Abstract
Although enterococci are normal commensals of the gastrointestinal tract of humans and many mammals, they have emerged as important pathogens causing nosocomial infections, including urinary tract, bloodstream, wound and surgical-site infections, in addition to their long-recognized importance as a cause of infective endocarditis (Murray, 1990). The binding of Enterococcus faecalis to heart valves and to various biomaterials (Joyanes et al., 1999; Toledo-Arana et al., 2001; Mohamed et al., 2004; Seno et al., 2005) and medical devices (Keane et al., 1994; Dautle et al., 2003; Sandoe et al., 2003) is the presumed initiating factor that then allows subsequent formation of a biofilm. Biofilm formation in E. faecalis has been reported to be influenced by various genes such as esp, initially reported by Toledo-Arana et al. (2001). Gelatinase, which strongly influences virulence in models of peritonitis, endocarditis (Singh et al., 1998, 2005), endophthalmitis (Engelbert et al., 2004) and in vitro translocation (Zeng et al., 2005), has been reported to influence biofilm formation (Mohamed et al., 2003, 2004; Hancock & Perego, 2004; Kristich et al., 2004). Other E. faecalis genes such as epa, atn (Mohamed et al., 2004), bop (Hufnagel et al., 2004), salA and salB (Mohamed et al., 2006) have also been shown to influence biofilms. Biofilm production has been shown to be regulated by quorum-sensing systems in several important pathogens, including fsr in E. faecalis, which was shown to have a pronounced effect on biofilms (Hancock & Perego, 2004; Mohamed et al., 2003, 2004; Pillai et al., 2004).
The fsr locus, a homologue of the staphylococcal agr loci, global regulators of virulence and metabolism (Dunman et al., 2001), positively regulates the expression of gelatinase and serine protease in E. faecalis (Qin et al., 2000) and autoregulates expression of the fsrB and fsrC genes (Nakayama et al., 2001; Qin et al., 2001). Most strains of E. faecalis are gelE+, although only about 60 % of them have the fsr locus and are gelatinase producers by standard assay (Roberts et al., 2004). We have shown recently that all three fsr mutants (phenotypically GelE SprE by standard assay; Qin et al., 2000; Singh et al., 2005), as well as gelE mutants (GelE SprE), showed decreased biofilm production (Mohamed et al., 2003, 2004), a finding confirmed by Hancock & Perego (2004); the latter also showed that the introduction of fsrABC and fsrABC/gelE/sprE into strain FA2-2 (GelE, with point mutations in the fsrB and fsrC loci) resulted in gelatinase production and increased biofilm production (Hancock & Perego, 2004). However, our initial results had also suggested that fsr may have a role, in addition to activating gelatinase-mediated biofilm formation, as the fsr mutants formed slightly more biofilm than the gelatinase/serine protease double mutant TX5128 (Mohamed et al., 2004). In order to study the effect of fsr on biofilm formation by E. faecalis, independent of activation of its gelatinase production, we used E. faecalis clinical isolates lacking gelE and fsr as hosts in this study.
The bacterial strains and plasmids used in this study are listed in Table 1. For biofilm experiments, bacteria were first grown overnight in tryptic soy broth plus 0.25 % glucose (TSBG) with or without erythromycin, as appropriate, and then in TSBG for biofilm formation. Hosts lacking fsr and gelE that produced strong, medium and weak biofilms from our culture collections and JH2-2 were used as recipients. pTEX5249 (pAT18 carrying fsrABC) and its appropriate control, pAT18, were introduced by electroporation and plated on ToddHewitt (TH) agar with 25 µg erythromycin ml1. Erythromycin-resistant colonies carrying pTEX5249 were identified by PCR and mini-plasmid preparations. Gelatinase production was tested at 72 h on TH agar plates containing 3 % gelatin (Qin et al., 2000). Biofilm formation and adherence to a polystyrene surface were assessed quantitatively as well as by phase-contrast microscopy, as described previously (Mohamed et al., 2004).