Development of a Novel Collagen Wound Model To Simulate the Activity and Distribution of Antimicrobials in Soft Tissue during Diabetic Foot Infection

Abstract

Diabetes has major implications for public health, with diabetic foot ulcers (DFUs) being responsible for significant morbidity and mortality. A key factor in the development of nonhealing ulcers is infection, which often leads to the development of biofilm, gangrene, and amputation. A novel approach to treating DFUs is the local release of antibiotics from calcium sulfate beads. We have developed a novel model system to study and compare the release and efficacy of antibiotics released locally, using collagen as a substrate for biofilm growth and incorporating serum to mimic the biochemical complexity of the wound environment. We found that our soft-tissue model supports the growth of a robust Pseudomonas aeruginosa biofilm, and that this was completely eradicated by the introduction of calcium sulfate beads loaded with tobramycin or gentamicin. The model also enabled us to measure the concentration of these antibiotics at different distances from the beads and in simulated wound fluid bathing the collagen matrix. We additionally found that a multidrug-resistant Staphylococcus aureus biofilm, nonsusceptible to antibiotics, nonetheless showed an almost 1-log drop in viable counts when exposed to calcium sulfate beads combined with antibiotics. Together, these data suggest that locally applied antibiotics combined with calcium sulfate provide surprising efficacy in diabetic foot infections and offer an effective alternative approach to infection management. Our study additionally establishes our new system as a biochemically and histologically relevant model that may be used to study the effectiveness of a range of therapies locally or systemically for infected DFUs.

FIG 1

Collagen wound model is designed to represent a grade 1B diabetic foot ulcer. (a) An example of a grade 1B ulcer on the University of Texas scale, which denotes a superficial soft-tissue wound with infection. (b) The model was designed to represent a grade 1B ulcer. The collagen matrix was situated in a tissue culture well insert in a 6-well plate that was bathed in media. A void was created in the model using a mold. (c) Appearance of a model with a 20-mm void after incubation with PAO1 containing 1.25 g loaded calcium sulfate beads. (d) The model utilized for all experiments has a void that is 12 mm in diameter and 10 mm deep. (e) Calcium sulfate beads (unloaded) or combined with antibiotic (loaded) were placed into the void of the model, and the movement of antibiotics from the void to the collagen matrix beneath the void (area 1), the collagen adjacent to the void (area 2), or the edge of the model (area 3) and the medium after 3 or 7 days of incubation was determined. (f) Following incubation of the model with loaded or unloaded calcium sulfate beads, the model was sectioned. Colony counts and concentrations of antibiotics were determined for each section.

FIG 2

Characterization of biofilm formation in the collagen matrix. P. aeruginosa was incubated in the collagen wound model for 24 h, and then the model was sectioned and histologically stained. (a) HVG staining shows the collagen matrix (pink) as well as the presence of a cluster of bacterial DNA (stained black) showing bacterial cells growing in a microcolony. (b) PASH staining shows the presence of polysaccharide extracellular matrix (purple) surrounding bacterial DNA (stained dark blue). (c) DAPI fluorescent stain shows clusters of bacterial DNA in microcolonies within the collagen matrix. P. aeruginosa (d) and S. aureus (e) then were incubated in the collagen matrix for 72 h, and sections were imaged by SEM. Rod-shaped P. aeruginosa (d) and S. aureus (e) cocci can clearly be seen in each model. Polymerized collagen fibrils indicated by red arrows form a mesh that the bacteria grow in and interact with. Evidence of extracellular matrix produced by the bacteria is indicated with black arrows.

FIG 3

Growth inhibition of P. aeruginosa biofilm by tobramycin-loaded calcium sulfate beads. (a) Colony counts from a collagen wound model incubated with P. aeruginosa for 24 h and then calcium sulfate beads for 72 h. Colony counts are plotted for each area of the model; (a) and (b) on the axis label denote sections from each side of the central void. Circles show bacterial growth in the presence of control beads not loaded with antibiotic. Squares show colony counts for models that were incubated with 100% tobramycin-loaded beads; no viable counts were recovered. (b) Log reductions in bacterial cell counts after 72 h of incubation of P. aeruginosa in the model followed by a further 72 h of incubation with calcium sulfate beads. Data show the decrease in viable organisms after exposure to tobramycin relative to numbers of viable organisms on exposure to unloaded control beads. (c) Concentration of tobramycin in area 1 of the model underneath the void, to which beads are added, through to area 3 at the edge of the model. (d) Mass of tobramycin detected in the collagen and medium phase of the model (tobramycin not detected is presumed to be retained in the beads).

FIG 4

Growth inhibition of P. aeruginosa by gentamicin-loaded calcium sulfate beads. (a) Log reductions in bacterial cell counts after 72 h of incubation of P. aeruginosa in the model followed by a further 72 h of incubation with calcium sulfate beads. Data show decreases in viable organisms after exposure to gentamicin relative to numbers of viable organisms on exposure to unloaded control beads. (b) Corresponding concentration of gentamicin in area 1 of the model underneath the void, to which beads are added, through to area 3 at the edge of the model. (c) Mass of gentamicin detected in the collagen and medium phase of the model (gentamicin not detected is presumed retained in the beads).

FIG 5

Growth inhibition of MDRSA after incubation with vancomycin- and gentamicin-loaded calcium sulfate beads for 72 h and 7 days. (a) Log reductions in bacterial cell counts after 72 h of incubation of MDRSA in the model followed by a further 72 h of incubation with calcium sulfate beads loaded with vancomycin and gentamicin. Data show the decrease in viable organisms after exposure to antibiotics relative to numbers of viable organisms on exposure to unloaded control beads. (b) Corresponding concentration of antibiotics in area 1 of the models shown in panel a underneath the void, to which beads are added, through to area 3 at the edge of the model. (c) Mass of antibiotics detected in the collagen and medium phase of the models shown in panel a. (d) Log reductions in bacterial cell counts after 72 h of incubation of MDRSA in the model followed by a further 7 days of incubation with calcium sulfate beads loaded with vancomycin and gentamicin. Counts from models exposed to antibiotics are subtracted from counts for unloaded control beads. (e) Corresponding concentration of antibiotics in area 1 of the models shown in panel d underneath the void, to which beads are added, through to area 3 at the edge of the model. (f) Mass of antibiotics detected in the collagen and medium phase of the models shown in panel d.