macrophage surface area (μm²) determination. On 20× objective of ECLIPSE Ci-L microscope with DS-Fi2 camera attachment (Nikon), the surface area of at least 200 macrophages/mouse was determined using Nikon camera software (Nikon).

To induce aspiration-induced bacterial infection, 10 μl of sterile PBS was administered intranasally (5 μl/nostril) in 2-d-old pups for four consecutive days (PND2-5). In experimental groups, 200 pg of rIL-10 (PeproTech, Rocky Hill, NJ) dissolved in 10 μl of PBS was administered.

BALF cytospins were prepared and fixed with 10% neutral buffered formalin at room temperature for 20 min. Fixed cytospins were washed with PBS and permeabilized with 0.1% Triton X-100. After PBS washes, cytospins were incubated with Odyssey Blocker (LI-COR Biosciences, Lincoln, NE) for 30 min at room temperature, after which the cytospins were incubated (overnight incubation at 4°C) with rabbit polyclonal Abs against inducible NO synthase (iNOS) (1:250, ab15323; Abcam, Cambridge, MA). After PBS washes, cytospins were incubated (1 h at room temperature) with Alexa Fluor 594–labeled goat anti-rabbit (Life Technologies, Carlsbad, CA) secondary Ab. After nuclear staining (NucBlue; Life Technologies), images were acquired on a live cell imager (ZOE Fluorescent Cell Imager; Bio-Rad Laboratories).

Gene expression analyses were performed as previously described (8).

Significant differences among groups were determined using ANOVA followed by Tukey post hoc test for multiple comparisons. Nonparametric data were compared by the Kruskal–Wallis test with Dunn post hoc test for multiple comparisons. Significant differences between the two groups were determined using Student t test assuming unequal variance. A p value of 0.05 was considered statistically significant. All data were expressed as mean ± SEM. GraphPad Prism 8.0 was used to perform statistical analyses (GraphPad Software, La Jolla, CA).

To quantify the relative levels of ILC subtypes (ILC1, ILC2, and ILC3) and their nonamnestic counterpart T helper cell subtypes (Th1, Th2, and Th17), whole lung digests from 10-d-old WT and Scnn1b-Tg⁺ pups were analyzed using flow cytometry. Although nonsignificant, the total counts for ILC2s (CD45⁺Lin⁻CD90⁺IL-33R⁺CD278⁺) trended higher in Tg⁺ pups as compared with WT counterparts (Fig. 1A). Notably, these cells were present in a significantly higher proportion of total harvested cells in Tg⁺ pups as compared with WT counterparts (Fig. 1A). Consistent with ILC2s, the total counts were trending higher, and the proportion of Th2 cells (CD3⁺CD4⁺IL-4⁺) was also higher in Tg⁺ lungs as compared with WT lungs (Fig. 1B). The total counts and proportion of Th1/Th2 hybrid cells were also elevated in Tg⁺ pups as compared with WT counterparts, but the differences were statistically insignificant (Supplemental Fig. 1B).

The total counts (p = 0.07) and proportion of ILC1 (p < 0.05) were lower in the Tg⁺ pups as compared with WT counterparts (Supplemental Fig. 2A). ILC3s were present at comparable proportions in both groups (Supplemental Fig. 2B). Although total counts as well as the proportion of Th1 cells were lower in Tg⁺ pups as compared with WT counterparts, the differences were not statistically significant (Supplemental Fig. 2C). Consistent with ILC3s, their nonamnestic equivalent (i.e., Th17) were also present at comparable proportions in both groups (Supplemental Fig. 2C). These data suggest that the Th2 response-associated key cell type (i.e., ILC2 and Th2) is enriched in the lungs of 10-d-old Tg⁺ pups.

Neonatal pups are prone to the aspiration of oropharyngeal secretions, along with microbes of oropharyngeal origin (6, 17). To determine whether the ablation of innate lymphoid system, adaptive immune system, or both systems affected the burden of aspirated oropharyngeal bacteria in neonates, aseptically harvested BALF from designated genetic strains were analyzed for bacterial burdens (Fig. 2A). Whereas only ∼38% immunocompetent WT (Il2rg^HET/Rag1^HET/WT) neonates (PND 6–7) had minimal levels of bacterial infection (mean CFUs = ∼49), ∼70% of Il2rg^KO/WT (Il2rg^KO/Rag1^HET/WT) neonates exhibited significantly increased bacterial burden (mean CFUs = ∼2.7 × 10³) (Fig. 2A). In contrast, only ∼18% of Rag1^KO/WT (Il2rg^HET/Rag1^KO/WT) neonates had minimal bacterial burden (mean CFUs = ∼4.5) (Fig. 2A). Interestingly, compared with Il2rg^KO/WT pups, the Il2rg^KO/Rag1^KO/WT neonates had remarkably lower (p value = 0.07) bacterial burden (mean CFUs = ∼34; prevalence = ∼25%), suggesting that the ablation of the adaptive immune system compensates for compromised bacterial clearance that was evident in Il2rg^KO/WT neonates (Fig. 2A).

