Gel-Entrapped Staphylococcus aureus Bacteria as Models of Biofilm Infection Exhibit Growth in Dense Aggregates, Oxygen Limitation, Antibiotic Tolerance, and Heterogeneous Gene Expression

Abstract

An experimental model that mimicked the structure and characteristics of in vivo biofilm infections, such as those occurring in the lung or in dermal wounds where no biomaterial surface is present, was developed. In these infections, microbial biofilm forms as cell aggregates interspersed in a layer of mucus or host matrix material. This structure was modeled by filling glass capillary tubes with an agarose gel that had been seeded with Staphylococcus aureus bacteria and then incubating the gel biofilm in medium for up to 30 h. Confocal microscopy showed that the bacteria formed in discrete pockets distributed throughout the gel matrix. These aggregates enlarged over time and also developed a size gradient, with the clusters being larger near the nutrient- and oxygen-supplied interface and smaller at greater depths. Bacteria entrapped in gels for 24 h grew slowly (specific growth rate, 0.06 h(-1)) and were much less susceptible to oxacillin, minocycline, or ciprofloxacin than planktonic cells. Microelectrode measurements showed that the oxygen concentration decreased with depth into the gel biofilm, falling to values less than 3% of air saturation at depths of 500 μm. An anaerobiosis-responsive green fluorescent protein reporter gene for lactate dehydrogenase was induced in the region of the gel where the measured oxygen concentrations were low, confirming biologically relevant hypoxia. These results show that the gel biofilm model captures key features of biofilm infection in mucus or compromised tissue: formation of dense, distinct aggregates, reduced specific growth rates, local hypoxia, and antibiotic tolerance.

FIG 1

Experimental system. (A) Square glass capillary tubes were loaded with agarose gel seeded with S. aureus and then submerged in air-sparged 1/10-strength TSB. (B) After 24 h of incubation, oxygen profiles were measured by probing with an oxygen microelectrode through the open end of the capillary tube. A small volume of air-sparged medium was placed on top of the gel biofilm to maintain hydration.

FIG 2

Growth curve for S. aureus AH2547 in the gel biofilm model system determined by viable plate counts. Data points show combined results from six independent experiments.

FIG 3

Representative confocal microscopy images of S. aureus AH2547 in the gel biofilm model at various times. The left edge of each image corresponds to the boundary of the gel directly exposed to nutrients and oxygen.

FIG 4

Growth curve for S. aureus AH2547 in the gel biofilm model system determined by image analysis of confocal images like those shown in Fig. 3. The area occupied by cells was determined on the basis of the green fluorescence. Data points show combined results from six independent experiments, with three images being analyzed at each sampling time.

FIG 5

Size of bacterial clusters versus their depth in the gel biofilm model system, based on analysis of images like those in Fig. 3. The data for each time point were fitted with a second-order polynomial. Each of the fitted curves was derived from between 1,026 and 6,796 individual data points.

FIG 6

Comparison of killing in planktonic culture (black bars), 2-h-old gel biofilms (gray bars), and 24-h-old gel biofilms (white bars) of S. aureus AH2547 by three antibiotics. OXC, oxacillin; MNC, minocycline; CIP, ciprofloxacin. *, statistical significance (P < 0.05) for comparisons between planktonic and gel biofilm conditions.

FIG 7

Oxygen concentration profiles in gel biofilms containing S. aureus AH2547. The data points are averages of the oxygen concentrations at five different spots in the gel. The results of two trials from two different gels are shown, and a representative error bar (standard deviation) is shown for each. Error bars on all other points were of similar magnitudes. Zero on the depth scale corresponds to the gel-nutrient medium interface.

FIG 8

Average cell intensity for a shake flask culture (open circles) and a static culture (closed circles) from confocal images of planktonic cells of strain UAMS-1/pEM87 containing the ldh::gfp reporter. Both flasks were grown on the shaker for 4 h, and then one was moved to a static incubator. *, statistical significance (P < 0.001) for comparisons between the two culture conditions.

FIG 9

Analysis of ldh::gfp reporter strain UAMS-1/pEM87 in the gel biofilm system from one representative experiment. (A) Confocal images over a 3-h time period. The left edge of each image corresponds to the gel interface directly exposed to nutrients and oxygen. (B) Green fluorescence images (ldh::gfp reporter strain) at 7 h overlaid with propidium iodide (red). (C) Average cluster green fluorescence intensity versus depth in the gel from 5 to 8 h.