The RetnlaTdT (RELMα^KO) mice were generated by genOway, as described previously (Li et al., 2021). These mice were backcrossed to C57BL/6J to generate KO and WT controls, bred in-house. The Retnla-floxed mice were also generated by genOway in a similar way to the RetnlaTdT mice, and bred in-house to generate the Retnla ^F/F mice. The B6.Cg-Tg(Itgax-Cre)1-1Reiz/J (strain #: 008068; common name: CD11c-Cre) and B6N.129S6(Cg)-Scgb1a1tm1(Cre/ERT)Blh/J (strain #: 016225; common name: Scgb1a1-CreER™ or CC10-Cre) mice were purchased from the Jackson Laboratory. The Cre heterozygous CD11c-Cre and CC10-Cre mice were bred in-house with Retnla ^F/F mice to generate RELMα^ΔCC10, RELMα^ΔCD11c, and corresponding F/F controls. The RELMα^ΔCC10 mice are tamoxifen-inducible and were administered five 100-µL doses of tamoxifen (Sigma) in corn oil at a concentration of 20mg/mL intraperitoneally every other day. Following the fifth dose, the mice were given 7 days to recover prior to helminth infection. The mice were sex and age matched (6–13 weeks) and housed in a specific pathogen-free vivarium with standard 12-hour light cycles. The mice had ad libitum access to standard chow and water. The animals were anesthetized with 2% volume-to-volume (v/v) isoflurane using a vaporizer approved for small animals, with a negative hind pedal reflex indicating sufficient anesthesia. Isoflurane was used for humane euthanasia followed by exsanguination for tissue harvest. The animal experiments were conducted in accordance with National Institutes of Health guidelines, the Animal Welfare Act, and the Public Health Service Policy on Humane Care and Use of Laboratory Animals. The protocols for the use of these animals were approved by and in accordance with the guidelines of the University of California Riverside Institutional Animal Care and Use Committee (IACUC, protocol number A-20210017).

The mice were infected subcutaneously with L3 larvae from N. brasiliensis, amounting to 600 worms in 200 µL of saline, as described previously (Batugedara et al., 2018; Li et al., 2021). The mice were given 30 days to recover from this single infection and were euthanized on day 30 following infection on day 0. For the acellular scaffold studies, mice were euthanized on day 80 post infection. Weight was monitored to prevent egregious weight loss in these animals in the few days after the initial infection.

After euthanasia, blood was collected from the renal arteries; the blood was allowed to coagulate and was then processed for serum by centrifugation. A small incision was made, revealing the trachea, to fit the cannula, a luer-lock shield from an 18-gauge shielded intravenous (IV) catheter (cat # 381447; BD Insyte™ Autoguard™). This was tied to the trachea with a nylon suture and used with a 1-mL syringe to perform the BAL using 2 × 800 µL of cold phosphate-buffered saline (PBS). The supernatant from the first BAL wash was used for protein analyses, whereas the cell pellets were used for flow cytometry. Prior to the excision of the heart and lungs en bloc, vascular immune cells were cleared from the lungs by cardiac perfusion with PBS. A nylon suture was tied around the right main bronchus, and the right lung lobes were removed and processed to create a single-cell suspension for flow cytometry and RNA analysis. The excised right lobes were minced using a scalpel and then incubated in a digestion solution [containing 5% fetal bovine serum (FBS), 1 mg/mL of Roche’s collagenase A, and 0.05 mg/mL of Sigma’s DNase I in Hank’s balanced salt solution (HBSS)] for 1 hour while being shaken at 37.5°C. At 15 minutes prior to the final 1-hour mark, the lungs were homogenized through a 18-gauge needle (10 times) and returned for the final 15 minutes of 37.5°C incubation. The incubated sample was poured over 70-µm cell strainers and pelleted. Red blood cell (RBC) lysis was performed in accordance with the manufacturer’s protocol [BioLegend (10X)]. The cells were resuspended in PBS and cell counts were quantified using an automated cell counter (DeNovix CellDrop™). Up to 1 million cells of each sample were used for RNA extraction and gene expression analysis by NanoString. Where indicated, lung suspensions were stained with αCD11c for sorting of CD11c⁺ cells on the MoFlo Astrios cell sorter (Beckman Coulter) prior to NanoString analysis, as previously described (Batugedara et al., 2018).

