Molecular Characteristics and Prevalence of Rifampin Resistance in Staphylococcus aureus Isolates from Patients with Bacteremia in South Korea

Rifampin resistance (RIF-R) in Staphylococcus aureus (S. aureus) with rpoB mutations as one of its resistance mechanisms has raised concern about clinical treatment and infection prevention strategies. Data on the prevalence and molecular epidemiology of RIF-R S. aureus blood isolates in South Korea are scarce. We used broth microdilution to investigate RIF-R prevalence and analyzed the rpoB gene mutation in 1615 S. aureus blood isolates (772 methicillin-susceptible and 843 methicillin-resistant S. aureus (MRSA)) from patients with bacteremia, between 2008 and 2017. RIF-R prevalence and antimicrobial susceptibility were determined. Multilocus sequence typing was used to characterize the isolate’s molecular epidemiology; Staphylococcus protein A (spa), staphylococcal cassette chromosome mec (SCCmec), and rpoB gene mutations were detected by PCR. Among 52 RIF-R MRSA isolates out of 57 RIF-R S. aureus blood isolates (57/1615, 0.4%; 5 methicillin-susceptible and 52 MRSA), ST5 (44/52, 84.6%), SCCmec IIb (40/52, 76.9%), and spa t2460 (27/52, 51.9%) were predominant. rpoB gene mutations with amino acid substitutions showed that A477D (17/48, 35.4%) frequently conferred high-level RIF resistance (MIC > 128 mg/L), followed by H481Y (4/48, 8.3%). RIF-R S. aureus blood isolates in South Korea have unique molecular characteristics and are closely associated with rpoB gene mutations. RIF-R surveillance through S. aureus–blood isolate epidemiology could enable effective therapeutic management.


Introduction
Staphylococcus aureus (S. aureus) is one of the most notorious and clinically significant pathogens in both community-acquired and nosocomial infections that causes various illnesses, which range from relatively minor local infections to life-threatening systemic infections [1].Widespread antibiotic use has contributed to the emergence of multidrugresistant S. aureus strains, particularly those that are resistant to methicillin, and this has become a global concern [2].Rifampin is an antimicrobial agent with activity against multidrug-resistant S. aureus, including methicillin-resistant S. aureus (MRSA).Thus, rifampin is currently indicated in combination therapy for S. aureus bacteremia, prosthetic joint infections, and prosthetic valve endocarditis [3][4][5].However, in MRSA, rifampin resistance (RIF-R) has emerged and is attributed to rpoB gene mutations [6].RIF-R prevalence has rapidly increased to create a high RIF-R resistance rate [7,8].The molecular characteristics of RIF-R and rpoB mutations in clinical MRSA [8,9] and those of S. aureus bloodstream isolates [10][11][12] have been previously evaluated.However, in South Korea, an accurate understanding of these characteristics is lacking, as there is limited information on the molecular characteristics of RIP-R and rpoB mutations in S. aureus infections.A recent study identified rifamycin resistance and rpoB gene mutations in Staphylococcus species isolates from prosthetic joint infections [13]; nonetheless, there is a paucity of data on the molecular epidemiology and prevalence of RIF-R and rpoB mutations in S. aureus isolates from patients with bacteremia in South Korea.With regard to treatment and infection control, the dissemination of RIF-R isolates in S. aureus bacteremia may pose a serious threat, and this highlights the importance of understanding the molecular epidemiology of S. aureus isolates.
Therefore, we aimed to investigate the prevalence of RIF-R and analyze the rpoB gene mutation in S. aureus isolates obtained from patients with bacteremia.
Among the 52 RIF-R MRSA isolates, 51 isolates (51/52, 98.1%) showed high-level rifampin resistance (MIC ≥ 8 mg/L), and 1 isolate (1/52, 1.9%) showed low-level rifampin resistance (MIC 4 mg/L) (Figure 1).No mutations in the rifampin resistance-determining region (RRDR) of the rpoB gene were detected in 4 of the 52 RIP-R MRSA isolates (4/52, 7.7%), though these isolates were phenotypically resistant to RIF.Amplification of the 714-bp RRDR region of the rpoB gene processed for DNA sequence analysis and the MIC distributions for RIP related to rpoB mutations are shown in Table 1.rpoB gene mutations were detected in 41 out of 48 (41/48, 85.4%) isolates compared with the rpoB gene sequence of the reference strain.We identified 19 different types of mutations in rpoB gene mutation analysis, with 33 single mutations (33/48, 68.8%) and 15 multiple mutations (15/48, 31.3%).Of the 48 isolates, 17 (17/48, 35.4%) showed a mutation at codon 477 (gct to gat) and presented the amino acid substitution A477D, which was associated with high-level RIF resistance (MIC > 128 mg/L).We found several amino acid substitutions of H481Y and S468P that were common mutations of the rpoB proteins and some relatively rare mutations of R484H, S486L, and D471G that some previous studies have reported [2,8].In addition, some isolates in our study carried rare mutations in RRDR, such as Q486L, Q486R, H481R, E568K, and R197L.However, when we sequenced 50 RIF-S MRSA strains as a negative control, we found that there was no detection of amino acid substitutions in any of the 50 RIF-S S. aureus isolates.

