Staphylococcus epidermidis in orthopedic device infections: the role of bacterial internalization in human osteoblasts and biofilm formation

Abstract

Background: Staphylococcus epidermidis orthopedic device infections are caused by direct inoculation of commensal flora during surgery and remain rare, although S. epidermidis carriage is likely universal. We wondered whether S. epidermidis orthopedic device infection strains might constitute a sub-population of commensal isolates with specific virulence ability. Biofilm formation and invasion of osteoblasts by S. aureus contribute to bone and joint infection recurrence by protecting bacteria from the host-immune system and most antibiotics. We aimed to determine whether S. epidermidis orthopedic device infection isolates could be distinguished from commensal strains by their ability to invade osteoblasts and form biofilms.

Materials and methods: Orthopedic device infection S. epidermidis strains (n = 15) were compared to nasal carriage isolates (n = 22). Osteoblast invasion was evaluated in an ex vivo infection model using MG63 osteoblastic cells co-cultured for 2 hours with bacteria. Adhesion of S. epidermidis to osteoblasts was explored by a flow cytometric approach, and internalized bacteria were quantified by plating cell lysates after selective killing of extra-cellular bacteria with gentamicin. Early and mature biofilm formations were evaluated by a crystal violet microtitration plate assay and the Biofilm Ring Test method.

Results: No difference was observed between commensal and infective strains in their ability to invade osteoblasts (internalization rate 308+/-631 and 347+/-431 CFU/well, respectively). This low internalization rate correlated with a low ability to adhere to osteoblasts. No difference was observed for biofilm formation between the two groups.

Conclusion: Osteoblast invasion and biofilm formation levels failed to distinguish S. epidermidis orthopedic device infection strains from commensal isolates. This study provides the first assessment of the interaction between S. epidermidis strains isolated from orthopedic device infections and osteoblasts, and suggests that bone cell invasion is not a major pathophysiological mechanism in S. epidermidis orthopedic device infections, contrary to what is observed for S. aureus.

Figure 1. Gentamicin protection assay.

MOI: Multiplicity of infection.

Figure 2. Invasion of human osteoblasts by S. epidermidis clinical isolates.

A. Determination of the optimal multiplicity of infection (MOI) for use in osteoblast invasion assays (reference strain NCTC11047). B. Evaluation of the rate at which S. epidermidis strains invade the MG63 osteoblastic cell line for nasal colonization (n = 22) versus orthopedic infection (n = 15) isolates. C. Comparison of the invasion rate in MG63 osteoblastic cells and primary human osteoblasts for S. aureus 8325-4, S. epidermidis NCTC11047, and one randomly chosen S. epidermidis clinical isolate. Data are represented as the means and standard deviations of three replicate cultures from one gentamicin protection assay. NS: Not significant.

Figure 3. Flow cytometric quantification of adhesion to MG63 osteoblasts by S. epidermidis.

Flow cytometry analysis was performed using an FSC/SSC stopping gate with uninfected cells (A1), with a fluorescence intensity marker M set to include <2% of uninfected cells (A2). The results from the original cytogram (1) and the distribution of FL1-positive events (2) of a representative experiment are shown, including the reference strains S. epidermidis NCTC1147 (B) and S. aureus 8325-4 (C).

Figure 4. Quantification of S. epidermidis adhesion to MG63 osteoblasts.

After 2 h of co-culture of MG63 osteoblasts with infective and colonizing S. epidermidis strains (n = 3 each) and S. aureus strains at an MOI of 500∶1 and 100∶1, respectively, adhered bacteria were labeled using the membrane-impermeable fluorochrome BODIPY FL vancomycin. The results are presented as the means and standard deviations of the proportion of cells bearing bacteria (A) and of arbitrary fluorescence units (AFUs) (B).

Figure 5. Evaluation of biofilm-formation ability of S. epidermidis clinical isolates.

A. Kinetics of early biofilm formation was assayed by the Biofilm Ring Test method for the reference strain NCTC11047 and for infective (n = 15) and colonizing (n = 22) S. epidermidis strains. B. Quantification of mature biofilm formation after 24 and 48 h by the crystal violet staining method for the S. epidermidis reference strain NCTC11047 and 37 clinical isolates.

Figure 6. Pulsed-field gel electrophoresis (PFGE) pattern analysis.

The number on horizontal lines indicates the percentage of homology by the Pearson correlation. BJI: Bone and joint infective strains; NC: Nasal carriage stains. One strain remained resistant to SmaI digestion and was assigned a specific non-restricted profile.