As reported previously (6, 17), as compared with WT littermates, the immunocompetent Tg⁺ neonates had significantly higher spontaneous bacterial infections (CFUs = ∼1.9 × 10³; prevalence = 100%). Next, we determined bacterial counts in the BALF from Il2rg^KO/Tg⁺, Rag1^KO/Tg⁺, and Il2rg^KO/Rag1^KO/Tg⁺ neonates (Fig. 2A). Although not significantly different, all the Il2rg^KO/Tg⁺ neonates had a higher bacterial burden (CFUs = ∼5.6 × 10³; prevalence = ∼100%) as compared with immunocompetent Tg⁺ neonates (Fig. 2A). The Rag1^KO/Tg⁺ neonates had lower bacterial burden (CFUs = ∼1.0 × 10³; prevalence = ∼87%; p value = 0.06) compared with Il2rg^KO/Tg⁺ neonates (Fig. 2A). As compared with Il2rg^KO/Tg⁺ neonates, the ablation of the adaptive immune system in innate lymphoid–deficient Scnn1b-Tg⁺ neonates (Il2rg^KO/Rag1^KO/Tg⁺) resulted in lower prevalence (∼69%) with lower bacterial burden (CFUs = ∼1.6 × 10³) (Fig. 2A).

To determine if the loss of the innate lymphoid immunity results in more persistent infections, BALF was aseptically harvested from juveniles (PND 22), and microaerophilic CFUs were enumerated. The early neonatal infection was mostly resolved in WT juveniles at PND 22, with the exception of one mouse with minimal infection (Fig. 2B). As compared with their higher bacterial burden at PND 6–7, ∼42% of the Il2rg^KO/WT juveniles had bacterial burden at lower level (mean CFUs = ∼8.2). Approximately 44% of Rag1^KO/WT and ∼33% of Il2rg^KO/Rag1^KO/WT juveniles also had bacterial burden of ∼16 and ∼10 CFUs, respectively. The spontaneous bacterial infection that was highly prevalent in immunocompetent Tg⁺ neonates was significantly cleared in ∼57% of immunocompetent Tg⁺ juveniles (CFUs = ∼87) (Fig. 2B). Whereas ∼83% Il2rg^KO/Tg⁺ juveniles still had a significantly higher bacterial burden (CFUs = ∼1.8 × 10³), ∼45% Rag1^KO/Tg⁺ juveniles exhibited a relatively lower bacterial burden (CFUs = ∼258) (Fig. 2B). Further, ∼67% Il2rg^KO/Rag1^KO/Tg⁺ juveniles had also relatively lower bacterial burden (CFUs = ∼588) as compared with Il2rg^KO/Tg⁺ juveniles.

To determine the effect of ablation of the immune systems on the diversity of microaerophilic microflora, bacterial 16S sequencing was performed on aseptically harvested CFUs from different neonate groups. The infected neonates from the eight experimental groups had Pasteurella sp. as a common bacterium (Fig. 2C). Except three infected mice in the immunocompetent WT group, none of the infected neonates in Il2rg^KO/WT, Rag1^KO/WT, and Il2rg^KO/Rag1^KO/WT neonates had Streptococcus sp. infection. However, Streptococcus sp. infection was prevalent in all the four groups on Tg⁺ background. Except for two incidences of bacterial infection due to Achromobacter sp. and Bacillus sp., infected neonates in all the four Tg⁺ groups exhibited a comparable diversity of microflora (Fig. 2C). These data suggest that the individual or combined loss of innate lymphoid and adaptive immune system does not affect the diversity of microaerophilic microflora in Tg⁺ neonates.

IL-10 is a known immunosuppressive anti-inflammatory cytokine that is involved in reduced bacterial clearance through the suppression of antibacterial functions of various cells, including T helper subsets, macrophages, dendritic cells, and neutrophils (20–23). Cells of both immune systems (i.e., innate lymphoid and adaptive immune systems), as well as macrophages and dendritic cells, are known producers of IL-10 (24, 25). To investigate the mechanism that accounts for the better clearance of bacterial infection in Il2rg^KO/Rag1^KO mice, we hypothesized that the reduction in the immunosuppressive effect of IL-10 on proinflammatory macrophage activation contributed to improved bacterial clearance. Accordingly, we assayed IL-10 levels in the BALF of all the experimental groups at PND 6–7 (Fig. 3A). Although nonsignificant, the BALF from Il2rg^KO/WT had higher levels of IL-10 (mean 14.24 ± 7.63 pg/ml) as compared with immunocompetent WT neonates (mean 1.17 ± 0.23 pg/ml) (Fig. 3A). As compared with the Il2rg^KO/WT neonates, the Rag1^KO/WT (mean 1.13 ± 0.36 pg/ml) and Il2rg^KO/Rag1^KO/WT (mean 2.39 ± 1.27 pg/ml) neonates had relatively lower BALF levels of IL-10 (Fig. 3A). A similar trend was seen among the four groups on Tg⁺ background. The immunocompetent Tg⁺ (mean 95.05 ± 61.16 pg/ml) and Il2rg^KO/Tg⁺ (mean 418.9 ± 405.5 pg/ml) neonates had comparable levels of IL-10 that were significantly elevated as compared with immunocompetent WT neonates (Fig. 3A). As compared with the Il2rg^KO/Tg⁺ neonates, the Rag1^KO/Tg⁺ had relatively lower IL-10 levels (mean 14.18 ± 8.23 pg/ml) without statistical significance, and the IL-10 levels were significantly reduced in Il2rg^KO/Rag1^KO/Tg⁺ neonates (mean 2.47 ± 1.34 pg/ml) (Fig. 3A).

