Repeated Emergence of Variant TetR Family Regulator, FarR, and Increased Resistance to Antimicrobial Unsaturated Fatty Acid among Clonal Complex 5 Methicillin-Resistant Staphylococcus aureus

Abstract

Resistance-nodulation-division (RND) superfamily efflux pumps promote antibiotic resistance in Gram-negative pathogens, but their role in Gram-positive pathogens, including methicillin-resistant Staphylococcus aureus (MRSA) is undocumented. However, recent in vitro selections for resistance of S. aureus to an antimicrobial fatty acid, linoleic acid, and an antibiotic, rhodomyrtone, identified H121Y and C116R substitution variants, respectively, in a TetR family regulator, FarR, promoting increased expression of the RND pump FarE. Hypothesizing that in vivo selection pressures have also promoted the emergence of FarR variants, we searched available genome data and found that strains with FarR^H121Y from human and bovine hosts have emerged sporadically in clonal complexes (CCs) CC1, CC30, CC8, CC22, and CC97, whereas multiple FarR variants have occurred within CC5 hospital-associated (HA)-MRSA. Of these, FarR^E160G and FarR^E93EE were exclusive to CC5, while FarR^C116Y, FarR^P165L, and FarR^G166D also occurred in nonrelated CCs, primarily from bovine hosts. Within CC5, FarR^C116Y and FarR^G166D strains were polyphyletic, each exhibiting two emergence events. FarR^C116Y and FarR^E160G were individually sufficient to confer increased expression of FarE and enhanced resistance to linoleic acid (LA). Isolates with FarR^E93EE were most closely related to S. aureus N315 MRSA and exhibited increased resistance independently of FarR^E93EE. Accumulation of pseudogenes and additional polymorphisms in FarR^E93EE strains contributed to a multiresistance phenotype which included fosfomycin and fusidic acid resistance in addition to increased linoleic acid resistance. These findings underscore the remarkable adaptive capacity of CC5 MRSA, which includes the polyphyletic USA100 lineage of HA-MRSA that is endemic in the Western hemisphere and known for the acquisition of multiple resistance phenotypes.

Keywords: MRSA; Staphylococcus aureus; TetR family regulator; efflux pumps; mechanisms of resistance.

FIG 1

Grapes plot of six major FarR clusters and associated variants (A) and mapping of FarR clusters on a proteome based Staphylococcus aureus phylogenetic map (B). For the Grapes plot (A), the minimum spanning tree was constructed by alignment of 564 full-length FarR proteins using GrapeTree, which identified 6 primary clusters and associated variants. Points represent groups of identical elements, with point size correlated with number of elements on a log scale. Phylogenetic distance is scaled to two single-amino acid polymorphisms (2SAAP). The asterisk on cluster 2 (*) marks FarR variant 2c where duplication of an amino acid codon at E93 alters the alignment. For the phylogenetic map (B), conceptually translated nucleotide sequences available from PATRIC are shown for 574 S. aureus and 1 S. argenteus strains using PhyloPhlAn3. Initial phylogenetic analysis used FastTree, which was refined with RaxML. The six primary FarR clusters are mapped on ring 1, while clonal complex associations are mapped on ring 2. The colored legend for clonal complex designations is presented in the same order of appearance as on ring 2, beginning with CC5 and progressing in descending order in each column from left to right, ending with the S. argenteus outgroup.

FIG 2

Distribution of farR variants within the representative worldwide S. aureus CC5 population. The phylogenetic distribution of strains with variant farR genes within the CC5 phylogeny was determined through a two-phase bioinformatics analysis. The first phase consisted of comparing polymorphisms in 119 CC5 strains with single amino acid substitutions in FarR to 598 CC5 reference strains (26), from which 26 reference genomes were selected that (i) subtended the nodes to which the new strains with FarR variants attached, (ii) provided examples of sister nodes of strains with FarR variants, and (iii) provided examples of the various CC5 clades previously defined. In the second phase, these 119 FarR variant strains and 26 reference genomes were analyzed with GATK to call SNPs and indels, as well as invariant core nucleotides relative to S. aureus JH1. The resulting bi-allelic SNPs and invariant core nucleotides were analyzed by PhyML and ClonalFrameML to generate a phylogeny and correct branch length for recombination. Here, strains with variant FarR proteins are placed in the context of CC5 phylogeny. The major CC5 clades, consisting of Basal, CC5-I, and CC5-II (-IIA and -IIB), as defined previously (26), are labeled on the root axis of the dendrogram. Biosample numbers from the NCBI genome sequence entries for each strain are shown adjacent to the branch structures. These strains are listed in the same order in Table S4, which also provides their common names (where available), SCCmec genotypes, and a list of polymorphisms for each strain relative to the S. aureus JH1 reference genome (83). Columns adjacent to the Biosample numbers provide information on country of origin (A), multilocus sequence type (MLST) designation (B), FarR variant (C), and presence or absence of mecA (D).

