Next Article in Journal
Hospital Antibiotic Consumption before and during the COVID-19 Pandemic in Hungary
Previous Article in Journal
Surgical Antibiotic Prophylaxis: A Proposal for a Global Evidence-Based Bundle
Previous Article in Special Issue
Evaluation of Enterococcal Probiotic Usage and Review of Potential Health Benefits, Safety, and Risk of Antibiotic-Resistant Strain Emergence
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Antibiotics
Volume 13
Issue 1
10.3390/antibiotics13010101
antibiotics-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Linezolid Resistance Genes and Mutations among Linezolid-Susceptible Enterococcus spp.—A Loose Cannon?

by
Jennifer K. Bender
1,*,
Carola Fleige
1,
Finn Funk
1,
Clara Moretó-Castellsagué
1,2,
Martin A. Fischer
1 and
Guido Werner
1
1
Nosocomial Pathogens and Antibiotic Resistances Unit, Department of Infectious Diseases, Robert Koch Institute, 38855 Wernigerode, Germany
2
Department of Microbiology and Parasitology, University Hospital of Bellvitge, 08907 Barcelona, Spain
*
Author to whom correspondence should be addressed.
Antibiotics 2024, 13(1), 101; https://doi.org/10.3390/antibiotics13010101
Submission received: 30 November 2023 / Revised: 16 January 2024 / Accepted: 17 January 2024 / Published: 19 January 2024
(This article belongs to the Special Issue Antibiotic Treatment and Antibiotic Resistance of Enterococcus Infections)
Browse Figure
Versions Notes

Abstract

:
The National Reference Centre for Enterococci receives an increasing number of linezolid-resistant Enterococcus isolates. Linezolid (LIN) resistance is mediated by G2576T 23S rDNA gene mutations and/or acquisition of resistance genes (cfr, optrA, poxtA). There are anecdotal reports that those resistance traits may be present in phenotypically linezolid-susceptible isolates. We aimed to determine the prevalence of LIN resistance genes and mutations in enterococci with a LIN MIC of 4 mg/L in broth microdilution (EUCAST = susceptible) isolated from German hospital patients 2019–2021. LIN MICs were additionally determined by ETEST® and VITEK2. Selected strains were subjected to LIN selective pressure and growth was monitored with increasing antibiotic concentrations. We received 195 isolates (LIN MIC = 4 mg/L). In total, 78/195 (40%) isolates contained either a putative resistance gene, the G2576T mutation, or a combination thereof. Very major error was high for broth microdilution. The ability to predict phenotypic resistance from genotypic profile was highest for G2576T-mediated resistance. Selection experiments revealed that, in particular, E. faecium isolates with resistance gene mutations or poxtA rapidly adapt to MICs above the clinical breakpoint. In conclusion, LIN resistance genes and mutations can be observed in phenotypically linezolid-susceptible enterococci. Those isolates may rapidly develop resistance under LIN selective pressure potentially leading to treatment failure.

