In-Host Evolution of Daptomycin Resistance and Heteroresistance in Methicillin-Resistant Staphylococcus aureus Strains From Three Endocarditis Patients

Abstract

Background

Daptomycin is considered an important alternative for the treatment of methicillin-resistant Staphylococcus aureus (MRSA). However, treatment failures associated with daptomycin nonsusceptibility isolates have been reported in recent years.

Methods

In this study, we investigated serial MRSA strains from 3 endocarditis patients who had breakthrough bacteremia, despite treatment with daptomycin. The strains were analyzed by whole-genome sequencing, molecular typing, and mutation screening. Population analysis and growth curves were also applied to evaluate heteroresistance and fitness cost.

Results

This series of MRSA strains belonged to ST5, ST59, and ST4513. The daptomycin minimum inhibitory concentrations for these MRSA strains increased after daptomycin exposure, whereas daptomycin-resistant strains emerged with mutations in mprF and yycH. Population analysis profiling results demonstrated the presence of a daptomycin-heteroresistant subpopulation among daptomycin-susceptible MRSA strains, and no significant fitness cost was observed within these heteroresistant MRSA clones.

Conclusions

We confirmed that daptomycin heteroresistance and resistance could emerge rapidly in MRSA strains of different lineages after daptomycin exposure. Further studies to fully understand the mechanism(s) underlying daptomycin resistance in MRSA are required.

The spread of multidrug-resistant bacteria such as methicillin-resistant Staphylococcus aureus (MRSA) is a worldwide problem undermining the clinical treatment of such infections [1]. Bloodstream infections and infective endocarditis caused by MRSA thus impose serious global social and economic burdens, leading to increased morbidity and mortality; however, the choice of antibiotics for the treatment of these infections remains limited [2]. Vancomycin treatment has recently been the mainstream method to treat MRSA bacteremia, but there are concerns about its efficacy [3]. Loss of susceptibility and heteroresistance to vancomycin, as well as high treatment failure rates for MRSA infections, are limitations to successful therapy with this drug [4]. Daptomycin provides an alternative to vancomycin for the treatment of S aureus bacteremia and endocarditis due to its rapid bactericidal activity [5]. Accordingly, this drug was proposed to be effective against S aureus bacteremia and right-sided endocarditis in a randomized trial [6]. Furthermore, daptomycin has a unique mode of action, with rare cross-resistance occurring with other antibiotics, making it a useful choice for the treatment of serious infections caused by MRSA [7].

Although daptomycin is a valid alternative to vancomycin for MRSA bacteremia and endocarditis therapy, treatment failure associated with daptomycin nonsusceptibility isolates has been reported. Both in clinical isolates and laboratory derivatives, the main mechanisms underlying this resistance involve cell membrane (mprF), cell wall (walKR and dltABCD), and ribonucleic acid polymerase subunits (rpoC and rpoB), and the dlt operon seems to contribute the most to this process [8–12]. In addition, heteroresistance can lead to daptomycin treatment failure against MRSA. “Heteroresistance” describes a phenomenon in which subpopulations of bacteria that appear to be genetically identical show a range of susceptibilities to a specific antibiotic. Unfortunately, a lack of standard methods to detect heteroresistance has led to inappropriate use and neglect or an underestimation of this phenomenon [2]. Of note, heteroresistance causes antibiotic failure for bacterial infections that are considered antibiotic susceptible, and this might continue to be a source of such treatment failures [13].

As is known, the predominant MRSA clones in healthcare settings and communities in China are different from those in United States and European countries. To date, studies on daptomycin resistance or heteroresistance in Chinese MRSA clones have been rarely reported. In this study, we report the host evolution of MRSA strains repeatedly isolated from 3 patients with endocarditis after daptomycin exposure.

MATERIALS AND METHODS

Patient and Isolate Data

Twenty sequential MRSA strains were isolated from blood cultures of 3 endocarditis patients (patient A, patient B, and patient C) at the Sir Run Run Shaw hospital from September 2017 to April 2018. The characteristics of isolates and related clinical data are listed in Table 1. Clinical data were obtained from patient electronic medical records.

Identification and Antimicrobial Susceptibility Testing

The identification of MRSA strains was performed using the VITEK 2 system (bioMérieux, Lyon, France). The susceptibility to cefoxitin, vancomycin, teicoplanin, linezolid, and daptomycin were evaluated using agar dilution and broth microdilution methods according to the Clinical and Laboratory Standards Institute ([CLSI] 2018) guidelines [14]. Staphylococcus aureus ATCC 29213 was used as a reference strain for quality control.

