Sequential evolution of virulence and resistance during clonal spread of community-acquired methicillin-resistant Staphylococcus aureus

Abstract

The past two decades have witnessed an alarming expansion of staphylococcal disease caused by community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA). The factors underlying the epidemic expansion of CA-MRSA lineages such as USA300, the predominant CA-MRSA clone in the United States, are largely unknown. Previously described virulence and antimicrobial resistance genes that promote the dissemination of CA-MRSA are carried by mobile genetic elements, including phages and plasmids. Here, we used high-resolution genomics and experimental infections to characterize the evolution of a USA300 variant plaguing a patient population at increased risk of infection to understand the mechanisms underlying the emergence of genetic elements that facilitate clonal spread of the pathogen. Genetic analyses provided conclusive evidence that fitness (manifest as emergence of a dominant clone) changed coincidently with the stepwise emergence of (i) a unique prophage and mutation of the regulator of the pyrimidine nucleotide biosynthetic operon that promoted abscess formation and colonization, respectively, thereby priming the clone for success; and (ii) a unique plasmid that conferred resistance to two topical microbiocides, mupirocin and chlorhexidine, frequently used for decolonization and infection prevention. The resistance plasmid evolved through successive incorporation of DNA elements from non-S. aureus spp. into an indigenous cryptic plasmid, suggesting a mechanism for interspecies genetic exchange that promotes antimicrobial resistance. Collectively, the data suggest that clonal spread in a vulnerable population resulted from extensive clinical intervention and intense selection pressure toward a pathogen lifestyle that involved the evolution of consequential mutations and mobile genetic elements.

Keywords: MRSA; antimicrobial resistance; evolution; virulence.

Fig. 1.

Bacterial phylogeny reveals the emergence and spread of a dominant clone (USA300-BKV) in the Orthodox Jewish community. Maximum-likelihood phylogenetic trees of 92 isolates obtained from patients residing in Orthodox-associated zip codes (shown in red) compared with (A) 16 representative USA300 isolates from adults and children in the same hospital or (B) 68 USA300 strains from northern Manhattan and the Bronx. Ninety-three percent (86/92) of isolates from patients residing in Orthodox-associated zip codes clustered within a unique clade (USA300-BKV; BKV clade in red). The remaining six isolates from patients residing in high-risk zip codes clustered with contemporary isolates from our hospital and isolates from northern Manhattan and the Bronx. The trees are rooted using USA300_FPR3757 and the distantly related S. aureus isolate MRSA131 as outgroups. Bootstrapping value for the USA300-BKV clade divergence branch is indicated. (C) Bayesian phylogenetic reconstruction of BKV isolates estimated from core genome mutations. USA300-BKV clade branches are highlighted in red. Predicted dates of divergence from a most recent common ancestor are indicated.

Fig. 2.

A single mutation in pyrR affects S. aureus fitness in vitro and in vivo. (A) Schematic of pyrR regulation of S. aureus pyrimidine biosynthetic and urea pathways through pyr operon gene repression. (B) Growth curves of USA300-BKV isolates (BKV isolates) and pyrR-complemented BKV isolates in pyrimidine-limited chemically defined medium. (C) Real-time qPCR validation of the impact of pyrR mutation on pyr operon genes carA and pyrB. Gene expressions were normalized to pyrR gene. Data represent mean ± SD (n = 3). *P < 0.05, **P < 0.01, ***P < 0.001 using ANOVA with multiple t test comparisons. (D) Number of unique mutations targeting pyr genes in independent BKV-isolate genomes. (E) Schematic of the mouse model of colonization and transmission. Parental mice were orally inoculated with a 1:1 ratio of 10⁷ cfu of USA300 strain JE2 + pyrR^WT vs. JE2 (a control strain of USA300) or strain USA300-BKV + pyrR^WT vs. BKV isolate. Ceca were harvested after 4 d of colonization. (F) Competitive colonization and transmission assay (competitive index, Left) and quantification of bacteria in stool (cfu, Right) were determined from the cecum of inoculated (colonization) and cohoused pups (transmission). Transmission indicates quantification of bacteria in the stool of uninoculated mice that were cohoused with inoculated mice. Median values are shown, and each symbol is the CI from one mouse (Left) or cfu from single mouse (Right). **P < 0.01 by Wilcoxon signed-rank tests.

Fig. 3.

USA300-BKV isolates (BKV isolates) characterized by the presence of mosaic ϕ11 (ϕ11-mos) demonstrate enhanced virulence in a murine abscess model of infection compared with phage-free control USA300 isolates. (A) Comparative genome map of the prototypical ϕ11 and the BKV isolate-associated ϕ11-mos. Arrows indicate predicted ORFs and the direction of the transcription in phage functional modules. Unique ORFs are indicated in blue. (B) Genes encoding proteins with low homology between the BKV isolate-associated ϕ11-mos and prototypical ϕ11 are indicated. (C) Distribution of ϕ11-mos among BKV isolates (blue). (D) Selected BKV isolates (n = 8) from the minimum spanning tree representing a set of phage-free (n = 3) and ϕ11-mos–containing BKV isolates (n = 5). Size of murine skin lesions [(n = 10) five mice per group with two abscesses per mouse] at 24, 48, and 72 h after s.c. infection with ∼1 × 10⁷ cfu of the indicated strain. Statistical analyses were performed with the Kruskal–Wallis test; ***P < 0.001. The results were corrected for multiple comparisons by using the Bonferroni-corrected threshold.

Fig. 4.

The mosaic portion of ϕ11 increases skin abscess size. USA300-LAC lysogens of mosaic ϕ11 [(ϕ11-mos) produced by induction of USA300-BKV_28] and prototypical ϕ11 (produced by induction of RN451), which differ only in the region corresponding to the mosaic block, were made and compared in a murine abscess model of infection. (A) Abscess size at 24, 48, and 72 h after s.c. infection with ∼1 × 10⁷ cfu of the indicated strain [(n = 10) five mice per group with two abscesses per mouse]. Statistical analyses were performed with the Kruskal–Wallis test after multiple comparison correction. (B) Representative pictures of murine skin abscesses at 3 d after s.c. infection with the indicated strain at either a 1 × 10⁷ (Top) or a 5 × 10⁷ cfu (Bottom) inoculum.

Fig. 5.

Emergence of a dominant clone coincides with acquisition of an S. epidermidis-recombinant mupirocin resistance plasmid, pBSRC1. (A) Structure and characteristics of pBSRC1. Gene names and location are indicated in the first inner circle (gray). mupA and qacA/B are highlighted in orange and purple, respectively. Other inner circles show perfect BLAST homology results to S. epidermidis (blue) and S. aureus (red) plasmid sequences. The site of S. aureus pUSA01 integration in pBSRC1 is indicated. (B) Minimum spanning tree (MST) based on a patristic genetic comparison of 86 isolates obtained from patients residing in Orthodox-associated zip codes compared with 22 control USA300 isolates from adults and children in the same hospital (16 representative USA300 isolates from patients residing in non–high-risk zip codes, and six USA300 strains from Orthodox-associated zip-codes that were not part of the USA300-BKV clade). The first MST (Top) shows control strains (blue) and the USA300-BKV isolates (red). The second MST (Middle) highlights the distribution of mupA⁺ pBSRC1 containing strains (orange), whereas the third MST (Bottom) represents the distribution of qacA/B pBSRC1 containing strains (purple). Each node represents a single strain. Distance between nodes is arbitrary, whereas node size is proportional to the number of connections. (C) Working model of the stepwise assembly of pBSRC1 in USA300-BKV isolates.