To test whether the beneficial effect of innate lymphoid and adaptive immune deficiency on bacterial clearance is caused by reduced IL-10 contents and enhanced phagocytic clearance of bacteria, we introduced bacterial infection in the Il2rg^KO/Rag1^KO/WT neonates through an intranasal administration model, as delineated in Fig. 3B. In this approach, 10 μl sterile PBS or IL-10 (200 pg in 10 μl of PBS) was administered intranasally (5 μl/nostril) from PND 2–5. As compared with PBS-treated pups, the IL-10–treated pups had ∼2-fold higher bacterial counts (Fig. 3C; p = 0.051). The total (Fig. 3D; p < 0.01), neutrophil (Fig. 3E; p < 0.01), and macrophage (Fig. 3F; p = 0.12) counts were also elevated in IL-10–treated versus PBS-treated neonates.

To determine the effect of IL-10 on macrophage activation, macrophage sizes were estimated. Whereas PBS-treated Il2rg^KO/Rag1^KO/WT neonates had macrophages with significantly larger sizes, macrophages from IL-10–treated Il2rg^KO/Rag1^KO/WT neonates were relatively smaller (Fig. 3G). Approximately 43% of BALF macrophages in PBS-treated Il2rg^KO/Rag1^KO/WT neonates were above 400 μm² (Fig. 3H); in contrast, only ∼13% of BALF macrophages in IL-10–treated Il2rg^KO/Rag1^KO/WT neonates were above 400 μm² (Fig. 3I). Further, we assessed the proportion of iNOS (a marker of classically activated bactericidal macrophages)-positive macrophages in PBS-treated or IL-10–treated Il2rg^KO/Rag1^KO/WT neonates. As expected, BALF from IL-10–treated Il2rg^KO/Rag1^KO/WT neonates had a significantly lesser proportion of iNOS-positive cells (Fig. 3J). These data suggest that morphologically (increased size) and functionally (iNOS-positive) activated macrophages are suppressed by IL-10.

We next determined if the ablation of the innate lymphoid system, adaptive immune system, or a combination of both systems affected the immune cell composition, numbers, or type of immune cells in the BALF from immunocompetent WT, Il2rg^KO/WT, Rag1^KO/WT, and Il2rg^KO/Rag1^KO/WT juveniles. In the absence of Scnn1b transgene, there was no significant difference in the total number of cells between the juveniles (Fig. 4A). In addition, the ablation of the innate lymphoid system, adaptive immune system, or a combination of both systems did not result in significant alterations in the total number of macrophages (Fig. 4B), neutrophils (Fig. 4C), eosinophils (Fig. 4D), and lymphocytes (Fig. 4E).

The total numbers as well as types of immune cells were also determined in the BALF of immunocompetent Tg⁺, Il2rg^KO/Tg⁺, Rag1^KO/Tg⁺, and Il2rg^KO/Rag1^KO/Tg⁺ juveniles. Compared with their WT counterparts, except immunocompetent Tg⁺, all the Tg⁺ groups had significantly increased numbers of total cells (Fig. 4A). Among the four Tg⁺ groups, there were no significant differences in the total cell (Fig. 4A), macrophage (Fig. 4B), neutrophil (Fig. 4C), or lymphocyte (Fig. 4E) counts. As compared with the immunocompetent Tg⁺ juveniles, whereas the Rag1^KO/Tg⁺ juveniles had equivalent eosinophil counts, the Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles had significantly reduced eosinophil counts (Fig. 4D).

Despite comparable numbers of macrophages in the four groups on WT as well as Tg⁺ backgrounds, we observed genotype-dependent significant changes in the surface area of macrophages. The proportions of macrophages with surface area >400 μm² were ∼4%, 0%, 10%, and 12% in immunocompetent WT, Il2rg^KO/WT, Rag1^KO/WT, and Il2rg^KO/Rag1^KO/WT juveniles, respectively (Fig. 4F, Supplemental Fig. 3A–D). As previously reported (6), a significant proportion of macrophages were morphologically activated, as indicated by their sizes, in all the four groups with Tg⁺ status. The genotype-dependent trends in morphological macrophage activation observed on WT background were also evident in groups on Tg⁺ background (i.e., ∼41%, ∼13%, ∼65%, and ∼68% macrophages in Tg⁺, Il2rg^KO/Tg⁺, Rag1^KO/Tg⁺, and Il2rg^KO/Rag1^KO/Tg⁺ juveniles, respectively, had surface area >400 μm²) (Fig. 4F, Supplemental Fig. 3E–H). These data suggest greater morphological activation in the absence of the adaptive immune system but also reduced activation in the absence of the innate lymphoid system.

To identify the roles of the innate lymphoid versus adaptive immune system in the establishment of cytokine/chemokine concentrations in the airspaces of WT and Scnn1b-Tg⁺ mice, we assessed the concentrations of designated chemokines and inflammatory mediators in the BALF from immunocompetent, Il2rg^KO, Rag1^KO, and Il2rg^KO/Rag1^KO mice. Neutrophil chemoattractants (i.e., KC [CXCL1], MIP-2 [CXCL2], and G-CSF) were undetectable in the BALF from WT, Il2rg^KO/WT, Rag1^KO/WT, and Il2rg^KO/Rag1^KO/WT juveniles (Fig. 4G–I). As compared with their WT counterparts, BALF levels of three neutrophil chemoattractants were significantly increased in the BALF from immunocompetent Tg⁺ and Rag1^KO/Tg⁺ juveniles (Fig. 4G–I). However, as compared with immunocompetent Tg⁺, the levels of neutrophil chemoattractants were significantly elevated in the BALF from Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles. As compared with Rag1^KO/Tg⁺ juveniles, the Il2rg^KO/Rag1^KO/Tg⁺ juveniles had significantly elevated levels of all the three neutrophil chemoattractants (Fig. 4G–I).