Antimicrobial Resistance Profiles of RIF-R MRSA Isolates
The antibiotic resistance profiles of the 52 RIF-R MRSA isolates are shown in Figure 2. Our results revealed that the resistance levels of 52 RIF-R MRSA isolates were as follows: 100% to ampicillin, 94.2% to clindamycin, 92.3% to ciprofloxacin, 94.2% to

δ-Hemolysin Activities As Genotypic Characteristics of S. aureus Isolates by Rifampin
Table 3 shows the differences in the genotypic characteristics of S. aureus isolates with RIF susceptibility and resistance.The agr functionality test was performed on 1615 S. aureus isolates.Fourteen (24.6%) functional agr (hemolytic strains) were resistant to RIF, and 835 (53.6%) agr (hemolytic strains) were susceptible to RIF.There were 43 (75.4%)dysfunctional agr (nonhemolytic strains) resistant to RIF and 723 (46.4%) dysfunctional agr (nonhemolytic strains) susceptible to RFP.

Discussion
In this study, we evaluated the molecular characteristics and prevalence of RIF resistance in S. aureus isolates obtained from patients with bacteremia.Our results suggested that RIF resistance was detected in 0.4% (57/1615) of S. aureus isolates associated with bloodstream infections, and ST5-SCCmec IIb-spat 2460 was the most common type of RIF-R S. aureus isolate.Amino acid A477D substitution was most common in RIF-R MRSA isolates that harbored multiple mutations, which conferred high-level RIF resistance (MIC > 128 mg/L).Notably, the RIF-R MRSA isolates showed high rates of resistance to various antibiotics.
Several recent studies have reported the geographically diverse molecular epidemiology of prevalent MRSA clones in bacteremia, with a high prevalence of ST5 and ST59 (People's Republic of China) [14][15][16], ST 239 (Turkey) [12], ST239 (Iran and Kuwait) [11,17], and ST1011 (Mexico) [10].However, limited information is available regarding the molecular epidemiology of S. aureus in patients with bacteremia in South Korea.ST5-SCCmecII and ST239-SCCmecIII clones were predominant among a relatively small number of blood isolates (n, 45-96) [18,19].Community-associated (CA) MRSA strains of ST72-SCCmec IV accounted for a major proportion of MRSA blood isolates (from 76 to 83) in South Korea [20][21][22].A key strength of our study was the evaluation of a large number of S. aureus blood isolates (n = 1615); ST5-SCCmecII (543/1615, 33.6%) and ST72-SCCmecIV (403/1615, 25.0%) were the major clones in South Korea, which reinforced the results of previous molecular epidemiology studies.To the best of our knowledge, this is the first study to evaluate the molecular characteristics of RIF-R S. aureus blood isolates in South Korea, and we identified molecular characteristics distinct from those reported in previous studies on RIF-R S. aureus clinical isolates [8,9,13,23].
Mutations conferring RIF-R are almost exclusively confined to the rpoB gene in most microorganisms [24], and rpoB mutations can be associated with high-level RIF resistance [25].In our study, we amplified and sequenced portions of rpoB from RIF-R S. aureus isolates.High-level RIF resistance may be attributed to additional mutations within the rpoB gene, as previously described [26].Though sequencing of the RRDR successfully identified mutations in RIF-R MRSA isolates, four RIF-R MRSA isolates (4/52, 7.7%) lacked any mutation in the RRDR region of the rpoB gene despite phenotypic resistance to RIF.This difference might persist because of genotypic variations that are prevailing worldwide, and the presence of mutations in the rpoB gene other than in the RRDR region could not be excluded [27].The lack of RRDR mutations in RIF-R MRSA isolates may be due to the presence of other rare rpoB mutations or another mechanism of RIF-R [27], which is supported by previous reports from several countries [28][29][30].We posit that the four isolates (7.7%) likely possessed additional mutations outside the RRDR because they exhibited higher RIF-R levels than those conferred by the mutations in their RRDR alleles.On the contrary, 15 multiple mutations (15/48, 31.3%) were identified in rpoB gene mutation analysis in our study, all of which were associated with high-level RIF-R (MIC ≥ 8 mg/L).Although most rpoB gene mutations were correlated with high-level resistance [6], it is difficult to identify how much each mutation contributed to the increased MIC.Further molecular analysis of rpoB gene mutations in RIF-R S. aureus isolates to identify the correlation between high-level MIC and mutation at specific codons will provide higher robustness on isolates' ability to survive and affect clinical practice in the region of endemicity.