1. Introduction

Enterococci are gut commensal organisms that might become life-threatening bacterial pathogens in immunocompromised individuals, mainly in hospital settings. Infections are difficult to treat due to their inherent insusceptibility to various antimicrobial substances. Since the year 2000, linezolid (LIN), a synthetic antibiotic, has been approved as a last-resort therapeutic option to treat vancomycin-resistant Enterococcus spp. (VRE). Although an increase in the prevalence of linezolid-resistant enterococci (LRE) has not yet been reported on a global scale [1,2,3,4], recent data from the German Antimicrobial Resistance Surveillance indicate an increase of LRE among invasive isolates (0.6% in 2019 to 1.2% in 2021; https://ars.rki.de/, accessed on 26 November 2023). Moreover, and although biased, the National Reference Centre for Staphylococci and Enterococci (NRC) at the Robert Koch Institute has identified an increasing number of LRE sent to the NRC for confirmatory resistance diagnosis [5].
In enterococci, LIN phenotypic resistance is either mediated by the expression of mobile resistance genes poxtA or optrA, both encoding ribosomal protection proteins, by the methyltransferase Cfr and variants thereof, or the result of chromosomal alterations mainly at position 2576 in the V domain of the 23S ribosomal RNA (reviewed by [6,7]).
In order to determine LIN resistance, antimicrobial susceptibility testing (AST) is performed by broth microdilution (BMD) (reference method), gradient strip tests (such as ETEST®), automated systems (such as VITEK2), or disk diffusion assays. According to the European Committee on Antimicrobial Susceptibility Testing (EUCAST), LIN resistance in Enterococcus spp. is defined as a minimal inhibitory concentration (MIC) of >4 mg/L. It has previously been demonstrated that enterococci harboring the resistance gene cfr/cfr(B), optrA, or G2576T mutations in low allelic frequencies would present with LIN MICs of 4 mg/L (or even below), thus being inadequately diagnosed as linezolid-susceptible isolates (EUCAST breakpoints) [8,9,10,11].
From 2019 to 2021, the NRC noted quite a substantial number of isolates demonstrating an MIC just below the EUCAST-defined clinical breakpoint. Thus, we compiled a strain collection of phenotypically linezolid-susceptible (EUCAST) or linezolid-intermediate (according to the Clinical and Laboratory Standards Institute (CLSI)) Enterococcus spp. (LSE) isolates (all MIC = 4 mg/L) as determined by the reference method BMD and aimed at: (i) determining the prevalence of LIN resistance genes and mutations in enterococci with a LIN MIC of 4 mg/L, (ii) comparing the performance of the reference method BMD with the results obtained from VITEK2, ETEST®, and CHROMagarTM LIN-R; (iii) analyzing the genotype–phenotype correlation between the presence of LIN resistance genes/mutations and LIN resistance, and (iv) evaluating the potential of “silent” resistance genes and/or G2576T ribosomal RNA gene mutations in supporting rapid adaption to LIN under antibiotic selective pressure.
Our study revealed a high proportion of LIN resistance genes and/or chromosomal mutations in phenotypically susceptible (according to EUCAST) or intermediate (CLSI) isolates that could potentially result in resistance development under LIN selective pressure. We further highlight the diagnostic dilemma for LIN AST by various methods and their partial inability to correctly assess phenotypic resistance in the presence of a potential underlying LIN resistance mechanism.

2. Results

2.1. Distribution of LIN Resistance Genes/Mutations among Phenotypically Susceptible Enterococci

From 2019 to 2021, we received a total of 196 Enterococcus spp. isolates displaying a linezolid MIC of 4 mg/L when applying the reference method BMD. Isolates were preserved as cryo-cultures shortly after receipt. One isolate could not be re-cultivated from stock and thus was excluded from further analyses.
Of the 195 isolates investigated, species determination confirmed 21 (11%) E. faecalis and 174/195 (89%) E. faecium. No putative resistance mechanism was detected for 4/21 (19%) E. faecalis and 113/174 (65%) E. faecium isolates (Table 1). Altogether, 78/195 (40%) of all isolates either contained a putative resistance gene or the G2576T 23S rRNA gene mutation (Table 1). Within E. faecalis, one isolate exhibited a 23S rDNA mutation and 16/21 (76%) harbored the gene for the ribosomal protection protein optrA. For E. faecium, a high proportion [49/174 (28%)] displayed G2576T chromosomal mutations, and, to a lesser extent, a single mobile resistance determinant, such as poxtA [7/174 (4%)] or a combination of resistance genes and/or G2576T mutations (Table 1). The gene for the methyltransferase Cfr was detected in two E. faecium isolates only (Table 1). It must be noted that we utilized a previously established multiplex-PCR to screen for cfr, optrA, and poxtA and that some variants thereof might not be identified due to reduced binding of the targeting primers. Thus, the total amount of transferable resistance genes may be underestimated.