IL-5, a key cytokine involved in the maturation, survival, and recruitment of eosinophils, was undetectable in the BALF from immunocompetent WT, Il2rg^KO/WT, Rag1^KO/WT, and Il2rg^KO/Rag1^KO/WT juveniles (Fig. 4J). As compared with their WT counterparts, BALF levels of IL-5 were elevated in immunocompetent Tg⁺ and Rag1^KO/Tg⁺ juveniles, a trend also shown by eosinophil counts in the BALF of these groups (Fig. 4D). In contrast, the BALF IL-5 levels were nondetectable in Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles (Fig. 4J).

To determine the contributions of innate lymphoid and adaptive immune systems in Th2 inflammation in Tg⁺ mice, we assessed molecular signatures of Th2 inflammation in immunocompetent, Il2rg^KO, Rag1^KO, and Il2rg^KO/Rag1^KO juveniles on WT and Tg⁺ backgrounds. Whereas BALF levels of IL-13, a primary Th2 cytokine, were below detection limits in all the eight experimental groups (data not shown), the expression levels of Il13 mRNA were elevated in the immunocompetent Tg⁺ as well as Rag1^KO/Tg⁺ juveniles (Fig. 5A). The transcript levels for Il13 were suppressed in Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles (Fig. 5A). The BALF levels of IL-4, another primary Th2 cytokine, were significantly elevated in the immunocompetent Tg⁺ and Rag1^KO/Tg⁺ groups (Fig. 5B). The BALF IL-4 levels were below the detection limits in the rest of the six experimental groups. The gene expression levels of Il4 were at the baseline in the Il2rg^KO, Rag1^KO, and Il2rg^KO/Rag1^KO juveniles (Fig. 5C). The immunocompetent Tg⁺ juveniles had significantly elevated levels of Il4, which were significantly reduced in Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles (Fig. 5C).

Next, we assessed gene expression levels for Th2 inflammation-associated signatures (i.e., Slc26a4 [Pendrin], Clca1/3 [Gob5], Retnla [Fizz1], Chi3l4 [Ym2], and Mmp12), which have been previously reported in Scnn1b-Tg⁺ mice (7–9). In the absence of Scnn1b transgene, the gene expression levels for these signatures were at baseline in all the four groups [e.g., Slc26a4 (Fig. 5D), Clca1/3 (Fig. 5E), Retnla (Fig. 5F) Chi3l4 (Fig. 5G), and Mmp12 (Fig. 5H)]. Both immunocompetent Tg⁺ and Rag1^KO/Tg⁺ groups had significantly elevated levels of these gene signatures. In contrast to immunocompetent Tg⁺, the expression levels of all the five genes were significantly reduced in the Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles (Fig. 5C–H).

Proinflammatory mediators (i.e., IP-10 [CXCL10], TNF-α, and MCP-1 [CCL2]) are present at significantly elevated levels in the BALF of immunocompetent Tg⁺ mice (8, 9, 17). Consistent with these reports, the immunocompetent Tg⁺ mice possessed these mediators at a concentration higher than those in immunocompetent, Il2rg^KO, Rag1^KO, and Il2rg^KO/Rag1^KO juveniles on WT background. As compared with immunocompetent Tg⁺ mice, the BALF levels for IP-10 and TNF-α were significantly elevated in Il2rg^KO/Tg⁺ juveniles (Supplemental Fig. 3I–K). The BALF levels for IP-10 and TNF-α were significantly reduced in the Rag1^KO/Tg⁺ juveniles versus Il2rg^KO/Tg⁺ juveniles. As compared with Il2rg^KO/Tg⁺ juveniles, whereas the Il2rg^KO/Rag1^KO/Tg⁺ juveniles had significantly reduced levels of IP-10, the contents for TNF-α (p = 0.12) and MCP-1 (p = 0.19) were reduced but without statistical significance (Supplemental Fig. 3I–K).

To determine the effect of ablation of the two immune systems on the MCM responses, a consistent feature of Scnn1b-Tg⁺ airway pathology (4, 5), we assessed histological signs and molecular signatures of MCM in all the eight experimental groups. As previously reported (18), the immunocompetent WT juveniles had significant linear density (∼33 cells/mm of basement membrane) of AB-PAS⁺–stained cells, consistent with intracellular mucopolysaccharide contents (Fig. 6A, 6B). Whereas the Il2rg^KO/WT (∼3 cells/mm of basement membrane) and Il2rg^KO/Rag1^KO/WT (∼4 cells/mm of basement membrane) juveniles were completely devoid of AB-PAS⁺ cells, the Rag1^KO/WT juveniles had AB-PAS⁺ cells (∼29 cells/mm of basement membrane) similar to immunocompetent WT juveniles (Fig. 6A, 6B).

As reported previously (4, 5), as compared with immunocompetent WT littermates, the immunocompetent Tg⁺ juveniles had significantly greater linear density (∼62 cells/mm of basement membrane) of AB-PAS⁺ cells. As observed on the WT background, although the Rag1^KO/Tg⁺ juveniles had equivalent numbers of AB-PAS⁺ cells (∼50 cells/mm of basement membrane) as immunocompetent Tg⁺ juveniles (Fig. 6A, 6B), the loss of innate lymphoid immunity significantly reduced the linear density of AB-PAS⁺ cells in Il2rg^KO/Tg⁺ (∼6 cells/mm of basement membrane) and Il2rg^KO/Rag1^KO/Tg⁺ (∼10 cells/mm of basement membrane) juveniles (Fig. 6A, 6B).

