Biofilm EPS can contain polysaccharides, proteins, nucleic acids (Flemming & Wingender, 2002), but few specific reports exist on the EPS matrix of P. putida biofilms. As a first approach to address the increased biofilm formation by the TOL-carrying strain, we attempted to extract EPS from biofilm-containing stagnant liquid cultures by a standard protocol using a cation-exchange resin (Frølund et al., 1996). However, it was not possible to extract the EPS component that caused the higher viscosity: both extracts had similar low viscosities [KT2440: 1.3 cSt BAY 80-6946 chemical structure vs. KT2440 (TOL) 1.8 cSt], and extractable carbohydrate and nucleic acid contents were similar for both strains (Table 3).

This suggested that a major part of the EPS was not extractable and remained cell associated. Further attempts to increase Protease Inhibitor Library extraction intensities rapidly caused cell lysis, as observed by rRNA release in extracts (results not shown). Consequently, we attempted to analyze EPS components directly in the biofilm-containing liquid cultures. Total carbohydrate contents were about double as high as in the extracts, but no striking differences between the two strains were found. Total DNA contents of the two cultures were, however, clearly different: The TOL-carrying strain contained twice the amount of total DNA, as compared with the respective plasmid-free cultures, as similar cellular biomass contents measured as particulate protein. The next approach

involved direct visualization by microscopy. To check for the differential presence of eDNA in liquid–solid interface biofilms, 7-day-old flow cell biofilms of the TOL-free or TOL-carrying KT2440 were stained with propidium iodide (PI). Diffuse PI fluorescence

was colocalized with the larger (presumably older) microcolonies formed by the TOL-carrying strain, indicating that eDNA was a major Sulfite dehydrogenase constituent (Fig. 1). The plasmid-free strains, again, did not form thick biofilms, and eDNA was not observed. Using similar approaches, eDNA has been observed in various pure-culture biofilms (e.g. Pseudomonas aeruginosa, Bacillus subtilis, Enterococcus faecalis, environmental isolate) (Whitchurch et al., 2002; Bockelmann et al., 2006; Thomas et al., 2009; Vilain et al., 2009), and eDNA is, therefore, increasingly being considered a potential core element of the biofilm matrix. eDNA has similarly been observed in flocs and unsaturated biofilms of environmental Pseudomonas (Palmgren & Nielsen, 1996; Steinberger & Holden, 2005). In contrast to our observations, this eDNA remained easily extractable by chemical or thermal treatment methods isolates (Palmgren & Nielsen, 1996; Steinberger & Holden, 2005). Ultimately, eDNA was successfully extracted from TOL-free and TOL-carrying KT2440 cultures by enzymatic treatment using cellulase and proteinase K followed by centrifugation (Wu & Xi, 2009), without concurrent increase in cell lysis as ascertained using live/dead cell staining.