Pulsed-Field Gel Electrophoresis

Pulsed-field gel electrophoresis (PFGE) was performed according to the method described by Bannerman et al [15] with some modifications. The restriction patterns were analyzed with BioNumerics 7.0 (Applied Maths BVBA, Sint-Martens-Latem, Belgium). Interpretation was performed according to Tenover’s criteria [16]. The PFGE profiles of the 20 isolates are shown in Supplementary Figure 1.

Whole-Genome Sequencing and Mutation Screening

The genomic deoxyribonucleic acid (DNA) of the 20 MRSA strains was extracted using a QIAamp DNA minikit (QIAGEN, Valencia, CA). The quality of DNA was determined by gel electrophoresis and a NanoDrop 2000 spectrophotometer (NanoDrop Technologies, Wilmington, DE). All strains were sequenced on an Illumina HiSeq X Ten platform using the 2 × 150-base pair paired-end mode (Illumina, San Diego, CA). The derived short reads were assembled into contigs with CLC Genomics Workbench software (version 9.5.1; CLC bio). Nanopore sequencing using a MinION sequencer (Oxford Nanopore Technologies, Oxford, UK) was performed for the genomic DNA of the SAW1 strain. The genomes were annotated with Prokka [17] version 1.11. Then, the mutations were detected using breseq [18], which mapped the sequence reads to the first strain as the reference genome. The regions containing the detected single-nucleotide polymorphisms were amplified by polymerase chain reaction (PCR), and the PCR products were sent to Biosune (Biosune, Hangzhou, China) for Sanger sequencing.

The genome of 20 MRSA strains were submitted to the NCBI (National Center for Biotechnology Information) database (Supplementary Table 1). The accession numbers of SAW1 genome are CP045468–CP045471.

Core Genome Multilocus Sequence Typing

Genome assemblies were imported into SeqSphere + software (version 4.1.9; Ridom) as FASTA files for core genome multilocus sequence typing (cgMLST) analysis. We applied stable cgMLST to identify cluster types with default parameters using the S aureus COL-type strain as a reference and a standard set of 1861 genes for gene-by-gene comparisons. For distance calculations, we used only cgMLST target genes that were present in all 20 isolates, meaning that 42 of the 1861 genes were not used. The minimum spanning trees for different MRSA groups were constructed using SeqSphere + software. The whole-genome sequencing (WGS) data were used for SCCmec typing and MLST using SeqSphere + software with standard procedures.

Fitness Cost

Fitness cost was analyzed as previously described [19]. In brief, the growth rate was measured at the start of the exponential phase, and the relative growth rate of each strain versus that of the first strain was calculated using an R script based on the OD₆₀₀ curves. A Student’s t test was used to assess differences between the means, with a significant probability considered at P ≤ .05.

Population Analysis

Population analysis profiling (PAP) was performed for daptomycin as described by Cui et al [20]. In brief, 50 μL of a ~10⁸ colony-forming unit (CFU)/mL suspension of each strain was dropped onto Tryptic Soy Broth plates containing 50 μg/mL calcium with daptomycin concentrations ranging from 0 to 4 or 8 μg/mL. Plates were incubated for 48 hours at 37°C, and the log-transformed number of CFUs was plotted against daptomycin with different concentrations. Community-associated MRSA (CA-MRSA) strain SA268 and MRSA strain COL were included as controls.

Statistical Methods

Statistical analysis was performed using GraphPad Prism 7.0. All error bars show the standard error of the mean, and data were analyzed using a Student’s t test for comparisons of fitness cost. P ≤ .05 were considered statistically significant.

Ethics

This study was approved by the local ethics committees of Sir Run Run Shaw hospital with a waiver of informed consent (Approval No. 20190426-2).

RESULTS

Methicillin-Resistant Staphylococcus aureus Strains From Endocarditis Patients

The clinical and microbiological data of the MRSA strains used in this study are shown in Figure 1 and Table 1. All clinical MRSA isolates were obtained from blood cultures of patients with endocarditis. Each series included the initial infecting isolate, which was susceptible to daptomycin with a minimum inhibitory concentration (MIC) of <1 μg/mL. In each case, an increase in daptomycin MICs was observed for subsequent isolates. Daptomycin-resistant MRSA strains were isolated from patients A and C, with MIC values of 2 and 4 μg/mL, respectively.