Transcript signatures associated with the molecular programming of MCM, including Spdef, Agr2, Tff2, and Gata3, were assessed in total RNA harvested from the lungs of all the eight experimental groups (Fig. 6C–F). In the four experimental groups on WT background, only Tff2, Agr2, and Gata3 were upregulated in two groups with a greater linear density of AB-PAS⁺ cells (i.e., immunocompetent WT and Rag1^KO/WT juveniles) (Fig. 6C–F). The immunocompetent Tg⁺ and Rag1^KO/Tg⁺ juveniles had comparable levels of expression for Spdef, Agr2, Tff2, and Gata3. As compared with immunocompetent Tg⁺ and Rag1^KO/Tg⁺ juveniles, the Il2rg^KO/Tg⁺ juveniles showed relatively lower expression levels of Spdef, Agr2, Tff2, and Gata3. These reductions were significant in Il2rg^KO/Rag1^KO/Tg⁺ juveniles (Fig. 6C–F).

To determine the effect of ablation of the two immune systems on the mucus obstruction, a consistent feature of Scnn1b-Tg⁺ airway pathology (4, 5), we examined lung sections from all the eight experimental groups. As expected, four experimental groups on WT background had no signs of mucopolysaccharide contents in the airway lumens (Fig. 7B). In agreement with previous reports (4, 5), the immunocompetent Tg⁺ juveniles had consistent signs of mucus obstruction in the airways (Fig. 7A, 7B). The Rag1^KO/Tg⁺ juveniles also had obstruction comparable to the immunocompetent Tg⁺ juveniles. Both Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles, consistent with their absence of MCM responses, had reduced mucus obstruction as compared with the immunocompetent Tg⁺ and Rag1^KO/Tg⁺ juveniles (Fig. 7A, 7B). These data are inconsistent with our observation that the BALF from Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles had smaller and fewer mucus plugs as compared with immunocompetent Tg⁺ and Rag1^KO/Tg⁺ juveniles.

Immunohistochemical staining and gene expression analyses for two gel-forming mucins, MUB5B and MUC5AC, were performed on lung sections. MUC5B immunostaining intensity was robust in immunocompetent WT and Rag1^KO/WT juveniles. In contrast, the Il2rg^KO/WT and Il2rg^KO/Rag1^KO/WT juveniles had reduced staining intensity (Supplemental Fig. 4A). Both immunocompetent Tg⁺ as well as Rag1^KO/Tg⁺ juveniles had robust MUC5B immunostaining of the airway epithelial cells and luminal contents (Fig. 7A). The intracellular staining for MUC5B was evident in Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles but was of lesser intensity as compared with immunocompetent Tg⁺ as well as Rag1^KO/Tg⁺ juveniles. Luminal contents in all the four Tg⁺ groups were intensely stained for MUC5B (Fig. 7A).

The Muc5b mRNA signatures were 1.23 ± 0.30–, 1.42 ± 0.54–, 2.35 ± 0.47–, and 0.80 ± 0.12–fold higher in immunocompetent WT, Il2rg^KO/WT, Rag1^KO/WT, and Il2rg^KO/Rag1^KO/WT juveniles, respectively. As compared with immunocompetent WT juveniles, the transcript levels for Muc5b were significantly elevated in the immunocompetent Tg⁺ (6.03 ± 1.16) and Rag1^KO/Tg⁺ (5.39 ± 1.19) juveniles (Fig. 7C). In contrast, the fold increase was significantly reduced in Il2rg^KO/Tg⁺ (3.62 ± 0.62) and Il2rg^KO/Rag1^KO/Tg⁺ (3.24 ± 0.56) juveniles (Fig. 7C).

As compared with immunocompetent WT juveniles, whereas MUC5AC immunostaining intensity was robust in Rag1^KO/WT, the Il2rg^KO/WT and Il2rg^KO/Rag1^KO/WT juveniles had reduced staining intensity (Supplemental Fig. 4B). Both immunocompetent Tg⁺ as well as Rag1^KO/Tg⁺ juveniles had robust MUC5AC immunostaining of the airway epithelial cells as well as luminal contents (Fig. 7A). In contrast, the intracellular as well as luminal immunostaining for MUC5AC in Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles was almost absent (Fig. 7A). Expression of Muc5ac gene was comparable between immunocompetent, Il2rg^KO, Rag1^KO, and Il2rg^KO/Rag1^KO juveniles on WT background (Fig 7D). Among the four groups with Tg⁺ background, only Rag1^KO/Tg⁺ juveniles had significantly elevated levels of Muc5ac transcripts (Figs 7D, 8)).

MCM and increased mucus production are Th2-associated responses (26–28), and IL-33 is a critical player in the manifestation of Th2 responses (29, 30). Accordingly, we recently investigated the effect of IL-33 deficiency on MCM and mucus obstruction in Scnn1b-Tg⁺ mouse (9), a well-accepted mouse model that recapitulates multiple features of human mucoinflammatory lung diseases (31). On the background of systemic IL-33 deficiency, the Tg⁺ juveniles were completely devoid of MCM, but the airway mucus obstruction persisted. In contrast to IL-33–sufficient Tg⁺ juveniles, the IL-33–deficient Tg⁺ juveniles had comparable gene expression levels of the two gel-forming mucins (i.e., Muc5b and Muc5ac), and the BALF contents of secreted MUC5AC protein, but not MUC5B, was significantly reduced. Although our findings revealed the role of IL-33 in the MUC5AC contributions to mucus obstruction in Tg⁺ juveniles, the identity and role of cellular players that mediated MUC5B-dominated mucus obstruction remained unclear.

