Lack of airway submucosal glands impairs respiratory host defenses

Abstract

Submucosal glands (SMGs) are a prominent structure that lines human cartilaginous airways. Although it has been assumed that SMGs contribute to respiratory defense, that hypothesis has gone without a direct test. Therefore, we studied pigs, which have lungs like humans, and disrupted the gene for ectodysplasin (EDA-KO), which initiates SMG development. EDA-KO pigs lacked SMGs throughout the airways. Their airway surface liquid had a reduced ability to kill bacteria, consistent with SMG production of antimicrobials. In wild-type pigs, SMGs secrete mucus that emerges onto the airway surface as strands. Lack of SMGs and mucus strands disrupted mucociliary transport in EDA-KO pigs. Consequently, EDA-KO pigs failed to eradicate a bacterial challenge in lung regions normally populated by SMGs. These in vivo and ex vivo results indicate that SMGs are required for normal antimicrobial activity and mucociliary transport, two key host defenses that protect the lung.

Keywords: host defense; lung; medicine; sus scrofa.

Figure 1.. CRISPR/Cas9 editing produced pigs with a disrupted EDA gene (EDA-KO).

(A) Diagram of interaction of EDA, EDAR, and EDARADD at the epithelial placode. (B) Schematic of EDA protein domains, including the transmembrane domain (TM), furin cleavage site, collagen domain, and TNF domain. Exon 4 of the EDA gene encodes the collagen domain. Relative positions of guide RNAs are shown. (C) PCR fragments from six edited pigs from one litter. Lane two shows position of predicted wild-type product. Pigs 1–5 were male; pig six was female.

Figure 1—figure supplement 1.. Nucleotide sequence of exon 4 of EDA gene and sequences of edited pigs.

Top shows wild-type DNA sequence with predicted amino acid sequence directly above. The glycines in the GXY motif are in red. Below that are DNA sequences of the pigs shown in Figure 1C. Deletions are shown as dashes. Pigs 1–5 were male, and pig 6 was female. Pigs 4 and 6 had two alleles.

Figure 2.. EDA-KO piglets had an appearance consistent with hypohidrotic ectodermal dysplasia.

(A) Physical appearance of wild-type (WT) and EDA-KO piglets. (B) Image of skin on the back of an EDA-KO pig showing alternating bands of hair and bare skin (dotted lines). (C) Image of skin from the top of head of EDA-KO showing bald spot. (D) Eyelids in EDA-KO pig lacked hair, although eyelashes were intact. Eyelids of a wild-type pig are shown for comparison.

Figure 3.. EDA-KO pigs lacked SMGs.

(A) Sections of wild-type and EDA-KO conducting airways, HE (top) and dPAS (middle and bottom) stains. Arrows point to SMGs. Scale bar = 125 mm (top and bottom panels) and 625 mm (middle panels). (B) Data are percentage of airway circumference containing SMG in indicated parts of the airway. Each data point is from a different pig. Trachea: n = 6 wild-type and 5 EDA-KO, *p=0.004. Secondary bronchi: n = 3 wild-type and 5 EDA-KO, p=0.018. Segmental bronchi: n = 3 wild-type and 5 EDA-KO, p=0.018. Bronchioles: n = 6 wild-type and 5 EDA-KO, p=1.00. Statistical analysis was by Mann-Whitney test.

Figure 3—figure supplement 1.. Histopathology of nasal mucosa, nasal planum, and skin of wild-type and EDA-KO piglets.

(A) Wild-type nasal mucosa had glands (arrows); these were absent in EDA-KO pigs. (B) Wild-type nasal planum had eccrine glands (arrows);these were absent in EDA-KO pigs. (C) EDA-KO pigs had sections of haired skin that had hair follicles (left, arrows); hairless regions of skin lacked hair follicles. HE stains. Scale bar = 60 µm (A), 150 µm (B), and 80 µm (C).

Figure 4.. MUC5AC, MUC5B, and β-tubulin IV immunolocalization is similar in airway surface epithelia of wild-type and EDA-KO piglets.