For patient A, MRSA prosthetic valve endocarditis (PVE) was diagnosed based on a positive result from blood culture and transthoracic echocardiography. Daptomycin was chosen for initial MRSA treatment because the patient was suffering from heart failure and acute renal failure. The MRSA strain SAZ1 was isolated at the beginning of daptomycin treatment. Although their condition subsequently improved after daptomycin treatment, this patient suffered from breakthrough MRSA bacteremia. After more than 20 days of daptomycin exposure, SAZ7 was isolated with a daptomycin MIC of 2 μg/mL. The blood culture was negative after the addition of rifampin.

Patients B and C were diagnosed as native valve endocarditis at other hospitals and were transferred to our hospitals, and both had paravalvular abscess detected by transthoracic echocardiography. For patient B, the initial strain SAW1 was obtained during moxifloxacin plus amoxicillin clavulanate potassium therapy, and it displayed resistance to fluoroquinolones such as ciprofloxacin and levofloxacin. Then, the treatment was changed to vancomycin for MRSA and ceftriaxone was added for pneumonia; SAW2 was isolated during this combination therapy. Afterward, daptomycin was used as an alternative to vancomycin for the clearance of MRSA from the bloodstream; however, the strains SAW3, SAW4, and SAW5 were isolated during daptomycin treatment. Although they were susceptible to daptomycin using 1 μg/mL as the breakpoint, these 3 strains exhibited 2- or 4-fold increases in daptomycin MIC. Likewise, the blood culture of patient B was negative after the addition of rifampin.

For patient C, daptomycin was used as an alternative agent because MRSA clearance from the bloodstream failed with vancomycin and linezolid. The daptomycin-resistant strains SAC4A (daptomycin = 4 μg/mL) and SAC4B (daptomycin = 4 μg/mL) were isolated after daptomycin exposure. It is interesting to note that these 2 strains exhibited restored susceptibility to cefoxitin and were identified as MSSA in VITEK 2. Subsequently, amoxicillin clavulanate potassium was added for this patient, and the blood culture turned negative after this. After 3 months, this patient was admitted to our hospital again due to recurrent bacteremia. Three MRSA strains, SAC5A, SAC5B, and SAC5C, which were susceptible to daptomycin, were isolated from 2 sets of blood culture bottles on the same day. Methicillin-resistant S aureus bacteremia was cured by combination therapy comprising linezolid and rifampin.

Molecular Typing and Mutations

To confirm the clonal relationship among these MRSA strains, PFGE and WGS-based typing were performed. The MRSA strains from patients A, B, and C belonged to ST5-SCCmec II (2A), ST59-SCCmec IVa (2B), and ST4513-SCCmec IVa (2B), respectively. The PFGE results showed that all strains had the same pattern within each case.

Meanwhile, a minimum spanning tree was constructed based on the cgMLST allelic profiles, showing that the MRSA strains of the same sequence type (ST) clustered together with several allelic differences within each ST (Figure 2). For patient A, 1 mutated gene was identified in the SAZ7 strain, which was further confirmed as a mutation in mprF causing an amino acid substitution (S295A). For patient B, 5 core genome mutations were found in the subsequent strains, such as those in gyrB, clpP, czcO, SA_01866, and mprF. A deletion was identified in mprF (I348del) of SAW5. For patient C, mutations were found in mprF (S337L) in daptomycin-resistant strains SAC4A and SAC4B, but for SAC5A, SAC5B, and SAC5C strains, which were susceptible to daptomycin and maintained wild-type mprF genes, acquired mutations in aspS (T568M and A199S) and prs (T233I) were found.

Because cgMLST analysis would miss mutations in the accessory genomes of these MRSA strains, we also screened for mutations by comparing the subsequent strains and their predecessors using the BRESEQ pipeline. The mutations identified by cgMLST were confirmed by BRESEQ results, and additional mutations such as those in yycH, ecfA2_1, yycl, dehH1, and mgt were found as listed in Table 2. All predicted mutations were validated by PCR with Sanger sequencing, and the primers are shown in Supplementary Tables 2 and 3.

Population Analysis and Heteroresistance

To identify daptomycin heteroresistance in MRSA strains associated with increased daptomycin MICs, we performed population analysis using MRSA strain COL and SA268 for comparisons (Figure 3). Through comparisons with the SA268 curve, MRSA strains from patient A displayed a substantial shift to the right in their daptomycin population curves. The highest inhibitory concentration (2 µg/mL) found for these daptomycin-susceptible MRSA strains (SAZ1-6) was 16-fold higher than the highest noninhibitory concentration (0.125 µg/mL) expected for the resistant strain (SAZ7). The highest inhibitory concentration of daptomycin against SAZ7 was not determined because of the limitations in the concentrations that we used; however, a resistant subpopulation was also observed in this strain at a concentration of 2 and 4 µg/mL, because the daptomycin MIC for this strain was 2 µg/mL.