Apart from Th2-dominated pathology, increased airway mucus production is also associated with inflammatory responses that are mediated through non-Th2 cells or their products (10, 32, 33). Therefore, to complement our recent study, where Th2 responses were blunted in IL-33–deficient Tg⁺ mice, we investigated mice in which Th2 as well as non-Th2 cells of the innate lymphoid and adaptive immune systems were modified as compared with their naive states in WT and Tg⁺ mice. This study addresses a critical question: do the two major immune defense systems, innate lymphoid and adaptive, regulate MCM and mucoinflammatory responses of airway epithelial cells to ASL layer dehydration?

Like B and T lymphocytes of the adaptive immune system, the ILC subsets originate from a common lymphoid precursor but lack Ag-specific receptors (34, 35). The transcriptional and functional machinery of ILCs mirrors that of adaptive immune cells (12, 34, 36). Accordingly, ILCs (i.e., ILC1, ILC2, ILC3, and NK cells) represent amnestic equivalents of adaptive immune cells Th1, Th2, Th17, and cytotoxic T cells, respectively (11–14). Because the development of ILC subtypes from a common lymphoid precursor requires signaling through the common γ-chain of the IL-2R (IL2RG), the Il2rg^KO lack ILCs but retain a limited number of T and B lymphocytes (15). In contrast, because of the absence of Rag1, Rag1^KO lack mature B and T lymphocytes of the adaptive immune system (16). The experimental design adopted in this study is unique in that direct comparisons between Il2rg^KO, Rag1^KO, and Il2rg^KO/Rag1^KO have been made in the context of Tg⁺ mucoinflammatory lung disease.

Our previous report on the longitudinal gene expression analyses in Tg⁺ mice at four disease-relevant age points revealed predominantly Th1-associated M1-like macrophage activation at PND3 and Th2-associated M2-like macrophage activation at PND42 (7). At PND10, the macrophage activation-relevant gene signatures reflected mixed activation responses (i.e., markers of both M1-like and M2-like macrophage activation were observed) (7). Therefore, we performed flow cytometry to characterize the immune cells present in the lungs of 10-d-old WT and Tg⁺ neonates. These analyses revealed that in contrast to the WT lungs, the Tg⁺ lungs had 1) increased proportions of ILC2 and Th2 cells (Fig. 1A, 1B), 2) reduced proportions of ILC1 and Th1 cells, 3) increased proportion of Th1/Th2 hybrid cells, and 4) comparable proportions of ILC3 and Th17 cells. These data suggest that ILC2/Th2 predominated airspace pathology in the 10-d-old Tg⁺ pups, which, most likely, progresses toward a more robust ILC2/Th2 pathology later in life as evidenced by robust M2-like macrophage activation in adult Tg⁺ mice (7). This is also supported by upregulated expression of M2-like macrophage activation-relevant gene signatures, including Mmp12, Retnla, and Chi3l4, in immunosufficient Tg⁺ juveniles (Fig. 5).

During early postnatal life, mice are prone to aspiration of oropharyngeal secretions that contain residential microbes, which are generally cleared by effective MCC and immune defense mechanisms (6). This bacterial burden was significantly elevated in 6–7-d-old WT neonates with ILC deficiency, suggesting beneficial roles of various subsets of ILCs in bacterial clearance. Because IL-33 is a critical factor in the development of ILC2s (37) and IL-33–deficient WT mice did not exhibit similar elevations in bacterial burden (9), it is plausible that the absence of non-ILC2s (ILC1, ILC3, or NK cells) results in increased bacterial burden in Il2rg^KO/WT neonates. Thymic stromal lymphopoietin–driven (i.e., IL-33–independent) ILC2 development has been reported (38), which suggests a possibility that these novel ILC2 subsets might contribute to the bacterial clearance in Tg⁺ mice. Therefore, the exact identity of the ILC subsets in bacterial clearance in Tg⁺ mice remains to be tested.

Interestingly, as compared with immunosufficient WT mice, although the loss of innate immune lymphoid system exacerbated the bacterial burden, the concomitant ablation of the adaptive immune system significantly compensated for the exacerbated bacterial burden that was caused by loss of the innate lymphoid system (Fig. 2). These data suggest that the cellular or humoral components of the adaptive immune system have suppressive effects on the innate lymphoid system–mediated bacterial clearance mechanisms. However, the mechanistic interactions between these two arms of the immune system and their influence on the bacterial clearance potential of phagocytes (i.e., macrophages and neutrophils) remain unexplored.

IL-10 is a known immunosuppressive and anti-inflammatory cytokine (23, 39, 40). Among others, the cells of the adaptive immune system, including T helper and T regulatory cells, are the known producers of IL-10 (22, 24, 25). Reduced levels of IL-10 in the BALF from four groups deficient in the adaptive immune system (Fig. 3A) suggest that the cells of the adaptive immunity are the primary source of IL-10 levels in the BALF. Based on the association between the magnitude of CFUs and the levels of IL-10 in BALF, we hypothesized that IL-10 may suppress the functioning of cellular players that are essential for bacterial clearance. Therefore, we used an intranasal aspiration model to introduce residential oropharyngeal microbes in the lungs of Il2rg^KO/Rag1^KO/WT pups that were either treated simultaneously with IL-10 or PBS (vehicle) (Fig. 3B). Interestingly, as compared with PBS-treated Il2rg^KO/Rag1^KO/WT pups, Il2rg^KO/Rag1^KO/WT pups that received IL-10 exhibited greater bacterial burden (Fig. 3C) and reduced macrophage activation (Fig. 3G–J). The critical role played by macrophages in bacterial clearance in WT as well as Tg⁺ neonates has been previously reported (17). Consistent with this report, our current study also suggests that airspace macrophages constitute a major antibacterial defense system in WT and immunodeficient mice.

