Abstract
We read with interest the recent paper by Purevdorj-Gage et al. (2005) on seeding dispersal in Pseudomonas aeruginosa biofilms. The authors compared a mucoid cystic fibrosis (CF) P. aeruginosa strain (strain FRD1) with strain PAO1 and found it to be seeding-dispersal-negative. They concluded that seeding dispersal might not be utilized by mucoid variants of CF strains, but rather be a transmission mechanism utilized by environmental strains of P. aeruginosa. We have been investigating mucoid CF isolates in a flow-through biofilm model (Webb et al., 2003) and have some observations that are relevant to this conclusion. Our studies have shown that CF isolates can exhibit biofilm developmental processes and seeding dispersal similar to strain PAO1 in this experimental model. We found that CF P. aeruginosa isolates (n=6) each exhibited a characteristic pattern of biofilm development and microcolony formation that was reproducible and true to strain in replicated biofilm experiments. Some CF strains exhibited a developmental pattern similar to that reported for strain FRD1, without obvious seeding dispersal within the time-frame of our experiments (7 days). However, other strains did form hollow structures with highly motile cells in the centre and seeding dispersal events, as described for strain PAO1, after 45 days of culture (Fig. 1a, b). Clearly, much still remains to be understood about the mechanisms underlying seeding-dispersal and hollow-colony formation. Previous studies using P. aeruginosa strain PAO1 linked this behaviour with bacteriophage-mediated lysis of a subpopulation of cells inside microcolony structures. We also observed dispersal-associated death in our CF isolates. BacLight LIVE/DEAD staining (Molecular Probes) of CF biofilms after day 6 of culture showed that all six strains tested exhibited regions of cell death within microcolonies. For at least one strain the pattern of cell death was identical to that seen with strain PAO1 (e.g. Fig. 1c). Coincident with this microcolony death, bacteriophage titres reached levels of >107 p.f.u. ml1 in flow-cell effluents. Our evidence to date thus suggests that death-associated dispersal mechanisms, as have been described for strain PAO1 (Webb et al., 2003, 2004), are also central to CF strain transmission.