The lung lobes were prepared for either formalin-fixed paraffin embedding (FFPE) or cryopreservation. For FFPE, the left lobe was partially inflated with 300 ML of 10% buffered formalin via the tied cannula and placed on an apparatus used to inflate the lungs under a constant pressure. This apparatus uses the pressure of 10% buffered formalin at 22 cm above the lungs for inflation, with the lungs surrounded by 10% buffered formalin, using reagent reservoirs; the lungs were sent to the Histology Core Facility at Sanford Burnham Prebys or the Translational Pathology Core Laboratory at UCLA for paraffin embedding. Hematoxylin and eosin (H&E) staining was also performed using these histology cores on 5-µm FFPE lung sections. For optimal cutting temperature (OCT) embedding, lungs were inflated with two parts OCT compound and one part formalin warmed to 60°C. The lungs were submerged in 30% sucrose formalin overnight, then in 30% sucrose PBS overnight. The lungs were embedded in OCT compound over dry ice and sectioned by using a cryotome at 5 µm. The sections were stored at −20°C until they were used for immunofluorescent staining. After washing slides in PBS to remove the OCT compound, slides were blocked for 1 hour (StartingBlock™ Blocking Buffer, cat. 37542; Thermo Scientific™). The slides were labeled with anti-murine CC10 (1: 200, clone: B-6; Santa Cruz Biotechnology), biotin-conjugated anti-murine RELMα (1: 200, cat. 500-P214; PeproTech®), in blocking buffer overnight at 4°C. The slides were washed three times in PBS with 0.05% Tween-20, then labeled with secondary antibody [goat anti-rabbit IgG fluorescein isothiocyanate (FITC), streptavidin-conjugated Cy5, 1: 300] for 2 hours at room temperature, washed three times, then coverslipped with mounting medium containing 4′,6-diamidino-2-phenylindole (DAPI; VECTASHIELD® antifade mounting medium with DAPI, cat. H-1200-10; Vector Laboratories).

The BAL cells were pelleted by centrifugation at 400 g for 5 minutes, then RBCs were lysed. The cells were resuspended in PBS and cell counts were quantified using an automated cell counter (DeNovix CellDrop™). For flow cytometry, the cells were pelleted in v-shaped-bottom 96-well plates. The panel of BioLegend antibodies used for flow analysis was as follows, used at 1: 400: FITC anti-mouse MERTK (Mer) antibody (clone: 2B10C42; BioLegend); Alexa Fluor® 700 anti-mouse major histocompatibility complex class II (MHCII; I-A/I-E) antibody (clone: M5/114.15.2; BioLegend); allophycocyanin (APC)/cyanine 7 (Cy7) anti-mouse CD11b antibody (clone: M1/70, BioLegend); Brilliant Violet 510™ anti-mouse Ly-6G antibody (clone 1A8; BioLegend); Zombie Aqua™ Fixable Viability Kit, Brilliant Violet 605™ anti-mouse CD11c antibody (clone: N418; BioLegend); Brilliant Violet 650™ anti-mouse F4/80 antibody (clone: BM8; Biolegend); phycoerythrin (PE)-CF594 rat anti-mouse Siglec-F (clone: E50-2440; BD Biosciences); and PE/Cy7 anti-mouse Ly-6C antibody (clone: Hk1.4; BioLegend). The samples were acquired using the NovoCyte Quanteon flow cytometer (Agilent Technologies) and flow data were analyzed using FlowJo™ version 10.

Recombinant murine RELMα protein, rabbit-anti-murine RELMα, and biotinylated-rabbit-anti-murine RELMα were purchased from PeproTech (cat # 450-26, 500-P214, and 500-P214BT, respectively). The plates were coated overnight at 4°C with rabbit anti-murine RELMα at a concentration of 0.5 µg/mL, and the ELISA was completed the following day. A detection antibody was also used at a concentration of 0.5 µg/mL.