Although this study suggests an immunosuppressive effect of IL-10 on macrophages as one plausible reason for the greater susceptibility to bacterial infections in Il2rg^KO/WT pups, further investigation of the role of IL-10 in modulating bacterial clearance responses is awaited. Two intriguing questions remain unanswered: 1) does IL-10 affect the composition of bacterial species in Tg⁺ mice, and 2) does the IL-10–mediated immunosuppression of other target cells, including ILCs (21), dendritic cells (41), and neutrophils (20), also contribute to the poor bacterial clearance response in Il2rg^KO/WT pups? Future studies employing approaches such as pharmacological inhibition or genetic ablation of IL-10 in Tg⁺ mice and the use of IL-10 reporter strains might address these critical questions.

The spontaneous bacterial infections are consistently present in the Tg⁺ neonates, but these infections are cleared as the neonates mature (6), a trend that was also observed in this study. Previous reports speculated that the clearance of bacterial infection in adulthood is most likely a consequence of age-associated maturation of the immune system (6, 42). Our data suggest that the presence of an intact innate lymphoid system, but not the adaptive immune system, is crucial for the improved bacterial clearance in the Tg⁺ juveniles. Indeed, as seen in Rag1^KO/WT neonates, the loss of the adaptive immune system in Tg⁺ neonates and juveniles appeared to promote bacterial clearance.

Mucus plugs provide a hypoxic but fertile microenvironment, a suitable niche for the colonization of microaerophilic bacterial species (43, 44). Accordingly, the degree of bacterial burden of microaerophilic species was expected to be lower in mice strain with reduced mucus obstruction (e.g., Il2rg^KO/Rag1^KO/Tg⁺ as compared with Rag1^KO/Tg⁺ juveniles) (Fig. 3B). Contrary to this expectation, both Il2rg^KO/Rag1^KO/Tg⁺ and Rag1^KO/Tg⁺ juveniles had a comparable bacterial burden, indicating that the reduced mucus obstruction was not beneficial to the former group. These findings point toward a possibility that the bacterial burden in the Tg⁺ mice persisted as a result of poor phagocytic capabilities of phagocytes rather than exaggerated mucus obstruction. This possibility was also suggested in a previous report on Tg⁺ mice in which partial depletion of macrophages severely compromised bacterial clearance, leading to lethal pneumonia in a significant proportion of neonates (17). In contrast, the magnitude of bacterial burden was poorly associated with the degree of mucus obstruction. One plausible reason for this may be that whereas mucus plugs provide a suitable niche for bacterial growth, the buildup of MUC5B, an antibacterial defense protein (45) and the primary contributor to mucus plugging in the Tg⁺ mice (46), could also counter bacterial growth within the mucus plugs.

We recently demonstrated that IL-33 is essential for the recruitment of eosinophils into the airspaces of Tg⁺ juveniles (9). IL-33 binds to various cell types, including naive ILCs and other non-ILC cells (47) that bear its receptor (ST2), to mediate eosinophil recruitment in Tg⁺ juveniles. Our current data suggest a dichotomy between the innate lymphoid and the adaptive immune systems in mediating eosinophil recruitment into the Tg⁺ airspaces. Similar to IL-33–deficient Tg⁺ juveniles, the Il2rg^KO/Tg⁺ juveniles lack airspace eosinophils, suggesting that, most likely, the ILCs, in particular ILC2, respond to IL-33 during eosinophilic recruitment to Tg⁺ airspaces. Conversely, Rag1^KO/Tg⁺ juveniles did not exhibit impaired eosinophil recruitment, suggesting that in the absence of adaptive immune cells, ILCs can independently respond to IL-33 during eosinophilic recruitment to Tg⁺ airspaces. Of note, following the trend of eosinophil recruitment, IL-5, a key eosinophil chemoattractant, was absent in Il2rg^KO/Tg⁺ but not in Rag1^KO/Tg⁺ juveniles. In this study, whereas we have demonstrated that adaptive immune cells are dispensable in the recruitment of eosinophils, the identity of indispensable cells other than ILCs, if any, in eosinophilic recruitment remains elusive.

The Tg⁺ mice exhibit elevated gene expression for the markers of Th2 inflammation, including Slc26a4 (Pendrin), Clca1/3 (Gob5), Retnla (Fizz1), and Chi3l4 (YM2) (7–9). The expression levels of each of these markers were comparable between immunocompetent Tg⁺ and Rag1^KO/Tg⁺ juveniles, suggesting that the adaptive immune system does not contribute to the Th2 inflammatory responses in Tg⁺ juveniles. In the absence of ILCs, the expression levels of all the four markers were suppressed to significantly lower levels. Interestingly, although significantly lower than immunocompetent Tg⁺, the Il2rg^KO/Rag1^KO/Tg⁺ juveniles had Slc26a4 and Chi3l4 expression above basal levels, suggesting the involvement of additional mechanisms in their upregulated expression in Tg⁺ lungs.

