Development of vancomycin and lysostaphin resistance in a methicillin-resistant Staphylococcus aureus isolate.

Glycopeptide resistance in Staphylococcus aureus is poorly understood. The diversity of change documented in cell walls of clinical glycopeptide-intermediate S. aureus (GISA) isolates is evidence that a single genetic or biochemical change cannot account for resistance in all isolates described to date. Therefore, identification of new GISA clinical isolates provides an opportunity to gain insight into the range of adaptive strategies employed by staphylococci to survive in the presence of glycopeptides. In April 1999, a GISA isolate was obtained from the blood of a 63-year-old dialysis patient in Illinois. This isolate was one of six clonally identical MRSA isolates (A-F) serially obtained from the blood of this patient who was receiving vancomycin therapy. All isolates were resistant to oxacillin (MIC > 256 mg/L). The initial isolate had an MIC of vancomycin of 1 mg/L. However, the presence of a subpopulation that could grow in the presence of 5 mg/L of vancomycin indicated that this isolate was predisposed to the acquisition of the GISA phenotype (MIC of vancomycin 10-12 mg/L), which occurred 13 days later, associated with an increased MIC of the endopeptidase lysostaphin and slightly increased cell wall thickness. The first and last isolates in the series, A and F, resisted killing when incubated in vancomycin 2 mg/L, resisted autolysis when incubated in Triton X-100 and had a decreased expression of a c. 116 kDa autolytic band, properties that were different from glycopeptide-susceptible control isolates. Lysostaphin resistance was not accompanied by alterations in the peptidoglycan pentaglycine cross-bridge or a decrease in oxacillin MIC. These data, when taken together with the demonstration of increased cross-linking in isolate F compared with isolate A, demonstrate that vancomycin resistance in these isolates probably occurred by a mechanism different from that of other GISA isolates described to date.