The H&E-stained slides were blinded and scored from 1 to 5 on alveolar destruction and vascular inflammation across the lung, which was divided into five equally spaced sections from the apex to the base. Each of the five spaces was assessed for the percentage of alveoli damaged and scored. The scores were reported as an average for each mouse. The scoring parameters were as follows: 0 for no detectable damage, 0.5 for 5% damaged, 1.0 for 10% damaged, 1.5 for 20% damaged, 2.0 for 30% damaged, 2.5 for 40% damaged, 3.0 for 50% damaged, 3.5 for 60% damaged, 4.0 for 70% damaged, 4.5 for 80% damaged, and 5 for ≥90% damaged.

The L3 N. brasiliensis larvae were isolated from fecal plates and treated with antibiotics prior to being placed in a sterile culture [containing 0.012 g of neomycin (cat. N1876; Sigma Aldrich), 1% penicillin–streptomycin (cat. 15140122; Fisher Scientific), and 30 mL of sterile PBS]. The lungs from RELMα KO and WT mice were collected on day 9 post infection with N. brasiliensis. The lungs were then digested, as mentioned above, and single-cell suspensions of lung homogenate were enriched for CD11c⁺ cells for use in the worm coculture experiments. Enrichment was performed by anti-CD11c magnetic bead isolation methods in accordance with the manufacturer’s protocol (cat. 130–125–835; Miltenyi Biotec). Flow cytometry and cell appearance on cytospin analysis were used for confirmation of CD11c+ cell enrichment. Then, 400,000 CD11c-enriched cells and 25 live L3 larvae were added to each well of a 24-well plate. The culture medium [500 µL per well; 10% FBS, 1% penicillin–streptomycin, 1 mM sodium pyruvate, 25 mM HEPES in Dulbecco’s modified Eagle’s medium (DMEM)] was also supplemented with serum collected from memory-infected mice deficient in RELMα. Cocultures were treated with an isotype control (cat. I413; Sigma Aldrich; rat IgG 100 microg/mL), anti-murine CD11a (cat. 101101; BioLegend; 2.5 µg/mL), or anti-murine integrin β7 (cat. 321202; BioLegend; 5 µg/mL) at the time of plating. The images were captured 24 hours after plating and cells bound to individual worms were counted from the images.

CD11c+ myeloid cells were sorted from lung cells at day 7 post infection (WT or RELMα KO), lysed, and directly hybridized to the mouse Myeloid Innate Immunity panel (cat. XT-CSO-MMII2-12; NanoString). The log2-transformed values were obtained after positive and negative normalization, and calculated as the geometric mean of corresponding values. Differential expression analysis was conducted with p-values identifying significantly altered genes at a p-value < 0.05 as indicated. Gene set analysis (GSA) was conducted using NanoString advanced analysis algorithms for the Myeloid Innate Immunity panel, facilitated by the ROSALIND® platform. The GSA scores were determined by averaging significance measures across genes within each pathway. Significance scores quantified overall pathway changes and were derived by amalgamating differential expression t-tests for all genes within a pathway.

For the lung cells, RNA was isolated from single-cell suspensions of lung homogenate 30 days post infection (WT/RELMα^ΔCC10/RELMα KO), in accordance with the manufacturer protocol (cat. Z3101, SV Total RNA Isolation System; Promega). Then, 50 ng of total RNA was hybridized to the mouse immunology panel (cat. XT-CSO-MIM1–12; NanoString). The NanoString gene expression data were processed and analyzed using the NanoTube package in R (The R Foundation for Statistical Computing, Vienna, Austria). The genes were filtered based on adjusted p-values (p _adj) with a threshold of < 0.05, identifying differentially expressed genes between experimental groups. The Qiagen Ingenuity Pathway Analysis (IPA) tool was employed to conduct the gene pathway analysis. Relevant differentially expressed pathways with their represented percentage of genes were identified through IPA analysis.

