No abstract

The initial attachment and the accumulation of Staphylococcus epidermidis on polymer surfaces in multilayered cell clusters embedded in amorphous slime, which together lead to the plastic-adherent phenotype detected by the adherence assay used in this study, have been proposed to be major virulence factors of these bacteria. An antigen specific for plastic-adherent S. epidermidis strains was detected by an indirect immunofluorescence test using absorbed antiserum raised against the strongly plastic-adherent S. epidermidis 1457. A coagglutination assay was established, which allowed the quantitation of the antigen in bacterial extracts under different physiologic growth conditions. Expression of the antigen and of plastic adherence depended significantly on the presence of glucose in the growth medium. Parallel to increased plastic adherence, a 32- to 64-fold increase in the amount of the antigen was detected in bacterial extracts of cells grown in tryptone soya broth (TSB) compared with that in extracts of cells grown in TSB lacking glucose. A parallel time-dependent increase of plastic adherence and expression of the antigen was observed after stimulation by glucose of stationary-phase cultures of plastic-adherent S. epidermidis strains grown in TSB lacking glucose. The antigen consisted most probably of polysaccharide, because its immunologic reactivity was completely abolished by periodate oxidation but was resistant to protease digestion. A significant proportion of cells of plastic-adherent as compared with nonadherent S. epidermidis strains grown in TSB were located in large cell clusters exceeding 50 cells, which completely disintegrated after periodate oxidation of the cell preparations. Periodate oxidation of adherent bacterial films in situ led to release of the adherent cells from the plastic surface. These results strongly indicate a functional relation of the antigen to adherence of S. epidermidis to polymer surfaces, most probably by mediating intercellular adhesion of cells leading to accumulation in multilayered cell clusters.

An association between adherent biofilm production on tissue culture plates and expression of a specific polysaccharide intercellular adhesion (PIA), which is functionally involved in cell clustering, was investigated for 179 Staphylococcus epidermidis isolates. Of the S. epidermidis strains, 50.8% were biofilm producers (A570 of > 0.1). There was a significant positive association between biofilm production and PIA expression: 86.8% of biofilm-producing S. epidermidis strains produced PIA as detected with a specific coagglutination assay. In contrast, 88.6% of the biofilm-negative isolates did not express PIA (P < .001). A linear association between the amount of PIA produced as detected by inhibition ELISA and the amount of biofilm produced was established for 49 S. epidermidis strains, representing a continuum from biofilm-negative to strongly biofilm-producing (r = .81, P < .001). Apparently, PIA is important for biofilm accumulation in the majority of clinical S. epidermidis isolates.

Hospital-acquired infections associated with implanted medical devices are most commonly caused by staphylococci. Current methods of species identification are slow, costly, and sometimes unreliable. We evaluated the ability of a Bruker Daltonics Microflex MALDI-TOF/MS in conjunction with MALDI Biotyper software to identify 158 characterized staphylococcal isolates from prosthetic joint infections, including 36 Staphylococcus aureus, 100 Staphylococcus epidermidis, 10 Staphylococcus capitis, 8 Staphylococcus lugdunensis, 2 Staphylococcus warneri, and 2 Staphylococcus haemolyticus isolates using the extraction method recommended by Bruker Daltonics. The suggested species identification by the MALDI Biotyper software was correct for all isolates, indicating reliable differentiation between S. aureus and coagulase-negative staphylococci. Applying the recommended criteria of the MALDI Biotyper software all 158 isolates gave scores ≥2.0, implying secure genus and probable species identification for all isolates. 34/36 S. aureus, 36/100 S. epidermidis, 5/10 S. capitis, 6/8 S. lugdunensis, 2/2 S. haemolyticus, 0/2 S. warneri displayed scores ≥2.3 implying highly probable species identification. For S. epidermidis 25/100 additional isolates had a score close to 2.3. It appears that additional clinically relevant staphylococcal isolates in the data base might aid in identification at scores implying highly probable species identification. The ability of the MALDI Biotyper software to recognize clonally-related strains within a species group (i.e. sub-typing) was investigated, and showed great potential. In conclusion, the MALDI-TOF/MS MALDI Biotyper system provides a promising rapid and reliable method of identifying clinical isolates from prosthetic joint infections to the species level, and has potential for sub-typing.

These data demonstrate for the first time that the biological effects of PIA are mediated through C5a in patients with PJI.