Smoking is associated with shortened airway cilia

Abstract

Background: Whereas cilia damage and reduced cilia beat frequency have been implicated as causative of reduced mucociliary clearance in smokers, theoretically mucociliary clearance could also be affected by cilia length. Based on models of mucociliary clearance predicting that cilia length must exceed the 6-7 microm airway surface fluid depth to generate force in the mucus layer, we hypothesized that cilia height may be decreased in airway epithelium of normal smokers compared to nonsmokers.

Methodology/principal findings: Cilia length in normal nonsmokers and smokers was evaluated in aldehyde-fixed, paraffin-embedded endobronchial biopsies, and air-dried and hydrated samples were brushed from human airway epithelium via fiberoptic bronchoscopy. In 28 endobronchial biopsies, healthy smoker cilia length was reduced by 15% compared to nonsmokers (p<0.05). In 39 air-dried samples of airway epithelial cells, smoker cilia length was reduced by 13% compared to nonsmokers (p<0.0001). Analysis of the length of individual, detached cilia in 27 samples showed that smoker cilia length was reduced by 9% compared to nonsmokers (p<0.05). Finally, in 16 fully hydrated, unfixed samples, smoker cilia length was reduced 7% compared to nonsmokers (p<0.05). Using genome-wide analysis of airway epithelial gene expression we identified 6 cilia-related genes whose expression levels were significantly reduced in healthy smokers compared to healthy nonsmokers.

Conclusions/significance: Models predict that a reduction in cilia length would reduce mucociliary clearance, suggesting that smoking-associated shorter airway epithelial cilia play a significant role in the pathogenesis of smoking-induced lung disease.

Figure 1. Cilia length in normal smokers and nonsmokers measured in fixed, paraffin-embedded endobronchial biopsies.

Endobronchial biopsies were fixed in 4% paraformaldehyde, processed for paraffin embedding, sectioned at a thickness of 5 m, and stained with hematoxylin and eosin. Brightfield images were evaluated for cilia length. (A) Endobronchial biopsy morphology in normal nonsmokers. Bar  = 10 m. (B) Endobronchial biopsy morphology in normal smokers. Bar  = 10 m. (C) Mean cilia length within the study populations. Shown are the mean standard error of cilia lengths for nonsmokers (n = 15) and smokers (n = 13). Between 4 and 13 individual measurements were made per study individual (median  = 7). (D) Distribution of cilia lengths in airway epithelial cells from normal nonsmokers and normal smokers. Distributions were constructed by creating histograms from 0.5 micron bins and then equally weighting each study individual.

Figure 2. Cilia length in normal smokers and nonsmokers measured in air-dried, un-fixed samples.

Suspensions of airway epithelial cells were applied to glass slides using a cytocentrifuge, air-dried, and stained with Diff-Quik. (A) Ciliated airway epithelial cell morphology in normal nonsmokers. Bar  = 10 m. (B) Ciliated airway epithelial cell morphology in normal smokers. Bar  = 10 m. (C) Quantitative assessment of cilia length of normal nonsmokers and normal smokers. Shown are the mean ; standard error of cilia lengths for nonsmokers (n = 18) and smokers (n = 21). Between 19 and 60 individual measurements were made per study individual (median  = 27). (D) Distribution of cilia lengths in airway epithelial cells from normal nonsmokers and normal smokers. Distributions were constructed by creating histograms from 0.5 micron bins and then equally weighting each study individual.

Figure 3. Cilia length in detached cilia from airway epithelial cells of normal smokers and nonsmokers.

Suspensions of airway epithelial cells were applied to glass slides using a cytocentrifuge, air-dried, fixed in 4% paraformaldehyde, and stained with an antibody against ∃4-tubulin. Fields include cilia that were detached from cells (arrows) as well as cilia that remain attached to the apical surface of cells (arrowheads). Only detached cilia that were visible from end-to-end were included in the analysis. (A) Detached cilia in airway epithelial cells from normal nonsmokers. Bar  = 5 m. (B) Detached cilia in airway epithelial cells from normal smokers. Bar  = 5 m. (C) Quantitative assessment of cilia length in detached cilia from normal nonsmokers and normal smokers. Shown are the mean standard error of cilia lengths for nonsmokers (n = 13) and smokers (n = 14). Between 5 and 103 individual measurements were made per study individual (median  = 54). (D) Distribution of cilia lengths in airway epithelial cells from normal nonsmokers and normal smokers. Distributions were constructed by creating histograms from 0.5 micron bins and then equally weighting each study individual.

Figure 4. Cilia length in normal smokers and nonsmokers measured in unfixed, hydrated cells.

Freshly isolated suspensions of airway epithelial cells were added to coverslip chambers and imaged using differential interference contrast microscopy. (A) Ciliated airway epithelial cell morphology in normal nonsmokers. Bar  = 10 m. (B) Ciliated airway epithelial cell morphology in normal smokers. Bar  = 10 m. (C) Mean cilia length within the study populations. Shown are the mean standard error of cilia lengths for nonsmokers (n = 6) and smokers (n = 10). Between 17 and 46 individual measurements were made per study individual (median  = 24.5). (D) Distribution of cilia lengths in airway epithelial cells from normal nonsmokers and normal smokers. Distributions were constructed by creating histograms from 0.5 micron bins and then equally weighting each study individual.

Figure 5. Theoretical effect of smoking on ciliary contribution to mucociliary clearance.

The models of cilia/mucus interaction propose that the tip of the cilia must pass through the periciliary fluid to contact the mucus in order to generate mucus movement , , , . Using the data generated in the study of hydrated, unfixed airway epithelial cells, the graph shows the proportion of cilia that would extend through the airway surface fluid (y-axis, “effective cilia”) at a variety of hypothetical airway surface fluid depths (x-axis).The range of reported normal depths of airway surface fluid is shown in gray. Proportions of effective cilia are shown for nonsmokers (blue) and smokers (red).