To permeabilize the bacteria for uptake of the FISH probe, the tissue was treated with 0.5 mg mL−1 lysozyme (Sigma-Aldrich, St Louis, MO) in 0.1 M Tris-HCl (Sigma-Aldrich) at pH 8.0 and 0.05 M Na2EDTA (Sigma-Aldrich) for 3 h at 37 °C and washed with ultrapure water. The samples were dehydrated in a graded series of ethanol washes (50%, 80%, and 100%) for 3 min at each concentration. FISH was performed as described previously (Hogardt et al., 2000; Kempf et al., 2000; Nistico et al., 2009) using the 16S ribosomal probe sequences: Sau 5′-(GAAGCAAGCTTCTCGTCCG)-3′(16S 69–87) (Kempf et al.,

2000) labeled JNK inhibitor in vitro with Cy3 (a green fluorescent fluorophore), which was specific for S. aureus. We used the nucleic acid stain Syto59 (red) as a general stain to stain all bacteria and host nuclei, so

that S. aureus would be dual stained both green and red and appear yellow or orange and non-S. aureus bacteria would only stain with the Syto59 stain and appear red. Bacteria stained with only the Syto59 are readily distinguished from host check details nuclei (which also take up the nucleic acid stain) on the basis of size (bacterial cocci are approximately 1 μm in diameter, whereas the nuclei of host cells are approximately 8 μm) and morphology (Hall-Stoodley et al., 2006; Nistico et al., 2009). For a positive FISH control, we stained MRSA cells grown from a patient with an infected elbow after revision surgery of a total elbow arthroplasty attached to a gelatin-coated slide. The individual cocci were readily discernible (data not shown). To control for nonspecific binding, we stained three pieces of tissue independently

with the NonEub338-Cy3 5′-(ACTCCTACGGGAGGCAGC)-3′ probe, which has no known complementation to any 16S rRNA sequences (Kempf et al., 2000; Manz et al., 1992). Reflected confocal microscopy with the 488-nm laser was used to visualize the tissue over a range of magnifications and a minimum of eight different fields of view in each specimen. The FISH-stained tissue was mounted in a 35-mm Petri dish on 0.5% low-temperature-setting Idelalisib supplier agarose and submerged in HBSS before imaging using CLSM. The Ibis assay positively identified both S. aureus and Staphylococcus epidermidis in the tissue, and also noted the presence of the mecA gene for methicillin resistance. The confidence based on the 16 primer sets was 1.00, 0.92, and 1.00, respectively. There were approximately 10 times more S. aureus than S. epidermidis based on counts of 3889 genomes per well and 452 genomes per well, respectively. The mecA gene returned 8184 genomes per well, suggesting, based on the numbers, that the S. aureus was an MRSA strain. However, from these data alone, we could not draw firm conclusions regarding the mecA status of either staphylococcal species, except that at least one was likely methicillin-resistant. No other bacterial species were detected.