Topical Antibiotic Use Coselects for the Carriage of Mobile Genetic Elements Conferring Resistance to Unrelated Antimicrobials in Staphylococcus aureus

Abstract

Topical antibiotics, such as mupirocin and fusidic acid, are commonly used in the prevention and treatment of skin infections, particularly those caused by staphylococci. However, the widespread use of these agents is associated with increased resistance to these agents, potentially limiting their efficacy. Of particular concern is the observation that resistance to topical antibiotics is often associated with multidrug resistance, suggesting that topical antibiotics may play a role in the emergence of multidrug-resistant (MDR) strains. New Zealand (NZ) has some of the highest globally recorded rates of topical antibiotic usage and resistance. Using a combination of Pacific Biosciences single-molecule real-time (SMRT) whole-genome sequencing, Illumina short-read sequencing, and Bayesian phylogenomic modeling on 118 new multilocus sequence type 1 (ST1) community Staphylococcus aureus isolates from New Zealand and 61 publically available international ST1 genome sequences, we demonstrate a strong correlation between the clinical introduction of topical antibiotics and the emergence of MDR ST1 S. aureus We also provide in vitro experimental evidence showing that exposure to topical antibiotics can lead to the rapid selection of MDR S. aureus isolates carrying plasmids that confer resistance to multiple unrelated antibiotics, from within a mixed population of competitor strains. These findings have important implications regarding the impact of the indiscriminate use of topical antibiotics.

Keywords: Staphylococcus aureus; coselection; fusidic acid; multidrug resistance; mupirocin; plasmids; topical antibiotics.

FIG 1

(A) Schematic diagram illustrating the genetic organization of the SCCfusC region of NZ14487 in comparison to the fusC-harboring SCC₄₇₆ element of MRSA476 (GenBank accession no. NC_002953). Regions of similarity are joined by gray lines. The fusC and ccrBA genes are highlighted. (B) Schematic diagram depicting the linear genomic comparisons of plasmid pNZAK1 from S. aureus strain NZ14487 and plasmid pMW2 from S. aureus strain MW2 (GenBank accession no. AP004832). Regions of DNA similarity are joined by gray lines. The location of predicted antimicrobial resistance genes (cadDX, blaRZ, qacRA, and mupA) are shown with red arrows. (C) Segregational stability of plasmid pNZAK1 in Staphylococcus aureus NZ14487. Cultures of S. aureus NZ14487 were serially passaged in either nonselective BHI broth or in broth supplemented with 100 mg/liter mupirocin. Plasmid loss was determined at the days indicated. The mean of triplicate independent cultures grown under each test condition are shown, with error bars representing the standard error of the mean (SEM).

FIG 2

Evolutionary history of the S. aureus ST1 lineage in NZ. (A) Maximum clade credibility (MCC) tree, generated in BEAST version 1.8.3 from 14,350-site filtered single-nucleotide polymorphisms (SNPs) in the core genome of 179 isolates, illustrating the global phylogeny of ST1 S. aureus. Branches are colored to reflect statistical support; those possessing a posterior probability of <95% are red. The 95% highest posterior density interval (HPDI) is shown by the blue horizontal lines. Branch annotations provide country of isolation (dark green, NZ; yellow, Australia; light blue, UK; red, Malaysia; light green, India; dark blue, Brazil; gray, Iraq, royal blue, France; purple, USA) and the mecA, fusC, blaZ, mupA, and qacA resistance gene status. The predicted pNZAK1 acquisition event is indicated. (B) Bayesian Skyline plot showing the change in the effective population size (EPS) of the ST1 lineage with time. The median EPS is indicated by the central black line, and the 95% HPDI is bordered by the blue lines above and below the median value. As described for panel A, the predicted pNZAK1 acquisition event is indicated.

FIG 3

(A) Competitive index assays showing the proportion of MRSA within a mixed population of competing S. aureus isolates following nonselective growth (red bars), or following exposure to either 6 mg/liter oxacillin (yellow bars) or 8 mg/liter FA (blue bars) for the number of days indicated. (B) The proportion of MDR isolates within a mixed population of competing S. aureus strains following 7 days of nonselective growth (red) or 7 days of exposure to 8 mg/liter FA (blue) or 100 mg/liter mupirocin (green). For both panels, the data represent the mean of the results of three independent assays, with error bars showing the SEM.