A significant increase in the proportion of airway mucous cells (i.e., MCM) is observed in the WT juveniles that wanes toward adulthood (18). Although this feature was consistently observed in immunocompetent WT and Rag1^KO/WT juveniles, the Il2rg^KO/WT and Il2rg^KO/Rag1^KO/WT groups were virtually devoid of AB-PAS⁺ airway mucous cells. As compared with WT counterparts, whereas the immunocompetent Tg⁺ and Rag1^KO/Tg⁺ juveniles had a significantly higher proportion of mucous cells, the Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles remained devoid of AB-PAS⁺ mucous cells. Along similar lines, the gene expression levels for MCM-relevant genes (i.e., Agr2, Tff2, and Gata3) trended lower in Il2rg^KO/WT and Il2rg^KO/Rag1^KO/WT juveniles as compared with immunocompetent WT or Rag1^KO/WT juveniles. Further, the Il2rg^KO/Tg⁺ and Il2rg^KO/Rag1^KO/Tg⁺ juveniles had significantly lower levels of Spdef, Agr2, Tff2, and Gata3 transcripts compared with immunocompetent Tg⁺ mice. These data are consistent with a previous report in which the ablation of ILCs resulted in diminished goblet cell hyperplasia in helminth-infected mice (32). These data suggest that the expression level of MCM-relevant genes and the intracellular accumulation of glycosylated mucins in the airway epithelial cells is regulated by ILCs. These outcomes were not mitigated in the absence of adaptive immune cells.

MCM responses are often associated with an increase in mucus obstruction (48–51). However, despite the presence of a significant number of AB-PAS⁺ mucous cells (∼30–33 cells/mm of basement membrane) in immunocompetent WT and Rag1^KO/WT juveniles, no mucus obstruction was evident (Fig. 6A). Interestingly, with just a 2-fold increase in the number of AB-PAS⁺ mucous cells (∼50–60 cells/mm of basement membrane) in immunocompetent Tg⁺ and Rag1^KO/Tg⁺ juveniles, mucus obstruction was clearly evident in the airways (Fig. 6A). In contrast, despite complete loss of AB-PAS⁺ mucous cells in Il2rg^KO/Tg⁺ juveniles, mucus obstruction persisted. Only Il2rg^KO/Rag1^KO/Tg⁺ mice had a significant reduction in mucus obstruction, which was associated with a reduction in Muc5b gene expression and MUC5B-specific staining of epithelial and luminal mucopolysaccharides (Fig. 7). As evident by studies on MUC5AC- and MUC5B-deficient mice crossed with Tg⁺ mice, ablation of MUC5AC was not associated with a reduction in mucus obstruction, whereas MUC5B was associated with ∼50% reduction in mucus obstruction (46). Therefore, the reduction in mucus obstruction observed in the Il2rg^KO/Rag1^KO/Tg⁺ juveniles could reflect in part the reduction in MUC5B expression, rather than MUC5AC expression.

Lung ILC2s are known to produce Th2 cytokines in diverse disease models, including allergic asthma and helminths infections (52–54). ILC2s were demonstrated as the primary source of IL-5 and IL-13 that were able to induce airway inflammation and airway hyperresponsiveness in murine models of asthma (46, 47). ILC2s from human asthmatics have also been identified as the producers of both IL-5 and IL-13 (55, 56). The contribution of ILC2s toward IL-4 production, however, remains controversial, although a recent study reported IL-4 production by IL-25–responsive pulmonary ILC2s (54). Based on our findings that imply a strong association between ILC deficiency and reduced levels of Th2 cytokines (Figs. 4J and 5A–C), it remains unclear whether ILC2s are the direct source of Th2 cytokines (i.e., IL-4, IL-5, and IL-13) in the Tg⁺ mice. This is because the deficiency of Il2rg not only disrupts ILC development (15) but also affects other signaling pathways that require γ-chain (e.g., type I IL-4R α, [heterodimeric type 1 IL4RαR complex that contains common γ-chain]) (57). Therefore, the reduced levels of Th2 cytokines in Il2rg^KO/Tg⁺ juveniles might be an upshot of disrupted signaling pathways prevalent in non-ILCs. Both IL-4 and IL-5 were found to be reduced in our previous study in which IL-33 was depleted in Tg⁺ mice, indicating that IL-33–regulated ILC2s might act as a source of these cytokines in these mice (9). Of note, IL-33 knockout mice are also known to exhibit thymic stromal lymphopoietin–regulated expansion of IL-13–expressing ILC2s (38), but these ILCs, if present, were not responsible for the maintenance of IL-4/IL-5 levels in IL-33–deficient Tg⁺ mice in our study. Despite this evidence, further studies pinpointing the source of Th2 cytokines in Tg⁺ mice are warranted in refined experimental models with a deficiency of individual ILC subsets.

In conclusion, this study revealed interesting contributions of the innate lymphoid and adaptive immune systems to the pathogenesis of Tg⁺ lung disease (Fig. 8). First, the innate lymphoid system is beneficial in the bacterial clearance in WT and Tg⁺ mice. Second, the ablation of the adaptive immune system promotes macrophage activation and likely facilitates bacterial clearance. Third, ILCs, not adaptive immune cells, are essential for the recruitment of eosinophils to the Tg⁺ airspaces. Fourth, the innate lymphoid system, but not the adaptive immune system, mediates Th2 responses in the Tg⁺ airspaces. Fifth, MCM is abolished in the absence of ILCs. Finally, the ablation of ILCs suppresses the expression of gel-forming mucins but not mucus obstruction. Collectively, these findings suggest critical roles of the innate lymphoid system in bacterial clearance, MCM, and mucin production and that these responses are independent of the adaptive immune system.

We thank Dr. Camille Ehre (University of North Carolina at Chapel Hill) for providing MUC5B and MUC5AC Abs. We thank Sherry Ring for histological tissue processing and Thaya Stoufflet for assistance with multiplex cytokine assays.

The authors have no financial conflicts of interest.