The lung tissue microbiome in chronic obstructive pulmonary disease

Abstract

Rationale: Based on surface brushings and bronchoalveolar lavage fluid, Hilty and coworkers demonstrated microbiomes in the human lung characteristic of asthma and chronic obstructive pulmonary disease (COPD), which have now been confirmed by others.

Objectives: To extend these findings to human lung tissue samples.

Methods: DNA from lung tissue samples was obtained from nonsmokers (n = 8); smokers without COPD (n = 8); patients with very severe COPD (Global Initiative for COPD [GOLD] 4) (n = 8); and patients with cystic fibrosis (CF) (n = 8). The latter served as a positive control, with sterile water as a negative control. All bacterial community analyses were based on polymerase chain reaction amplifying 16S rRNA gene fragments. Total bacterial populations were measured by quantitative polymerase chain reaction and bacterial community composition was assessed by terminal restriction fragment length polymorphism analysis and pyrotag sequencing.

Measurement and main results: Total bacterial populations within lung tissue were small (20-1,252 bacterial cells per 1,000 human cells) but greater in all four sample groups versus the negative control group (P < 0.001). Terminal restriction fragment length polymorphism analysis and sequencing distinguished three distinct bacterial community compositions: one common to the nonsmoker and smoker groups, a second to the GOLD 4 group, and the third to the CF-positive control group. Pyrotag sequencing identified greater than 1,400 unique bacterial sequences and showed an increase in the Firmicutes phylum in GOLD 4 patients versus all other groups (P < 0.003) attributable to an increase in the Lactobacillus genus (P < 0.0007).

Conclusions: There is a detectable bacterial community within human lung tissue that changes in patients with very severe COPD.

Figure 1.

Quantitative polymerase chain reaction results of the average number of bacterial cells in 1,000 human cells in the four groups of lung samples (n = 8). *P = 0.0002 versus all other groups. COPD = chronic obstructive pulmonary disease; GOLD = Global Initiative for COPD.

Figure 2.

Nonmetric multidimensional scaling analysis of the terminal restriction fragment length polymorphism profiles that provided adequate number of amplicons. Three distinct groups can be discerned, with the negative control subjects clustering close to the cystic fibrosis sample group. COPD = chronic obstructive pulmonary disease; GOLD = Global Initiative for COPD.

Figure 3.

Simpson inverse index of bacterial diversity in the four groups of lung samples and the negative control subjects where higher values correspond to higher diversity. Histograms represent average ± SEM (n = 8 for each lung sample group, n = 6 for the negative control group). *P < 0.0001 versus the cystic fibrosis and negative control groups. COPD = chronic obstructive pulmonary disease; GOLD = Global Initiative for COPD.

Figure 4.

Comparison among the nonsmoker control, smoker control, COPD (GOLD 4), cystic fibrosis, and negative control groups of the percent abundance of the major identified phyla within a sample. Histogram represents average ± SEM. *P < 0.007 versus all other groups. ^†P < 0.003 versus nonsmoking, smoking, and negative control groups. COPD = chronic obstructive pulmonary disease; GOLD = Global Initiative for COPD.