The amino acid substitution A477D was the most prevalent in our study, which was inconsistent with findings reported from previous studies that showed the H481Y MRSA clone was predominant [8,9,13,31].Furthermore, the transformation of A477Dmutated rpoB into wild-type S. aureus strains resulted in a RIF-R phenotype, indicating that these mutations contribute to rifampicin resistance in S. aureus.In the A477D mutant, this substitution places a negatively charged carboxylate unit in close proximity to an existing carboxylate from H481, which is situated near the protein-DNA interface and thereby increases the negative charge on the surface of the protein and destabilizes the enzyme-DNA interaction, owing to electrostatic repulsion [32].Even if the abovementioned substitution does not face the RIF target region, it can induce a conformational change that indirectly prevents antibiotic binding to the target site, which determines the RIF-R [8].Another interesting finding of our study was that the analyzed isolates lacked the L466S mutation, which was one of the most commonly found mutations in the clinical isolates of other studies [7,9].We posit that there should be regional differences in the mutation codon of RIF-R blood S. aureus isolates, given that only limited information is available regarding the molecular characteristics of rpoB mutations of blood S. aureus isolates in multiple countries, including South Korea.Future multi-center studies using isolates of national or global pathogen resource networks to further characterize the mutations associated with RIF-F and investigate new mutations are essential.
rpoB mutations are associated not only with resistance to RIF but also resistance to other last-line antibiotics for MRSA, including vancomycin and daptomycin [33,34].In particular, the reduced susceptibility to vancomycin that is linked to rpoB mutations may result in poor clinical outcomes and persistent MRSA infections [35].We posit that the potentially decreased susceptibility to vancomycin promoted persistent infection which was associated with the rpoB-A477D mutation; this aspect may be a critical concern in the treatment of MRSA bacteremia because the A477D mutant was most prevalent in our 52 RIF-R MRSA isolates from patients with bacteremia [2,33,34].In addition, given that RIF exposure can be the main selective pressure for decreased susceptibility to RIF and vancomycin and that a higher prevalence of RIF resistance in ST5 was prevalent in previous large-scale S. aureus genome studies [2], judicious use of RIF and preventive infection-control measures against the spread of RIF-R S. aureus clones should be prioritized based on large-scale genomic surveillance in the fight against S. aureus bacteremia.Further research is required to understand whether cross-resistance to other antibiotics develops during prolonged RIF exposure or whether resistance rapidly evolves during exposure to antibiotics [36].
Our study has certain limitations.First, this study was conducted at a single center.Second, we analyzed the clinical outcomes and risk factors of RIF-R S. aureus isolates obtained from patients with bacteremia.Third, we did not evaluate the molecular characteristics of the RIF-R MSSA isolates.We posit that our study could be strengthened further if we had sequenced more sufficient RIF-R and RIF-S S. aureus isolates, and it has limitations in being able to acquire strong confirmation that RIF-R in S. aureus isolates was caused by a rpoB mutation.However, our study included a relatively large number of S. aureus isolates from bacteremia (n = 1615) with RIF-R MRSA isolates (n = 52) and was conducted in a 2700-bed tertiary-care teaching hospital where patients from across the country were admitted during a 10-year period.Therefore, our study has important clinical implications for antibiotic use and provides useful information for preventive infection-control measures based on the molecular epidemiology of RIF-R S. aureus blood isolates in South Korea.

