Mucus concentration-dependent biophysical abnormalities unify submucosal gland and superficial airway dysfunction in cystic fibrosis

Abstract

Cystic fibrosis (CF) is characterized by abnormal transepithelial ion transport. However, a description of CF lung disease pathophysiology unifying superficial epithelial and submucosal gland (SMG) dysfunctions has remained elusive. We hypothesized that biophysical abnormalities associated with CF mucus hyperconcentration provide a unifying mechanism. Studies of the anion secretion-inhibited pig airway model of CF revealed elevated SMG mucus concentrations, osmotic pressures, and SMG mucus accumulation. Human airway studies revealed hyperconcentrated CF SMG mucus with raised osmotic pressures and cohesive forces predicted to limit SMG mucus secretion/release. Using proline-rich protein 4 (PRR4) as a biomarker of SMG secretion, CF sputum proteomics analyses revealed markedly lower PRR4 levels compared to healthy and bronchiectasis controls, consistent with a failure of CF SMGs to secrete mucus onto airway surfaces. Raised mucus osmotic/cohesive forces, reflecting mucus hyperconcentration, provide a unifying mechanism that describes disease-initiating mucus accumulation on airway surfaces and in SMGs of the CF lung.

Fig. 1.. Studies of anion transport–inhibited porcine airways.

(A) Solid concentration (% solids) of porcine SMG mucus after treatment with Ach without or with DMA, Bum, or the combination of DMA and Bum. n = 6 to 8 per group. Welch’s analysis of variance (ANOVA) test followed by Dunnett’s multiple comparison test. (B) Osmotic pressure (in pascals) of porcine SMG mucus after treatment with Ach without or with DMA, Bum, or DMA/Bum. n = 6 to 8 per group. Welch’s ANOVA test followed by Dunnett’s multiple comparison test. (C) Linear correlation between % solids and osmotic pressure of pig SMG mucus. n = 29. (D) Transmission electron microscopy (TEM) images of porcine SMG ducts after Ach treatment without or with DMA and Bum. (E) Quantitation of SMG (i) mucus retention and (ii) cilial height in ducts for each treatment condition. (i) n = 4 to 10 ducts from three airway specimens for each condition from two pigs. One to eight regions of mucus retention per duct were evaluated. Each score was used for plot and statistical analyses. n = 72 regions from 25 ducts in total. Kruskal-Wallis test followed by Dunn’s multiple comparison test. (ii) n = 4 to 10 ducts from three airway specimens for each condition from two pigs. One to eight regions of cilial height per duct were measured. Mean of cilial height for each duct was used for plot and statistical analyses. Welch’s ANOVA test followed by Dunnett’s multiple comparison test.

Fig. 2.. Studies of non-CF and CF human SMGs.

(A) Solid concentration (% solids) of human SMG mucus from non-CF (n = 35) and CF (n = 11) subjects. Welch’s t test. (B) Osmotic pressure (in pascals) of human SMG mucus from non-CF (n = 11) and CF (n = 8) subjects. Welch’s t test. (C) Linear correlation between % solids and osmotic pressure of human SMG mucus. (D) Cohesive strength (in millinewtons per centimeter) of human SMG mucus from non-CF (n = 8) and CF (n = 5) subjects. Welch’s t test. (E) SMG ducts of healthy control versus CF airways after Ach treatment captured by TEM. Scale bars, 2 μm. (F) Quantitation of human SMG (i) mucus ductal retention and (ii) cilial height in healthy control versus CF groups. (i) n = 37 regions from 15 healthy control ducts and n = 40 regions from 12 CF ducts. One to eight regions of mucus retention per duct were evaluated depending on TEM image availability. Each score was used for plot and statistical analyses. Wilcoxon’s signed-rank test. (ii) n = 15 ducts from three healthy control donors and n = 12 ducts from three CF donors. One to seven regions of cilial height per duct were measured. Mean of cilial height for each duct was used for plot and statistical analyses. Welch’s t test.

Fig. 3.. Characterizations of SMG mucus.

(A) Solubility of HBE cell culture mucus and human and porcine SMG mucus in PBS. (B) Molecular size, i.e., radius of gyration (R_g), of the soluble fraction of SMG mucus and HBE mucus as determined by light scattering. n = 15 for HBE, n = 4 for human and n = 3 pig SMG mucus. Welch’s ANOVA test (P = 0.1003). (C) (i) Immunofluorescent staining of MUC5B and MUC5AC in non-CF SMG mucus strands. Scale bars, 10 μm. (ii) Intensity of MUC5B and MUC5AC signals on the yellow line in the left panel. (D) Glycan abundance of less complex (cores 1 and 3) glycans in HBE versus SMG mucus. Unpaired t test. n = 3 for each. (E) Abundance of glycans with shorter (≤3) monosaccharides in HBE versus SMG mucus. Unpaired t test. n = 3 for each.

Fig. 4.. Characteristics of SMG mucus strands.

(A) Concentration of strands within SMG mucus samples. Values were normalized by sample volume, imaged, and divided by the number of mean healthy strands. (B) Mean strand width of non-CF versus CF SMG mucus. (C) Mean strand length of non-CF versus CF SMG mucus. (D) SEM images of SMG strands from non-CF subjects. Inset was falsely colored according to angular alignment of pixels quantified with OrientationJ. (E) SEM images of CF SMG strands. Inset was falsely colored according to angular alignment of pixels. Samples from n = 9 (non-CF) and n = 4 (CF) donors. One to five strands per donor were measured. Mean of each donor was plotted. (F and G) Alignment of structures within non-CF versus CF SMG mucus. (F) Histograms of the relative alignment angles of SEM images in (D) and (E). Non-CF sample exhibits increased alignment of surface structures, whereas CF strand does not. (G) Quantification of alignment fraction for each clinical sample based on decreased numbers of bins of (D) and (E) to reflect slight deviations in angle from strand directionality. n = 3 each for non-CF and CF. (A) Mann-Whitney test. (B, C, and G) Unpaired t test.

Fig. 5.. PRR4 in airways and clinical samples.

(A and B) RNA-ISH and IHC for PRR4 in (A) healthy control and (B) CF human airways. Scale bars, 500 μm (low power images) and 200 μm (magnified images). (C) Immunofluorescent costaining of PRR4 and MUC5B in human proximal airways. SMG ductal mucus shows copositivity for PRR4 and MUC5B (white arrow). Scale bars, 200 μm. (D) Immunofluorescent costaining of PRR4 and MUC5B in human SMG mucus strands. (E and F) Label-free quantitative mass spectrometry (MS) analysis of (E) PRR4 and (F) human neutrophil elastase (HNE) peptides after tryptic digestion of induced sputum samples from healthy donors (n = 18), CF (n = 20), non-CF bronchiectasis (NCFB) (n = 10), and primary ciliary dyskinesia (PCD) (n = 9). Kruskal-Wallis test followed by Dunn’s multiple comparison test. (G) PRR4 content in bronchoalveolar lavage (BAL)–obtained flakes. (i) Immunofluorescent costaining of WGA and PRR4 for healthy control and CF mucus flakes (scale bars, 100 μm) and (ii) ratio of staining intensity of PRR4 to WGA in mucus flakes. n = 3 for healthy control and n = 6 for CF. Mann-Whitney test. (H) (i) Immunofluorescent staining of PRR4 and MUC5B for SMGs in healthy control versus CF tissues. Scale bars, 100 μm. (ii) Quantification of serous cell (PRR4) versus mucous cell (MUC5B) signals in SMG in healthy control versus CF. n = 9 for healthy control and n = 6 for CF. Welch’s t test.