Extracellular DNA-dependent biofilm formation by Staphylococcus epidermidis RP62A in response to subminimal inhibitory concentrations of antibiotics

Abstract

We measured the ability of Staphylococcus epidermidis to form biofilms in the presence of subminimal inhibitory concentrations (sub-MICs) of vancomycin, tigecycline, linezolid and novobiocin. Six strains that produce different amounts of biofilm were tested. The three strains that produced the highest amounts of biofilm exhibited steady-state or decreased biofilm formation in the presence of sub-MIC antibiotics, whereas the three strains that produced lower amounts of biofilm exhibited up to 10-fold-increased biofilm formation in the presence of sub-MIC antibiotics. In two of the inducible strains (9142 and 456a), antibiotic-induced biofilm formation was inhibited by dispersin B, an enzyme that degrades poly-N-acetylglucosamine (PNAG) biofilm polysaccharide. In the third inducible strain (RP62A), dispersin B inhibited biofilm formation in response to sub-MIC vancomycin, but not to sub-MIC tigecycline. In contrast, DNase I efficiently inhibited biofilm formation by strain RP62A in response to sub-MIC tigecycline and vancomycin. DNase I had no effect on antibiotic-induced biofilm formation in strains 9142 and 456a. Our findings indicate that antibiotic-induced biofilm formation in S. epidermidis is both strain- and antibiotic-dependent and that S. epidermidis RP62A utilizes an extracellular DNA-dependent mechanism to form biofilms in response to sub-MIC antibiotics.

Fig. 1

Biofilm formation in 96-well microtiter plates by six strains of S. epidermidis in the presence of subminimal inhibitory concentrations of four antibiotics. The names of the strains are indicated along the top, and the antibiotics are indicated on the right. The x-axis of each graph shows the antibiotic concentration (in µg/ml), and the y-axis indicates the relative crystal violet binding compared to the drug-free control (A_{[plus antibiotic]}/A_{[minus antibiotic]}). Values show mean and range for duplicate wells.

Fig. 2

Effects of DspB and rhDNase on biofilm formation by S. epidermidis strains RP62A (left panels), 9142 (middle panels), and 456a (right panels) in the presence of 1/4 MIC tigecycline (top panels) or 3/4 MIC vancomycin (bottom panels). Cultures were grown in the absence of enzymes (filled circles), or in the presence of 10 µg/ml DspB (open squares) or 10 µg/ml rhDNase (gray triangles). The y-axis indicates the relative crystal violet binding compared to the drug-free control, calculated as described in the legend to Fig. 1. Graphs show mean values for duplicate wells. Error bars were omitted for clarity.

Fig. 3

Detachment of antibiotic-induced S. epidermidis RP62A, 9142 and 456a biofilms by DspB and rhDNase. Biofilms were cultured in 1/4 MIC (0.02 µg/ml) tigecycline (Tig), 3/4 MIC (1.7 µg/ml) vancomycin (Van) or no antibiotic as a control (None). Biofilms were then rinsed and treated with 10 µg/ml DspB (filled bars) or 10 µg/ml rhDNase (gray bars). Control biofilms were treated with enzyme buffer alone. Values show mean percent biofilm detachment for duplicate wells and error bars indicate range. Percent biofilm detachment values were calculated the formula 1 − (A_{[plus enzyme]}/A_{[minus enzyme]}) × 100.