Images are confocal immunofluorescence of wild-type and EDA-KO trachea. In all panels, actin is labeled with phalloidin in grey and nuclei are labeled with DAPI in blue. Scale bar = 20 µm. (A) Images show goblet cells expressing MUC5AC (green) and MUC5B (red). A portion of a SMG is indicated; SMGs were detected in wild-type only. (B) Images indicate ciliated cells (β-tubulin IV, green) and MUC5B (red).

Figure 5.. ASL of EDA-KO pigs has impaired killing of S.aureus in vivo.

(A) Schematic showing S. aureus attached to gold grids by biotin-streptavidin linkages. The grids were placed on ASL of trachea for 1 min in vivo. Then the bacteria were counted and the percentage that were dead was determined. (B) Example of live (green)/dead (red) staining of bacteria after the grid was removed from the airway. (C) Percentage of dead bacteria. *p<0.0012 by Mann-Whitney test.

Figure 6.. Loss of SMGs eliminates mucus strands and impairs MCT ex vivo.

(A) Data are confocal images of tracheal surface of wild-type and EDA-KO pigs. Mucus was labeled with fluorescent nanospheres (red). Left and middle panel show movement of mucus strand with time. Strands were not observed in EDA-KO airway in right panel. The white spots in the middle of the field are reflected light. (B) Number of mucus strands crossing the microscopy field in 15 min. N = 8 wild-type and 5 EDA-KO pigs. *p=0.0016. (C) Ciliary beat frequency on trachea under methacholine stimulation. N = 8 wild-type and 5 EDA-KO pigs. p=0.4351. (D) Metallic spheres were dropped onto the airway surface and movement was tracked with time. Mucus was labeled with fluorescent nanospheres (red). Images from wild-type airway show a mucus strand attached to a sphere and pulling it across the field. Images from EDA-KO airway show mucus attached to sphere that was rolling in place. (E) Fraction of metallic spheres that moved at least 1 mm during a 15-min observation period. N = 7 wild-type and 4 EDA-KO pigs. * indicates p=0.0061. For panels B, C, and E, each dot represents a different pig. Statistical significance was evaluated with a Mann-Whitney test.

Figure 7.. EDA-KO pigs have impaired MCT in vivo.

MCT was assessed by insufflating tantalum microdisks in sedated, spontaneously breathing pigs followed by acquisition of high-resolution CT scans every 9 s for 6.3 min (total 44 scans). Positions of individual microdisks were tracked. Pigs were studied under basal conditions and after stimulating submucosal gland secretion with intravenous methacholine. N = 6 wild-type and 3 EDA-KO pigs. Statistical significance between data from wild-type and EDA-KO was evaluated with a Mann-Whitney test. (A) Examples of individual microdisks (different colors) tracked in wild-type (left) and EDA-KO (right) pigs. Position of carina is indicated. (B) Percentage of time microdisks were moving. * indicates p=0.0238 under basal conditions and p=0.0238 under methacholine-stimulated conditions. (C) Mean speed (mm/min) of microdisks that moved more than 10 mm. One EDA-KO pig had no microdisks moving >10 mm under basal conditions; therefore, only two data points and the range are shown in that case. * indicates p=0.8571 under basal conditions and p=0.1667 under methacholine-stimulated conditions. (D) Percentage of microdisks that reached the larynx during the study. p=0.1667 under basal conditions. * indicates p=0.0238 under methacholine-stimulated conditions.

Figure 8.. EDA-KO pigs have decreased eradication of Staphylococcus aureus from the lung.

S. aureus were aerosolized into the airways and 4 hr later samples were obtained by tracheal washes, bronchoalveolar lavage (BAL), and lung homogenates. Data are the log colony-forming units recovered. N = 5 wild-type and 8 EDA-KO pigs. * indicates p=0.0016 for trachea washes, * indicates p=0.0186 for BAL, and p=0.3908 for distal lung homogenates. Statistical analysis was with a Mann-Whitney test.