Figure 2.

Biofilms and planktonic cells are observed across infection types. (A–C) Representative projections of confocal images of sputum samples of CAP with no detected pathogen (A), COPD with Moraxella sp (B) and CF with Pseudomonas aeruginosa (C). Specimens were stained with Tamra-5 (red) using PNA-FISH probes specific to bacterial 16S rRNA and DAPI m(blue). Scale bar is 10 µm. (D) Median total biomass of bacteria by infection type. Bacterial biomass was calculated on each sample (n=43) by counting the voxels representing bacteria after image analysis pipeline. There was no significant difference in sample biomass between infection types (p<0.05, Kruskal-Wallis test). (E) Comparing median sample intensity across infection type. Bacterial objects on each sample were identified, classified as either planktonic cells (≤5 µm³) or biofilms (>5 µm³) and their per cent contribution to total biomass was calculated. We found CAP samples to have the higher median intensity than CF samples (p<0.05, Kruskal-Wallis test), while COPD and CF samples, typically described chronic infections, have equivalent median intensity. (F) Comparing median sample intensity in biofilm (red) vs planktonic cells (black). We also found the median intensity of voxels in biofilms is higher than in planktonic cells (p<0.0001, Wilcoxon test). CAP, community-acquired pneumonia; CF, cystic fibrosis; COPD, chronic obstructive pulmonary disease; DAPI, 4′,6-diamidino-2-phenylindole; PNA-FISH, peptide nucleic acid fluorescent in situ hybridisation.