The mice were euthanized on day 80 post infection and perfused by cardiac perfusion with PBS until they turned white. Acellular scaffolds were prepared using an adapted protocol for the decellularization of human lungs from our previous publication (Nordgren et al., 2018). A cannula was placed in the trachea, as in the BAL washes above, and tied tightly with nylon suture wire. The lungs were gently excised, inflated with 1 mL of 0.1% Triton X in deionized, distilled water (ddH₂O), and the syringe was left attached to maintain the volume in the lungs. The lungs were then placed in a plastic container and covered with ddH₂O for overnight incubation at 4°C. The next day, the lungs were deflated and reinflated with 2% deoxycholic acid (Spectrum) and placed back in the container with fresh ddH₂O for overnight incubation at 4°C. On the final day, the lungs were deflated and rinsed with 2% p/s PBS (inflated deflated). Prior to vibratome sectioning, the lungs were inflated with 2% low-melting-point agarose in PBS and chilled on ice to set the agarose. The lungs were sectioned with the manufacturer’s 300-µm-thick tungsten carbide blades and placed in 100% ethanol for sterilization and cold storage (Precisionary Instruments, Compresstome®; speed setting: 1, oscillation setting: 10). To standardize lung size, sections were identified by cut number and only cuts 6–10 were used in these studies. Hoechst staining of representative scaffolds revealed successful decellularization. Prior to the assay, scaffolds were incubated in 2% penicillin–streptomycin in PBS overnight at 37.5°C to clear the agarose. The following day, the 2% p/s in PBS was replaced for two more 20-minute washes at 37.5°C. The murine lung epithelial cell line MLE 12-CRL2110™ [American Type Culture Collection (ATCC®), passages 6–8] was plated at 200,000 cells per scaffold in 48-well plates to assay for ex vivo lung repair. The cells were plated directly onto the scaffolds in 100 µL of ATCC-recommended HITES medium for overnight incubation at 37.5°C in a humidified cell culture incubator at 5% CO₂. The following day, an additional 200 µL of HITES medium was added to each scaffold. On day 3, the medium was replaced. On day 6, the scaffolds were stained with Hoechst dye (1: 2,500 of 20 mM, cat #) and transferred to a new 24-well plate for imaging with a 4X objective using the Keyence BZ-X810 fluorescence microscope. After imaging, scaffolds were gently transferred to new 24-well opaque plates for analysis of cellular ATP, in accordance with the manufacturer’s protocol (cat # G9681, CellTiter-Glo™ 3D Cell Viability Assay; Promega). This allowed for the quantification of ATP from only those cells associated with the scaffold. The protein content of each scaffold was determined by bicinchoninic acid assay (Pierce™ BCA kit). Cellular ATP was normalized to the respective scaffold protein content to adjust for differences in available surface for cell adherence. A nuclei count analysis was performed using an automated protocol from the University of Chicago (Labno, 2014).

Statistical analyses were performed using GraphPad Software Prism 9 (GraphPad Software Inc., CA, USA) and R software (version 2023). With the exception of the NanoString analyses, which were performed on one experiment only (CD11c⁺ cells), and specimens from pooled experiments (lung cells), the experiments were repeated at least twice with a number (n) of three or more per group. Outliers were removed by ROUT. The group analyses were performed via two-way ANOVA followed by Tukey’s post hoc multiple comparisons analyses and plotted as bar graphs showing the mean with the standard error. In vitro analyses were averaged per group per experiment and analyzed by one-way ANOVA or t-test. These were graphed as box and whisker plots. All graphs were formatted using the “colorblind safe 4” color palette in Prism.

