Fig. 5

The taxonomic composition of bacterial communities in murine whole lung tissue is distinct from the background-dominant taxonomic composition of BAL fluid and similar to that of the oral microbiome, a biologically plausible source community. a Whole lung tissue clusters separately from BAL fluid and sampling controls by principal component analysis of Hellinger-transformed 16S rRNA gene sequencing data. Individual data points represent specimens grouped by sample or control type. b Whole lung tissue, but not BAL fluid, clusters near tongue samples by principal component analysis of Hellinger-transformed 16S rRNA gene sequencing data. Individual data points represent specimens grouped by sample type. c Bacterial communities in whole lung tissue are more similar to matched (within-mouse) oral communities than BAL fluid. Similarity of lung bacterial communities, grouped by sampling approach, to matched oral communities was quantified using Bray-Curtis dissimilarity index. Median, IQR, and individual data points representing within-mouse comparisons of oral and lung communities are shown. d Relative abundance of bacterial taxa in whole lung tissue are similar to that of oral bacterial communities. In contrast, the relative abundance of bacterial taxa in BAL fluid is similar to that of negative controls. Bars are ranked by mean abundance in whole lung tissue and represent mean ± SEM percent relative abundance of the top 50 bacterial taxa (OTUs) in whole lung tissue across sample types. Labels denote genus (or most specific taxonomic level if no genus was assigned) and unique identifier for each OTU. Overall significance was determined by (a, b) permutational multivariate ANOVA (p = 0.00009 for both). Pairwise significance was determined by (a, b) two-sample PERMANOVA (a only: pooled sampling controls were compared to each lung sample type), and c two-sample unpaired Mann-Whitney U test. Significance key: ns p > 0.05; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001