Extracellular polymeric substance (EPS)-degrading enzymes reduce staphylococcal surface attachment and biocide resistance on pig skin in vivo

Abstract

Staphylococcal extracellular polymeric substances (EPS) such as extracellular DNA (eDNA) and poly-N-acetylglucosamine surface polysaccharide (PNAG) mediate numerous virulence traits including host colonization and antimicrobial resistance. Previous studies showed that EPS-degrading enzymes increase staphylococcal biocide susceptibility in vitro and in vivo, and decrease virulence in animal models. In the present study we tested the effect of EPS-degrading enzymes on staphylococcal skin colonization and povidone iodine susceptibility using a novel in vivo pig model that enabled us to colonize and treat 96 isolated areas of skin on a single animal in vivo. To quantitate skin colonization, punch biopsies of colonized areas were homogenized, diluted, and plated on agar for colony forming unit enumeration. Skin was colonized with either Staphylococcus epidermidis or Staphylococcus aureus. Two EPS-degrading enzymes, DNase I and the PNAG-degrading enzyme dispersin B, were employed. Enzymes were tested for their ability to inhibit skin colonization and detach preattached bacteria. The effect of enzymes on the susceptibility of preattached S. aureus to killing by povidone iodine was also measured. We found that dispersin B significantly inhibited skin colonization by S. epidermidis and detached preattached S. epidermidis cells from skin. A cocktail of dispersin B and DNase I detached preattached S. aureus cells from skin and increased their susceptibility to killing by povidone iodine. These findings suggest that staphylococcal EPS components such as eDNA and PNAG contribute to skin colonization and biocide resistance in vivo. EPS-degrading enzymes may be a useful adjunct to conventional skin antisepsis procedures in order to further reduce skin bioburden.

Fig 1. System for measuring attachment of bacteria to porcine skin.

(a) Photograph of 24 skin inoculation sites arrayed on a 90 × 140-mm hydrocolloid dressing. Each inoculation site was isolated by means of a polypropylene cloning cylinder attached to the skin with high vacuum grease. A total of four dressings were placed on each animal, two on each side. (b) Schematic of a single skin inoculation site in cross section.

Fig 2

Attachment of S. epidermidis strain 5 (a) and strain NJ9712 (b) to porcine skin. A volume of 400 μl of bacteria (5–10 × 10⁷ c.f.u. ml^-1) was inoculated onto the skin. After 1 or 2 h, inoculation sites were aspirated, excised with a biopsy punch, rinsed with PBS, homogenized, diluted, and plated for c.f.u. enumeration. Values indicate mean c.f.u. mm^-2 counts for six inoculation sites from a single animal and error bars indicate sd.

Fig 3

Dispersin B (DspB) inhibits attachment of S. epidermidis strains 5 (a) and NJ9712 (b) to porcine skin. Attachment assays were carried out for 1 h as described in Fig 2 except that some inocula were supplemented with 10 μg ml^-1 dispersin B (DspB), 10 μg ml^-1 DNase I (DNase), or 10 μg ml^-1 of both enzymes (Both). Enzyme buffer (Buffer) served as a control. Dots show individual c.f.u. mm^-2 values from six inoculation sites from a single animal for each condition. Horizontal lines indicate means. *, significantly different from buffer control (P < 0.005) as determined by one-way ANOVA with Tukey’s post hoc analysis.

Fig 4. EPS-degrading enzymes detach S. epidermidis from porcine skin.

S. epidermidis strains 5 (a) and NJ9712 (b) were allowed to attach for 1 h. Inoculation sites were then aspirated and treated with 10 μg ml^-1 dispersin B (DspB), 10 μg ml^-1 DNase I (DNase), or 10 μg ml^-1 of both enzymes (Both) for 20 min. Enzyme buffer (Buffer) served as a control. Dots show individual c.f.u. mm^-2 values from three to six inoculation sites from a single animal for each condition. In the experiment shown in panel A, three Buffer samples were lost during processing. Horizontal lines indicate means. *, significantly different from buffer control (P < 0.005) as determined by one-way ANOVA with Tukey’s post hoc analysis.

Fig 5. EPS-degrading enzyme cocktail detaches S. aureus strain MZ100 from porcine skin and sensitizes it to killing by povidone iodine.

(a) Bacteria were allowed to attach to skin for 1 h. Inocula were then aspirated and skin was treated with enzyme cocktail (10 μg ml^-1 dispersin B plus10 μg ml^-1 DNase I) for 20 min. Control skin was treated with enzyme buffer alone. Dots show individual c.f.u. mm^-2 values from six inoculation sites from a single animal for each condition, and horizontal lines indicate means. *, P = 0.002 compared to no enzyme control by two-tailed t-test. (b) S. aureus strain MZ100 was allowed to attach to porcine skin for 1 h. Inoculation sites were then aspirated and treated with enzyme cocktail (or enzyme buffer control) for 10 min, followed by 0.4% povidone iodine (or water control) for 5 min. Dots show individual c.f.u. mm^-2 values from six inoculation sites from a single animal for each condition. *, significantly different from enzyme buffer/no drug control (P < 0.005) as determined by one-way ANOVA with Tukey’s post hoc analysis.