Likewise, sequential MRSA strains from patients B and C displayed right-shifted curves after comparing their curves with those of their predecessor strains and the SA268 strain. The daptomycin-susceptible strains SAW4, SAC5A, SAC5B, and SAC5C were associated with the highest inhibitory concentration at 2 µg/mL, which was 16-fold higher than their highest noninhibitory concentration (0.125 µg/mL). The daptomycin-resistant strains SAW5, SAC4A, and SAC4B also showed a population-wide variable response to daptomycin with the highest inhibitory concentration ≥8 µg/mL and the highest noninhibitory concentration equal to 0.5 µg/mL. Meanwhile, daptomycin heteroresistance was not observed in strains SAW1, SAW2, SAW3, SAC1, SAC2, and SAC3, which had similar curves compared with that of the SA268 strain.

Fitness

To evaluate the fitness cost associated with the daptomycin-resistant or daptomycin-heteroresistant strains, we measured the growth rates of these strains and compared it to those of their predecessor strains (Figure 4). For patient A, the daptomycin-resistant strain SAZ7 exhibited a significant reduction in maximum relative growth rate (1.000 ± 0.027 vs 0.8513 ± 0.007; P < .0001), whereas the other serial MRSA strains with increased daptomycin MICs showed no significant adaptive cost, compared with fitness of the SAZ1 strain.

In contrast, the predecessor strains SAW1 and SAC1 had relatively low growth rates compared with those of the subsequent MRSA strains. For patient B, no significant change in growth rate was observed among SAW3, SAW4, and SAW5, which were associated with 2- or 4-fold increases in daptomycin MICs. For patient C, the daptomycin-resistant strain SAC4A exhibited no significant reduction in growth rate, whereas the other resistant strain SAC4B had a lower growth rate compared with that of SAC1. The daptomycin-susceptible strains from patient C, except for strain SAC5C, showed relatively higher growth rates when compared with that of strain SAC1. Strain SAC5C, which was isolated on the same day from blood cultures of SAC5B and SAC5A, showed a significantly reduced growth rate compared with that of SAC1, SAC5B, and SAC5A.

DISCUSSION

Daptomycin, which is rapidly bactericidal to susceptible strains, is recommended as an important alternative to vancomycin for the treatment of endocarditis; however, combination therapy and the development of resistance have been highlighted in complicated clinical or microbiological circumstances such as PVE cases and infections with vancomycin-nonsusceptible strains [21, 22]. Because there are increasing reports of daptomycin resistance among clinical MRSA isolates from other countries, we first analyzed the microbiological and genetic characteristics of daptomycin-resistant or nonsusceptible MRSA strains from endocarditis cases in China [9].

According to previous studies [23, 24], ST5 and ST239 MRSA are predominant in hospital settings in China, whereas ST59 is the most commonly isolated lineage from CA-MRSA infections. In our study, we showed the in-host evolution of MRSA strains from infective endocarditis patients, among which 1 serial strain belonged to ST5, whereas strains from the other 2 patients belonged to CC59, with 1 allelic difference. Antimicrobial susceptibility testing results showed that these MRSA strains developed resistance rapidly after exposure to antimicrobials including daptomycin. This potential for resistance based on evolution in these MRSA lineages might explain their predominance in both hospital and community environments.

The use of daptomycin for the treatment of infections caused by MRSA has resulted in the evolution of daptomycin-nonsusceptible MRSA strains, which are usually associated with multiple phenotypic and genotypic changes [25]. Such phenotypic changes in daptomycin-nonsusceptible strains include enhanced positive cell surface charges, increased cell wall thickness, and altered cell membrane fluidity due to changes in fatty acid composition. These phenotypic changes can be explained by 1 or multiple gene mutations such as those of mprF, dltABCD, and walK [7]. As expected, different mutations were detected in mprF genes of daptomycin-resistant MRSA strains in our study. We reviewed the studies on mprF mutations and found that the A883C variant of mprF has not been reported previously. It is interesting to note that MRSA SAW5 with another novel mutation, namely, 1042_1044delATC in mprF, was still susceptible to daptomycin according to the CLSI breakpoint, but it was associated with a 4-fold increase in daptomycin MICs. mprF mutation is the most common gene implicated in daptomycin resistance and appears to occur early under daptomycin exposure [7]. In our case, the emergence mprF mutations occurred after more than 1 week of daptomycin use. Moreover, several mutations in other genes were detected before mprF mutated, which maybe associated with the selection of other antimicrobial agents such as fluoroquinolones and β-lactams. The contribution of other mutations besides mprF to daptomycin resistance need to be further studied. Mutations in other daptomycin resistance-related genes such as yycH and yycI were also identified in our strains [26]. A previous study also showed that the yycHI mutation could lead to the reduced activation of WalRK, which was described as a 2-component regulator associated with daptomycin resistance [27].