RELMα was expressed by macrophages and lung epithelial cells at day 7 post infection, as shown by immunofluorescent staining in the airway and mesenchyme of lung sections ( Figure 1A ). We validated the effectiveness of our complete, myeloid (CD11c-Cre)-, and epithelial (CC10-Cre)-specific RELMα knockout systems ( Figure 1B ) using a variety of methods (e.g., immunofluorescence, intracellular flow cytometry, and ELISA). Because flow cytometry of epithelial cells was challenging owing to the low viability of these cell subsets, RELMα^ΔCC10 mice were validated by immunofluorescent staining of the airways with fluorescent quantification, revealing a significant reduction in RELMα fluorescent intensity in infected Cre^+/– mice (RELMα^ΔCC10) at day 7 compared with Cre^–/– (RELMα^F/F) mice ( Figure 1C ). RELMα^ΔCD11c mice were validated by intracellular flow cytometry, which demonstrated a reduction in the frequency of RELMα⁺CD11c⁺ cells and in the mean fluorescence intensity (MFI) for RELMα. The CD11c⁺ subset comprised interstitial macrophages and dendritic cells, which are also likely sources of RELMα, in addition to alveolar macrophages, which were higher in frequency than the other subsets (Cook et al., 2012; Svedberg et al., 2019). The cell-specific knockouts allowed us to clarify if epithelial or myeloid cells were responsible for maintaining RELMα levels in the peripheral and lung-specific compartments at day 30 post infection ( Figure 1D ). The RELMα^ΔCC10 mice had reduced levels of RELMα in the bronchoalveolar lavage fluid (BALF), with levels unaffected in serum. Conversely, CD11c⁺ cells contributed to the serum RELMα levels, since RELMα^ΔCD11c mice exhibited a significant reduction in RELMα levels in the serum but not in the BALF. Thus, RELMα concentration at day 30 post infection was maintained by specific cell populations in a compartment-dependent manner, with lung-specific RELMα protein contributed by CC10⁺ lung epithelial cells, and systemic expression in the blood contributed by CD11c⁺ cells.

Lung infection or injury markedly alters macrophage populations, causing the recruitment of monocyte-derived alveolar macrophages (Mo-AMs) from the blood that replenish the tissue-resident AM (TR-AM) population, which are originally derived from the fetal liver and yolk sac (Guilliams et al., 2013; Misharin et al., 2017). Mo-AMs recruited following N. brasiliensis infection were also shown to express RELMα. The long-term effects of N. brasiliensis-induced lung inflammation and AM populations in the RELMα transgenic models were evaluated at day 30 post infection. Owing to the extended time point, intestinal worm counts could not be performed. Instead, infections of these mice were assessed by fecal egg burden ( Figure S1 ). Quantification and flow cytometry of the BAL cells were performed at day 30 post infection in WT mice or mice where RELMα was constitutively deleted or deleted in CC10- or CD11c-expressing cells ( Figure S2 ). AMs in the BALF were evaluated and grouped based on CD11c and Siglec-F co-expression, followed by MHCII expression; TR-AMs are gated as MHCII^lo while Mo-AMs are MHCII^HI, based on previous studies (Li et al., 2022). Interstitial macrophages and dendritic cells are also MHC class II^HI; however, they are not Siglec-F positive nor are they typically present in the BAL. Uniform manifold approximation and projection (UMAP) analysis was performed and revealed infection-induced eosinophils and decreased AM populations ( Figure 2A ). Quantification of AMs and eosinophil frequencies in naive and day 30 post infection samples across all genotypes was performed ( Figure 2B ). Overall, AM frequencies, particularly that of Mo-AMs, were reduced with infection. Across all genotypes, eosinophil frequencies were increased with infection, even at this late time point. When comparing genotype-specific differences, no significant changes were observed with infection, in either constitutive or cell-specific RELMα KO mice. However, under naive conditions, Mo-AM frequency was significantly reduced in constitutive RELMα KO mice. In summary, there was an infection-dependent reduction in macrophage frequency, and an increase in eosinophil frequency in the BAL that persisted at the 30-day time point but was not RELMα dependent.