2.2. Comparison of BMD with Additional AST Methods

2.2.1. Performance of AST Methods on the Set of Isolates with a LIN MIC of 4 mg/L

We carried out additional AST using ETEST® and VITEK2 and investigated the growth of all strains on CHROMagarTM LIN-R plates after 24 h and 48 h of incubation. The MIC results of the different AST methods are displayed in Table 2. Compared to BMD, which is considered the reference method for (LIN) AST in enterococci and which resulted in 100% susceptible isolates (MIC = 4 mg/L) applying EUCAST breakpoints, ETEST® and VITEK2 determined 6.7% and 29%, respectively, as linezolid-resistant (Table 2). Incubation on CHROMagarTM LIN-R resulted in 38% and 42% resistant isolates when evaluated after 24 h and 48 h, respectively, at 37 °C (Table 2).
To estimate the number of false-resistant (major error, ME) and false-susceptible (very major error, VME) isolates as determined by the different AST methods, we made the following assumption: isolates showing none of the known and tested LIN resistance mechanisms (N = 117) were considered genotypically/phenotypically susceptible, whereas those exhibiting either a G2576T 23S rRNA gene mutation, a mobile resistance gene cfr/optrA/poxtA, or a combination thereof (N = 78) were considered genotypically/phenotypically resistant.
Due to our study design (only isolates with LIN MIC = 4 mg/L in BMD), ME (false-resistant = genotypically susceptible, but phenotypically resistant) could not be determined for BMD (n.a., not applicable, Table 3). For BMD, VME was 40%, corresponding to isolates that were diagnosed as susceptible but with a genotypic profile capable of mediating phenotypic resistance.
Compared to BMD, the percentage of false-susceptible (= VME) isolates gradually declined from 33% for ETEST® and 13% for VITEK2 to 2.1% for CHROMagarTM LIN-R (48 h of incubation) (Table 3). In contrast, false-resistant (= ME) isolates increased from 0% using ETEST® and 2.1% by VITEK2 to 3.1% and 4.1% when streaking on CHROMagarTM LIN-R agar plates (24 h and 48 h incubation, respectively) (Table 3).

2.2.2. Predictive Ability of Different AST Methods for Genotype-Phenotype Correlation

As outlined above, different AST methods varied with respect to their overall performance. Thus, we further investigated whether these methods exhibit tendencies for the prediction of phenotypic resistance based on their specific genotypic profile. We selected all isolates that either harbored a single resistance gene or the 23S rRNA gene mutation G2576T (n = 75) to estimate the positive predictive value (PPV). Isolates with multiple resistance genes and/or chromosomal mutations were excluded due to the inability to differentiate their relative contribution to phenotypic resistance.
The different AST methods varied greatly in their predictive ability according to the underlying potential resistance mechanism (Table 4). Among the LIN resistance genes, ETEST® and VITEK2 showed the highest PPV for optrA-positive isolates, 58.8% (10/17) and 11.8% (2/17), respectively, whereas CHROMagarTM LIN-R majorly detected poxtA-positive isolates (100%, 7/10) (Table 4).
ETEST® failed to predict phenotypic resistance among poxtA-positive strains (0/7, 0%). With only one isolate available, cfr-based genotypical resistance demonstrated phenotypic resistant results, and thus a high PPV (100%), for VITEK2 and CHROMagarTM LIN-R, the latter only after 48 h of incubation. With respect to the G2576T chromosomal mutations, ETEST® and VITEK2 demonstrated the highest PPV as compared to gene-mediated resistance (except for cfr). For example, VITEK® correctly assessed 72% of G2576T-haboring isolates as resistant as compared to 58.8% and 42.9% for optrA and poxtA, respectively. In contrast, CHROMagarTM LIN-R generally demonstrated a lower predictive value for the G2576T 23S rRNA gene mutations, especially at 24 h of incubation, when compared to gene-based LIN resistance (86% G2576T vs. 94.1% and 100% for optrA- and poxtA-positive isolates) (Table 4).

2.3. LIN Selection Experiments

We hypothesized, that isolates with a false-susceptible result, hence being genotypically resistant, could potentially adapt to linezolid selective pressure more rapidly. To investigate this, we carried out in vitro selection experiments by gradually increasing LIN concentrations. We selected seven E. faecium and five E. faecalis with (n = 10) and w/o (n = 2) G2576T 23S rRNA gene mutations or mobile resistance determinants (Supplementary Table S1). As we have not detected cfr- or poxtA-positive E. faecalis in our study (Table 1), no such isolate could be included. We also excluded combinations of chromosomal mutations and resistance genes, given the generally low number of isolates identified in our study and assuming that frequent recombination of mutated 23S rDNA alleles could shadow the impact of resistant gene expression. After six steps of LIN challenge from sub-inhibitory concentrations of 0.5 mg/L to 16 mg/L, 3/10 of the isolates containing a putative resistance trait had reached MICs above the clinical breakpoint (EUCAST > 4 mg/L), and one isolate (E. faecalis UW21555 optrA-positive) presented with 4 mg/L and 8 mg/L in the two independent selection experiments (Figure 1 and Supplementary Table S1).
The two strains without a resistance trait as identified in our study (E. faecium UW19609 and E. faecalis UW22498) and the susceptible negative control E. faecalis ATCC29212 were not able to adapt and remained with MIC values below the threshold at the end of the selection experiments (Figure 1). With the sole exception of one optrA-positive E. faecalis (UW21555, final LIN MIC 4 mg/L and 8 mg/L), no other selected E. faecalis was able to adapt to increasing concentrations of LIN (Figure 1). In contrast, the two poxtA-positive E. faecium (UW19892, UW21431) and an E. faecium harboring the G2576T 23S rRNA gene mutation (UW20036) reached MICs of 16 mg/L (Figure 1, Supplementary Table S1). E. faecium with cfr, optrA, or a combination of optrA and poxtA were not able to adapt to LIN MIC above 4 mg/L (Figure 1, Supplementary Table S1). In order to rule out that this lack of adaptation is due to the loss of the respective resistance determinant, we examined the presence of the initial resistance trait after LIN selective pressure for all isolates. In one instance, a previously optrA-positive E. faecium (UW22166, Figure 1, Supplementary Table S1) had lost the resistance gene during the course of both independent experiments. Mutations in the 23S rRNA gene did not revert, as assessed by NheI-restriction experiments, or were detected in previously G2576T mutation-negative isolates.