Collection of S. aureus Isolates
The study sample consisted of S. aureus strain isolates obtained from Asan Medical Center, Seoul, Republic of Korea, from 2008 to 2017.The S. aureus samples were plated on a blood agar plate.This sterile medium was streaked with a cotton swab, and the plates were incubated overnight at 37 • C. The isolate was grown to screen for and analyze S. aureus.The strains were stored in 20% glycerol-tryptic soy broth at −80 • C (Becton Dickinson, Sparks, MD, USA).We determined the methicillin resistance of S. aureus isolates based on the oxacillin MIC and the presence of the mecA gene.

Agr Functionality Test
We used δ-hemolysin activity to determine agr functionality by cross-streaking vertically to RN4220 and a test strain on a sheep blood agar plate (BAP).The β-hemolysin produced by RN4220 enables detection of δ-hemolysin [37].δ-hemolytic activity was indicated by an enhanced area of hemolysis at the intersection of the streaks.

Antimicrobial Susceptibility Tests
The antimicrobial susceptibility profiles of the S. aureus isolates were determined using the broth microdilution method according to the Clinical and Laboratory Standard Institute (CLSI) guidelines [38].For the broth microdilution method, serial two-fold dilutions were carried out in cation-adjusted Mueller-Hinton II broth (Becton Dickinson, Sparks, MD, USA) in microtiter plates according to standard criteria [38].The MIC was determined using the broth microdilution method, and each spot was inoculated with 10 6 CFU.After incubation at 37 • C for 16-20 h, the MIC value was considered when the bacteria did not grow at the minimum antibiotic concentration.The reference strain from the American Type Culture Collection 29,213 was used for quality control.The MIC for rifampin (Sigma, Darmstadt, Germany) was determined using the broth microdilution method, following standard criteria [38].Based on the CLSI guidelines, the isolates were classified as RIF susceptible (MIC ≤ 1 mg/L) or RIF resistant (MIC ≤ 4 mg/L).

Molecular Typing 4.4.1. Detection of the mecA Gene
The mecA gene sequence (532 bp) of all the MRSA isolates was amplified by PCR.The amplification of the mecA gene was achieved using mecA1 (5 AAA ATC GAT GGT AAA GGT TGG C 3 ) and mecA2 (3 AGT TCT GCA GTA CCG GAT TTG C 5 ) primers and sequence analysis.The PCR conditions involved an initial denaturation for 3 min, followed by 30 cycles at 94 • C for 30 s, 55 • C for 30 s, and 72 • C 30 s and a final extension at 72 • C for 4 min.The PCR products (10 µL) were separated on a 1% agarose gel in 0.5 × Tris-borate-EDTA buffer at 100 V and visualized using RedSafe.

SCCmec Typing
The SCCmec typing of the MRSA isolates was performed using the multiplex PCR method described by Oliveira and de Lencastre [40].The eight loci (A through H) and specific pairs of primers for SCCmec types and subtypes I, II, III, and IV have been described previously [41].The multiplex PCR conditions included an initial denaturation at 4 min, followed by 30 cycles at 94 • C for 30 s, 53 • C for 30 s, and 72 • C for 1 min as well as a final extension at 72 • C for 4 min.The PCR products (10 µL) were separated on a 1.8% agarose gel in 0.5 × Tris-borate-EDTA buffer at 135 V and visualized using RedSafe.

Spa Typing
The Staphylococcus protein A (spa) variable repeat region from each MRSA isolate was amplified using simplex PCR oligonucleotide primers, as described previously [40,42].The purified spa PCR products were sequenced, and the typing of spa was performed using the public spa database website (http://spa.ridom.de/,accessed on 10 August 2023) for all S. aureus isolates.

Statistical Analysis
The statistical distribution and clinical characteristics of the patients were compared using cross-analysis and the chi-square test, respectively.Statistical significance was set at p < 0.05.All statistical analyses were performed using SPSS version 24.0 (SPSS Inc., Chicago, IL, USA).

Table 1 .
Correlation of rpoB gene mutations and the level of resistance to rifampin in MRSA isolates.

Table 1 .
Correlation of rpoB gene mutations and the level of resistance to rifampin in MRSA isolates.

Table 2 .
Molecular characteristics of the 52 rifampin-resistant MRSA isolates.

Table 2 .
Molecular characteristics of the 52 rifampin-resistant MRSA isolates.

Table 3 .
Genotypic characteristics of the S. aureus isolates stratified by their rifampin-resistance status.