The lungs were inflated with formalin at a constant pressure to ensure physiological airspaces for the comparison across mice of variable lung volumes prior to paraffin embedding ( Figure 3A ). Histopathological scoring of H&E-stained lungs was performed on blinded slides to assess the role of RELMα expression in alveolar destruction following helminth infection ( Figure 3B ). When compared with naive controls, infection resulted in higher scores for alveolar destruction regardless of RELMα status. Deletion of RELMα from either CC10⁺ or CD11c⁺ cells alone was insufficient to result in statistically significant changes; however, constitutive RELMα KO mice had increased levels of alveolar destruction at day 30 post infection, in line with previous studies conducted at earlier time points (Chen et al., 2016; Batugedara et al., 2018; Sutherland Id et al., 2018), which suggests a tissue-protective role for RELMα during helminth infection that can be derived from either epithelial cells or CD11c⁺ cells.

Our previous investigations demonstrated distinctions in activity and function between lung myeloid cells from WT and those from constitutive RELMα KO mice (Batugedara et al., 2018). Notably, enhanced worm binding was observed by KO cells compared to WT cells. We aimed to extend our study into downstream genes responsible for enhanced worm binding that may also contribute to increased lung injury. CD11c⁺ cells were isolated from day 7 post infection lungs from WT and KO mice, and purity was confirmed by flow cytometry and cytospin analysis ( Figure 4A ). An additional anti-CD11c fluorescence minus one (FMO) control was performed and validated anti-CD11c specificity ( Figure S3 ). The CD11c⁺ cell lysates were hybridized to the Myeloid Innate Immunity NanoString panel. Heatmap analysis of the top differentially expressed genes (DEGs) indicated that RELMα WT CD11c⁺ cells had increased levels of expression of genes associated with wound healing (Fn1, Adam8, Arg1, and complement protein C1q), while RELMα KO CD11c⁺ cells had increased levels of gene expression associated with Th1 cytokine-activated macrophages (Cd80, Il12b, and Stat1) ( Figure 4B ). Consistent with these findings, pathway analysis revealed that the DEGs mapped to functions associated with associated with cell migration, adhesion and ECM remodeling. RELMα KO CD11c⁺ cells exhibited Th1-skewed pathways, with increased significance scores for Th1 activation, toll-like receptor signaling, and T-cell activation ( Table 1 ). This skewing toward inflammatory macrophage activation was at the expense of factors involved in the repair of the microenvironment needed to resolve alveolar damage and destruction. We identified two genes with increased levels of expression in RELMα KO CD11c⁺ cells that are involved in cell migration/adhesion: Itgal and Itgb7 ( Figure 4C ) (Schittenhelm et al., 2017). We investigated these as potential effectors mediating the enhanced activation and binding of RELMα KO CD11c⁺ cells to the N. brasiliensis larvae. Blocking antibodies against Itgal and Itgb7 were used in cocultures of N. brasiliensis larvae with WT and KO CD11c⁺ cells isolated from N. brasiliensis-infected mice. RELMα KO CD11c⁺ cells showed significantly increased binding to worms, which was abrogated when CD11a or Itgb7 was blocked ( Figure 4D ). These data indicate that RELMα KO CD11c⁺ cells gain an enhanced worm binding ability through increased expression of integrin alpha L, Itgal, and beta 7, Itgb7.