3. Discussion

Linezolid resistance in vancomycin-susceptible, and more importantly, in vancomycin-resistant enterococci, is considered a threat to public health due to limited available treatment options. Therefore, it is important to correctly asses the resistance potential of Enterococcus spp. in diagnostic laboratories. We herein examined a set of phenotypically LIN-susceptible isolates (ECUAST susceptible = LIN MIC = 4 mg/L) as determined by the reference method BMD and observed that: (i) 40% of those isolates harbored potential resistance genes or G2576T ribosomal gene mutations, (ii) different AST methods displayed varying results concerning the number of false-susceptible and false-resistant isolates detected, (iii) these methods varied in their ability to predict phenotypic resistances based on the isolates’ genotypical profile, and (iv) “silent” resistance genes could potentially trigger rapid adaptation to LIN under selective pressure.
BMD is considered the reference method for AST of rapidly growing aerobic bacteria [12]; however, assessing resistance to LIN in BMD is a challenging task due to the phenomenon of “trailing growth”. This is defined as the fading of growth over two to three wells [12]. Hence, manual read-out is not only influenced by the growth phenomenon of Enterococcus spp. per se, but also by individual visual inspection of the wells, resulting in subjective classification of isolates as resistant or susceptible. As a limitation of our study, our data set only constituted isolates with a LIN MIC of 4 mg/L as determined by BMD and thus, might be prone to a certain degree of measurement bias.
The performance and impact of different AST methods to identify LRE have been assessed previously [13,14,15,16,17]. We found a high proportion (40%) of VME (false-susceptible) for BMD and gradually declining VMEs of 33% for ETEST®, 13% for VITEK2, and <5% for CHROMagarTM LIN-R plates. MEs (false-resistant) were assessed as 0% for ETEST®, 2.1% for VITEK2, and 4.1% for CHROMagarTM LIN-R (48 h of incubation). A study by Dejoies et al. [13] yielded comparable estimates when applying EUCAST breakpoints with rates of 2.1% ME for VITEK2 and 2.1% ME for ETEST®; but in contrast, their VME results showed higher rates of 38.5% for VITEK2 and 23.1% for ETEST®. It must be noted that the error rates reported by our colleagues were obtained by using BMD results (susceptible or resistant) as the reference group without considering potential genotypic resistance profiles. This is an important piece of information, implying that reported VME and ME rates for the various methods and subsequent comparisons of the error calculations must be taken with caution, given the different clinical breakpoints applied (EUCAST or CLSI) and the definition of the reference group used.
We categorized our isolates according to EUCAST guidelines for which a LIN MIC = 4 mg/L is considered susceptible, an MIC value that is considered “intermediate” when applying CLSI criteria. Under our study assumption of a genotype–phenotype correlation, the error calculation of VME and ME was substantially different when using CLSI breakpoints, resulting in no VME for BMD (compared to 40% for EUCAST breakpoints), 21% for ETEST® (33% EUCAST), and 7.2% for VITEK2 (13% EUCAST). It seems reasonable to consider CLSI prior to EUCAST for determination of LIN resistance and to flag isolates with an MIC of 4 mg/L, hence drawing the attention of clinicians to a possibly untreatable isolate. With the potential for LIN resistance development under LIN selective pressure for those isolates harboring potential resistance determinants, as seen in our study, the discussion must be continued following further investigations.