FIG 3

Graphs showing 24-h endpoint growth of USA100 FarR variants in tryptic soy broth (TSB) + 1,200 μM linoleic acid (LA). Cultures of USA300, N315, FAR7, or CC5 MRSA harboring either FarR^C116Y from the United States (A) or Canada (B), FarR^E93EE (C), FarR^E160G (D), or FarR^P165L from the United States (E) or Canada (F) were inoculated into triplicate tubes containing 3 mL of TSB + 1,200 μM LA + 0.1% dimethyl sulfoxide (DMSO) at an optical density at 600 nm (OD₆₀₀) of 0.01, followed by incubation at 37°C with orbital shaking. Growth (OD₆₀₀) was determined after 24 h. Each data point represents the mean ± standard deviation (SD) from triplicate cultures. Statistically significant differences (****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05) compared to N315 were determined by Tukey’s multiple-comparison test.

FIG 4

Growth of representative FarR variant strains in TSB + 200 μM LA. Cultures were inoculated to OD₆₀₀ = 0.01 into 96-well microtiter plates containing 200 μL of TSB supplemented with a subinhibitory concentration of 200 μM LA + 0.1% DMSO. Plates were incubated at 37°C with orbital shaking, and growth (OD₆₀₀) was monitored hourly. (A) FarR^C116Y (M1545, NM0256, NM0256) and FarR^E160G (hp.082 and hp.099) strains. (B) FarR^E93EE (M1516) and FarR^P165L (DAR1157, DAR948, NM0135) strains. Growth was compared to the ST-5 HA-MRSA (health care-associated MRSA) reference strain N315 and FAR7 (FarR^H121Y). Each data point represents the mean ± standard error of the mean (SEM) from 6× 200-μL wells in 96-well microtiter plates.

FIG 5

farR^C116Y and farR^E160G are sufficient to confer either increased resistance to, or growth advantage on exposure to, LA. Growth (OD₆₀₀) of farRΦNE harboring either pLI50 vehicle, pLIfarR, or variants farR^H121Y, farR^C116Y, farR^E160G, farR^P165L, or farR^E93EE was assessed in TSB containing 1,200 μM (A) or 200 μM (B to D) LA + 0.1% DMSO. For panels A and B, endpoint growth (OD₆₀₀) in culture tubes was assessed after 24 h, while panels C and D represent growth in microtiter plates with automated monitoring over 24 h. Each value represents the mean ± SD of triplicate 3-mL culture tubes (A and B) or the mean ± SEM of 6× 200-μL wells in 96-well microtiter plates (C and D). Statistically significant differences (***, P < 0.001; **, P < 0.01) compared to farRΦNE + pLIfarR wild type were determined by Tukey’s multiple-comparison test.

FIG 6

farR variants promote increased expression of farE in a farRE::lux luciferase reporter construct. farRΦNE was transformed with either pGYfarRE::lux or variant lux¹ (farR^H121Y), lux² (farR^C116Y), or lux³ (farR^E160G) constructs where luciferase activity is driven from the P_farE promoter under the control of wild-type or variant farR genes. Cultures were grown in 125-mL flasks containing 25 mL TSB for 3 h followed by supplementation with 40 μM LA + 0.1% DMSO. (A) Growth (OD₆₀₀) and luciferase activity (relative light units [RLU]/OD₆₀₀) was quantified at 30-min intervals. (B) Comparison of luciferase activity (RLU/OD₆₀₀) at 1 h after addition of LA. Each data point represents the mean OD₆₀₀ ± SD from triplicate flasks. Statistically significant differences (****, P < 0.0001) compared to wild-type farER were determined by Tukey’s multiple-comparison test.

FIG 7

Resistance and growth phenotypes of USA300 harboring transposon insertions in genes that exhibit frameshift mutations in FarR^E93EE strains. Cultures of USA300 or isogenic variants ΦΝΕ178 (A), ΦΝΕ1511 (B), or ΦΝΕ1154 (C), with transposon insertions in leuA, lysR, or uhpT, respectively, were grown in TSB containing the indicated concentrations of LA + 0.1% DMSO. Each datapoint represents the mean OD₆₀₀ ± SD from triplicate 3-mL tube cultures after 24 h growth. (D) Growth of USA300 and isogenic ΦΝΕ1511 (lysR::tn) in chemically defined medium containing 0.4% glucose (CDM-G) or 0.25% histidine (CDM-H) as a carbon source. All data points represent the mean ± SD from triplicate 3-mL tube cultures after 24 h growth. Statistically significant differences (****, P < 0.0001; ***, P < 0.001; *, P < 0.05) compared to S. aureus USA300 were determined by Tukey’s multiple-comparison test.