To determine gene expression changes in the whole-lung compartment that may be associated with deficient tissue repair, gene expression analysis using the NanoString Mouse Immunology panel was performed on lung cell RNA from infected RELMα KO and RELMα^ΔCC10 mice at day 30 post infection, along with their respective infected WT counterparts (RELMα WT and RELMα^F/F). Principal component analysis (PCA) of RELMα WT versus RELMα KO indicated clustering by genotype ( Figure 5A ). There were 83 DEGs that met the cutoff criterion of p _adj<0.05 and a heatmap of the top 30 DEGs was plotted ( Figure 5B ). RELMα KO lung cells had an overrepresentation of genes involved in Th1-skewed or viral activation pathways such as the NK cell activation ligand Klrc2, Nos2, and the calcium-binding proteins S100a8 and S100a9, which are biomarkers for inflammation. In contrast, WT lung cells expressed tissue repair-associated genes, such as Fn1, and chemokines. Ingenuity pathway analysis (IPA) of the DEGs between WT and RELMα KO lung cells indicated changes in JAK, cytokine, Tnfr2, antigen presentation, iNOS, toll-like receptors, and Th1 signaling pathways ( Figure 5C ; Table 2 ). When comparing gene expression in lung cells infected RELMα^ΔCC10 and RELMα^F/F mice at 30 days post infection, PCA analysis did not reveal distinct clusters, indicating that CC10-specific deletion of RELMα was insufficient to alter overall lung cell gene expression ( Figure 5D ). Indeed, only one gene was significantly downregulated in RELMα^ΔCC10 lung cells, Fkbp5, which is involved in glucocorticoid activity ( Figure 5E ). IPA analysis for global gene set patterns indicated alterations in wound healing, toll-like receptors, pathogen-induced cytokine storms, and iNOS signaling pathways ( Table 3 ). Pathways associated with airway inflammation in asthma were also affected. Many of these pathways were shared in the constitutive KO and WT lung cells, suggesting that the CC10-specific RELMα deletion had similar downstream effects as the constitutive deletion, but that the effects were more subtle. Overall, these transcriptional analyses indicate that constitutive but not CC10-specific RELMα deletion leads to the downregulation of genes associated with tissue remodeling and upregulation of inflammatory and Th1 cytokine activation pathways.

In the weeks following a single infection with N. brasiliensis, there are gross alterations to the lung appearance, with infected lungs marked by enlarged, nodular lobes even after the worms have left the lungs ( Figure 6A ). Stitched images demonstrate the diffuse alveolar destruction of H&E-stained FFPE lung sections. Using 20× images taken of the alveolar spaces, the mean linear intercept was determined by semi-automated quantification methods using an ImageJ plugin devised by Nolan et al (Crowley et al., 2019). An increasing chord length suggests further distances between tissue intersections, demonstrating a loss of alveolar surface area. These methods indicated that infected mice had significantly longer average chord lengths than naive mice, which suggests an emphysematous-like pathology post infection, as has been previously described by others (Marsland et al., 2008; Sutherland Id et al., 2018).

We sought to determine if RELMα deficiency during the infection recovery period affects the extracellular matrix (ECM), either directly through its role in collagen cross-linking or indirectly through RELMα-responsive cells with ECM remodeling roles, which could contribute to the lung repair process via epithelium–ECM interactions (Knipper et al., 2015). The decellularized lungs from RELMα WT and RELMα KO mice, both naive and infected, were inflated with agarose for vibratome sectioning (300-µm-thick sections) for in vitro re-epithelialization ( Figure 6B ). A mouse lung epithelial cell line (MLE-12-CRL2110) was plated onto each scaffold to assess the effectiveness of re-epithelialization over a period of 6 days. Hoechst staining of the scaffolds revealed epithelial integration onto each scaffold type, and alveolar destruction on the infected scaffolds was also apparent ( Figure 6C ). Nuclei counts were performed on 4× images of the scaffolds and only WT-infected mice (as opposed to WT-naive mice) showed significantly reduced average nuclei counts ( Figure 6D ). Cellular ATP, as an indicator of cell viability and metabolic activity, was measured and revealed a significant reduction in ATP from KO-infected scaffolds when compared with both WT and KO-naive scaffolds, suggesting that the epithelial cells on the scaffold from infected KO mice were less viable and/or less metabolically active. To normalize for potential differences in scaffold protein content between the groups, protein was extracted from each scaffold following lysis of cells during the ATP assay, which indicated that there were no significant differences in the protein content of the scaffolds used in these studies ( Figure 6E ). Acellular scaffolds were also assessed for soluble collagen content and no significant differences among the groups were observed ( Figure 6F ). Together, these acellular scaffold assays revealed that the lung architecture suffers from long-lasting changes in response to helminth infection, which negatively impacts subsequent re-epithelialization. The deficiency in RELMα did not significantly alter this deficit.