Heteroresistance describes a phenomenon wherein subpopulations of seemingly isogenic bacteria exhibit a range of susceptibilities to a particular antibiotic, and daptomycin heteroresistance in Staphylococcus has been described previously [2]. Cui et al [20] reported a laboratory MRSA strain with dual heteroresistance to daptomycin and vancomycin caused by an rpoB (A621E) mutation. In our 3 cases, before the emergence of daptomycin-resistant strains, MRSA strains associated with increasing daptomycin MICs were isolated continuously during daptomycin treatment. We assumed that a daptomycin-resistant subpopulation might exist in these daptomycin-susceptible MRSA populations, which ultimately led to treatment failure. Despite the lack of a standard method and criteria to define daptomycin-heterogeneous resistance in MRSA, we performed PAP for our strains to evaluate the heteroresistant subpopulation, because this method is considered the gold standard for detecting this phenomenon by CFU counts [2]. As expected, daptomycin heteroresistance was observed in MRSA strains from all 3 cases. The initial populations from patients B and C were negative for daptomycin heteroresistance based on PAP results, but heteroresistance phenotype emerged in the sequential strains with increased daptomycin MICs, indicating that intrinsic heteroresistance could be induced after initial exposure to daptomycin [2]. It is interesting to note that the first strain SAZ1, from patient A, had a significant resistant subpopulation, even without daptomycin exposure. Previous studies demonstrated that the development of daptomycin heteroresistance could be associated with other antimicrobial agents such as vancomycin [9]. Thus, we could not rule out the possibility that this initial strain from patient A, which belonged to ST5, had been evolving in the healthcare environment for a long time. Our results indicated that daptomycin heteroresistance could emerge with different phenotypes among different MRSA lineages with varied genetic backgrounds. Some novel mutated genes were also identified in our daptomycin-heteroresistant strains, such as mgt, aspS, and prsA. Further studies should be conducted to reveal the mechanism underlying daptomycin heteroresistance in these clinical MRSA isolates.

Fitness cost has been considered an important factor associated with the emergence of antibiotic resistance in a bacterial population [28]. Fitness cost in daptomycin-resistant MRSA strains isolated from the bloodstream was reported by Li et al [29]; however, these resistant mutants were obtained by daptomycin selection in vitro. In our study, we evaluated the fitness cost of these daptomycin-resistant or daptomycin-heteroresistant MRSA strains from clinical patients. According to growth rate results, we found that the daptomycin-heteroresistant strains suffer little or no fitness cost, based on comparisons with their predecessor strains. However, we also found that it was difficult to evaluate the fitness cost in clinical strains, because the predecessor strain usually already exhibited some fitness cost due to exposure to other antimicrobials.

Based on the development of resistance during prolonged MRSA infection, it is worth mentioning that the rational use of antimicrobial agents should be emphasized for complicated MRSA infections such as endocarditis. For patient A, combination rifampin treatment was delayed for several reasons including concerns over kidney function, delayed consultation with an infectious disease specialist, and insufficient clinical evaluation. In addition, the gentamycin was unavailable at our hospital at the time. Although the evidence is poor, adding rifampin for the treatment of staphylococcal PVE is recommended by the guidelines, although treatment might be associated with microbial resistance, hepatotoxicity, and drug interactions [21, 22].

CONCLUSIONS

To conclude, we demonstrated in-host adaptation and the development of daptomycin resistance in MRSA strains from 3 patients with infective endocarditis. Also of note, we confirmed that daptomycin heteroresistance and resistance could emerge rapidly in MRSA strains form different lineages such as ST5 and ST59 after daptomycin exposure. Our results also showed that WGS-based typing such as cgMLST could be a useful tool to detect mutations associated with resistance, but that the mechanism of daptomycin resistance could be different and complicated for MRSA strains with varied genetic backgrounds. Further study to fully understand the mechanism associated with daptomycin resistance in MRSA is urgently needed to establish efficient antimicrobial treatment strategies for difficult-to-treat bacterial infections such as endocarditis.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

References

Notes

Financial support. This work was funded by the Key Research and Development Programme of Zhejiang (Grant 2015C03046).

Supplement sponsorship. This supplement was supported by MSD.

Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest.

Author notes