3, Fig S2). Similar strong effects of DNase treatment on biofilm integrity has been observed for P. aeruginosa, Streptococcus mutans, and Streptococcus intermedius (Whitchurch et al., 2002; Petersen et al., 2005). Hence, eDNA may be responsible for the development or stabilization of the air–liquid interface biofilm formed by KT2440 TOL. Its removal by DNase treatment reduces the cohesiveness of the pellicle and probably results in a higher turnover of the pellicle. eDNA release in biofilms (P. aeruginosa, E. faecalis) is often caused by cell lysis under control of density-dependent TGF-beta inhibitor mechanisms (Allesen-Holm et al., 2006; Qin et al., 2007; Thomas et al., 2009), while in other cases, the mechanisms
of its excretion are not clear (Bockelmann et al., 2006; Vilain et al., 2009). Hence, we examined differential culture viability in the static cultures. Using a live/dead staining procedure and flow cytometric quantification of cells, three core observations were made (Table 2). First, TOL carriage delayed initial increase in culture densities, but final densities of both cultures were similar. Second, the fraction of dead cells increased at the end of incubation, but was not affected by plasmid carriage. Third, cell sizes increased slightly buy Y-27632 with culture age, and this effect was strongest for the TOL-carrying strain (Table 2). Exocellular β-glucosidase activity increased in both cultures with time, and sharply
after 7 days, but with little relation to TOL carriage. Therefore, we could not obtain proof for plasmid-carriage-dependent cell lysis as the reason for increased eDNA concentrations. Similar cell counts and live/dead fractions were observed in static cultures of both strains irrespective of plasmid carriage, and measures of released cellular
enzymatic activity were similar. The stimulatory role of plasmid carriage on biofilm formation was first documented and examined with E. coli K-12. The effect was restricted to derepressed plasmids, and pointed to the need for traA-like gene expression, suggesting a direct involvement of conjugal pili as adhesion factors (Ghigo, 2001; Reisner et al., 2003). Observations with a range of E. coli isolates confirmed that Oxymatrine biofilm stimulation was contingent on active conjugal plasmid transfer (Reisner et al., 2006). Although some direct proof of IncF-mating pili involvement in initial biofilm establishment has been provided (May & Okabe, 2008), the exact mechanisms responsible for plasmid-mediated biofilm enhancement remain unresolved. Yang et al. (2008) have shown that enhanced biofilm formation caused by the presence of R1drd19 in E. coli is contingent on the envelope stress response system, speculating that pili synthesis imposes stress on membranes. The virulence plasmid pO157 enhances biofilm formation in E. coli 0157:H7 due to increased exopolysaccharide production (Lim et al.