Modelling the bacterial communities associated with cystic fibrosis lung infections.

In many human diseases that cystic fibrosis (CF) patients suffer from, for example, lung infections, bacteria have been considered to grow as biofilms. The ability of key CF pathogens such as Pseudomonas aeruginosa to resist antibiotic therapies may be due to the poor drug penetration of these biofilms. The overall aim of this study was to develop biofilm models in vitro that resembled the bacterial species composition of CF sputa. Here, this was a step towards a longer term goal of forming multiple bacterial biofilm models in vitro that would serve, in turn, as better assays of antibiotic susceptibilities than conventionally grown cells. Biofilm models were constructed from 31 CF sputum samples, using a modified microtitre plate assay. Three forms of assessment of these biofilms were made, namely, the mass, microscopic analysis and species composition. Species composition in sputa and biofilms, characterised by terminal restriction fragment length polymorphism (T-RFLP) analysis of ribosomal gene polymerase chain reaction (PCR) products amplified from directly extracted nucleic acids, indicated that the bacterial community in sputa was well reproduced in the biofilm models. Typically, fresh sputa contained 4.6 +/- 2.3 bacterial species, with the species number decreasing to 4.0 +/- 1.6 over 5 days-this was not statistically significant (p = 0.29). This study outlines a novel methodology by which to generate and study bacterial biofilms communities. It is also hoped that the versatility of this in vitro approach, combined with its simplicity and high reproducibility, will make it an effective system to study CF sputum biofilm development and, in the longer term, serve as a means of assessing antibiotic susceptibilities.