Sonic hedgehog signalling as a potential endobronchial biomarker in COPD

Abstract

Background: The hedgehog (HH) pathway has been associated with chronic obstructive pulmonary disease (COPD) in genome-wide association studies and recent studies suggest that HH signalling could be altered in COPD. We therefore used minimally invasive endobronchial procedures to assess activation of the HH pathway including the main transcription factor, Gli2, and the ligand, Sonic HH (Shh).

Methods: Thirty non-COPD patients and 28 COPD patients were included. Bronchial brushings, bronchoalveolar lavage fluid (BALF) and bronchial biopsies were obtained from fiberoptic bronchoscopy. Characterization of cell populations and subcellular localization were evaluated by immunostaining. ELISA and RNAseq analysis were performed to identify Shh proteins in BAL and transcripts on lung tissues from non-COPD and COPD patients with validation in an external and independent cohort.

Results: Compared to non-COPD patients, COPD patients exhibited a larger proportion of basal cells in bronchial brushings (26 ± 11% vs 13 ± 6%; p < 0.0001). Airway basal cells of COPD subjects presented less intense nuclear staining for Gli2 in bronchial brushings and biopsies (p < 0.05). Bronchial BALF from COPD patients contained lower Shh concentrations than non-COPD BALF (12.5 vs 40.9 pg/mL; p = 0.002); SHH transcripts were also reduced in COPD lungs in the validation cohort (p = 0.0001).

Conclusion: This study demonstrates the feasibility of assessing HH pathway activation in respiratory samples collected by bronchoscopy and identifies impaired bronchial epithelial HH signalling in COPD.

Keywords: Airway epithelial cells; Bronchoscopy; Chronic obstructive pulmonary disease; Hedgehog signalling pathway.

Fig. 1

Airway progenitor basal cell population is enriched in COPD. a. Study flow chart. b. Representative micrograph showing a Region of Interest (ROI) containing AEC obtained by bronchial brushing in a non-COPD patient stained for cilia (Arl13b, green); mucins (Muc5ac, red); basal cells (p63, white) and cell nuclei (DAPI, blue). Magnification corresponding to the selected area is shown. C. Dot plot with median showing the percentage of ciliated, goblet and basal cells in both non-COPD (n = 15) (black circle) and COPD patients (n = 15) (red circle). *, p < 0.05 and ***, p < 0.0001; non-COPD vs COPD

Fig. 2

Gli2 expression is decreased in AEC from COPD patients. a. Representative micrograph showing a ROI of a bronchial brushing stained for Gli2 (Gli2, red) and cell nuclei (DAPI, blue) in both non-COPD (upper panel) and COPD patients (lower panel). Magnification corresponding to the selected area is shown. b. Dot plot with median showing the total percentage of Gli2-positive cells in non-COPD (n = 15) and COPD patients (n = 15). *, p < 0.05

Fig. 3

Gli2 expression is decreased in airway progenitor basal cell nuclei from COPD patients. a. Representative micrograph showing a ROI of a bronchial brushing stained for cilia (Acetylated tubulin, green); Gli2 (Gli2, red); basal cells (p63, white) and cell nuclei (DAPI, blue) in both non-COPD (upper panel) and COPD patients (lower panel). Magnification corresponding to the selected area is shown. Insets depict localization of the Gli2 transcription factor. b. Dot plot with median showing the percentage of Gli2-positive basal cell nuclei in non-COPD (n = 15) and COPD patients (n = 15). ***, p < 0.0001. c. Linear regression of the percentages of Gli2-positive basal cell nuclei according to FEV₁ (% predicted) for non-COPD (n = 15) and COPD patients (n = 15). Non-COPD patients are represented by black circles and COPD patients are represented by red circles

Fig. 4

Gli2 transcription factor is decreased in whole bronchial epithelium from COPD patients. a. Representative micrograph showing a ROI of a bronchial biopsy stained for cilia (Acetylated tubulin, green); Gli2 (Gli2, red); basal cells (p63, white) and cell nuclei (DAPI, blue) in both non-COPD (upper panel) and COPD patients (lower panel). Magnification corresponding to the selected area is shown. Insets depict localization of the Gli2 transcription factor. b. Dot plot with median showing the intensity of Gli2 mean grey value (Arbitrary units, AU) in whole bronchial epithelium in non-COPD (n = 12) and COPD patients (n = 19). **, p < 0.001 C. Linear regression of the intensity of Gli2 mean grey value according to FEV₁ (% predicted) in non-COPD (n = 12) and COPD patients (n = 19). Non-COPD patients are represented by black circles and COPD patients are represented by red circles

Fig. 5

Shh activating ligand is deficient in COPD bronchi. a. Representative micrograph showing a ROI of a bronchial brushing stained for basal cells (pancytokeratin, KP, white); Gli2 (green); Ptch1 (red, left panel); Hhip (red, right panel); and cell nuclei (DAPI, blue). Magnification corresponding to the selected area is shown. b. Representative micrograph showing a ROI of a bronchial biopsy stained for cilia (Acetylated tubulin or Arl13b, green); Ptch1 (red, left panel); Hhip (red, right panel); and cell nuclei (DAPI, blue). Magnification corresponding to the selected area is shown. c. Representative micrograph showing a ROI of a bronchial biopsy stained for cilia (Arl13b, green); Shh (red) and cell nuclei (DAPI, blue). Magnification corresponding to the selected area is shown. d. Dot plot with median representing Shh concentrations measured by ELISA in bronchial BALF from non-COPD (n = 15) and COPD patients (n = 15). **, p < 0.001

Fig. 6

Shh transcript levels are decreased in COPD lung tissues. Dot plot with median showing normalized expression of Log2-transformed SHH expression in lung tissue from non-COPD (n = 91) and COPD patients (n = 145). ***, p < 0.0001. Transcript expression microarrays were extracted from the GSE47460 dataset available at (https://www.ncbi.nlm.nih.gov/gds). Non-COPD patients are represented by black circles and COPD patients are represented by red circles