Abstract
Staphylococcal biofilm development on implanted biomaterials represents an important virulence determinant in the pathogenesis of device-related infections. The prevailing environmental milieu and growth media supplements such as NaCl, ethanol, glucose and sub-inhibitory concentrations of antibiotics strongly influence biofilm development. In the laboratory, the microtitre plate assay is one of the commonest methods used to measure the biofilm-forming capability of Staphylococcus aureus isolates. This semi-quantitative measurement of biofilm formation involves growing bacterial cultures in individual wells of 96-well polystyrene plates according to the method of Christensen et al. (1985). In our experiments bacteria were grown at 37 °C for 24 h before being vigorously washed three times with distilled H2O and dried for 1 h at 56 °C as recommended by Gelosia et al. (2001) prior to staining with a 0.4 % crystal violet solution. The absorbance of adhered, stained cells was measured at 492 nm using a Multiskan plate reader (Flow Laboratories).
We have previously observed that the restriction-deficient laboratory strain RN4220, normally used as an intermediate cloning host, is capable of variable but significant levels of biofilm formation and that NaCl can further induce biofilm development in this strain. RN4220, which is a derivative of NCTC 8325-4 is known to harbour a small deletion in the rsbU gene and is therefore deficient in the stress responsive sigma factor, σB. In addition, these observations appear to contradict previous results, which demonstrated that in the absence of σB, biofilm formation could not be induced in this strain (Rachid et al., 2000). Significantly, Valle et al. (2003) reported that isolates of 8325-4 from different laboratories may have different biofilm forming capacities. To investigate this possibility further and to assess the σB status of our RN4220 strain we used PCR to verify the presence of the 11-bp deletion in the RN4220 rsbU gene as described previously by Kullik et al. (1998) (data not shown). In addition we used RT-PCR to demonstrate that transcription of the σB-dependent genes asp23 (Gertz et al., 1999; Giachino et al., 2001) and csb9 (Gertz et al., 2000) were severely repressed in RN4220 and its parent strain 8325-4 compared to the rsbU-repaired derivative of 8325-4, SH1000 (Horsburgh et al., 2002) (data not shown). Therefore, given that our strain of RN4220 appeared to have the correct genotype we decided to examine whether differences in the culture media used or chemical properties of the polystyrene 96-well plates could explain the difference between our biofilm assay data and that of Rachid et al. (2000). The experiments of Rachid et al. (2000) were performed on cells grown in TSB medium, whereas in this study cells were grown in BHI broth. We therefore repeated our analysis of RN4220 using TSB-grown cells. This revealed that, consistent with the findings of Rachid et al. (2000), the biofilm-forming capacity of RN4220 was diminished when grown in TSB (Fig. 1), indicating that the choice of culture media strongly influences biofilm development in S. aureus. In addition, differences between the Oxoid TSB used in this study and the Difco TSB used by Rachid et al. (2000) may account for the weak but significant biofilm formed by RN4220 in the Oxoid TSB (Fig. 1). However, we still observed a strong induction of biofilm-forming capacity when RN4220 was grown in TSB supplemented with 3 % NaCl (Fig. 1).