In our study, we made a strong assumption that genotypically resistant isolates would correlate with a resistant phenotype. To further investigate whether AST methods differ in their potential to detect the respective underlying resistance trait, we assessed their predictive ability for phenotypic resistance based on the genotypic profile. We observed variability to obtain a resistant phenotype; for example, ETEST® and VITEK2 displayed a much higher PPV for optrA-positive isolates compared to poxtA. Comparable results were obtained by Dejoies et al., in which the majority of optrA-positive isolates reached MIC values above the clinical breakpoint (EUCAST and CLSI) compared to isolates harboring poxtA [13]. Furthermore, the prediction of mobile resistance gene determinants by ETEST® and VITEK2 was outperformed by the prediction of phenotypic resistance mediated by G2576T 23S rRNA gene mutation, reaching PPVs as high as 72% (VITEK2). This is somewhat expected as it has been demonstrated that mutation of a single 23R rDNA allele in E. faecium is sufficient to elevate MICs above the EUCAST clinical breakpoint [18]. In contrast to VITEK2 and ETEST®, CHROMagarTM LIN-R demonstrated higher PPVs for gene-containing isolates (optrA and poxtA) compared to G2576T chromosomal mutations. Considering the small sample size for cfr- and poxtA-positive isolates as a limitation for generalization, this is still an interesting observation and requires more in-depth investigations.
We would like to emphasize that the total number of gene-positive isolates may be underestimated and thus affect our estimates for VME, ME, and PPV due to the possible reduced binding of our screening primers to target gene variants. It should also be mentioned that our genotypic assessment focuses on the most frequent 23S rDNA mutation and resistance-mediating transferable genes. Our analysis does not include less relevant 23S rDNA mutations or ribosomal protein alterations, which may result in reduced LIN susceptibility but not LIN resistance in enterococci [19,20,21,22]. This is a limitation of our study.
It is well known that mobile LIN resistance genes may fail to confer phenotypic resistance in some instances that are not well understood [8,23,24,25,26]. Thus, our VME rates are somewhat overestimated. Nevertheless, it is unclear how LIN silent resistance genes, here defined as gene presence but an absence of phenotypic resistance, will respond to LIN selective pressure under LIN therapy. We hypothesized that these silent resistance genes or mutations could potentially allow rapid adaption and resistance development under LIN selective pressure. Our selection experiments confirmed this assumption for E. faecium with a G2576T mutation or containing the poxtA gene and for one optrA-positive E. faecalis. However, in some instances, resistance development was not achieved. The results are difficult to interpret at this stage, as the mutation persisted or genes were consistently present (with one exception) over the course of the independent experiments. For E. faecalis it was demonstrated that two out of four allelic copies mutated could result in an MIC of 4 mg/L [9], and thus our observation of the G2576T-harboring E. faecalis that remained below the resistance threshold might be due to the low number of mutated rRNA alleles that did not increase during LIN selection in vitro. Of note, our results were obtained under laboratory conditions that can only approximate but do not reflect in vivo situations of prolonged LIN therapy. LIN treatment, which is prescribed in concentrations above the sub-inhibitory levels used in our selection experiments, was shown to be associated with an increased risk of developing LIN resistance in E. faecium [27]. Moreover, a hospital-wide increasing number of LRE were linked with increased LIN consumption [28]. Thus, failure to develop resistance in 7/10 of our selected isolates investigated might reflect the laboratory conditions and cannot be generalized to a LIN-treated human host.
Relying on phenotypic resistance determination is insufficient to provide the full picture. As seen in our study, 40% of all isolates contained a genotypic resistance marker but were susceptible in BMD (EUCAST breakpoints). It is widely accepted that gene-associated resistance development is highly dependent on the promotor structure, plasmid copy number, or regulation of gene expression [29,30,31], which was not further examined in our study. Silent resistance genes are not unique to enterococci and have been widely reported across the bacteria phylum (reviewed by [32]). However, they can become an unprecedented threat due to the silent transmission of resistance determinants in the hospital environment, as reported for blaOXA-23-carrying, but imipenem-susceptible Acinetobacter baumanii [33]. Furthermore, the transfer of silent genes into a new host may activate the respective resistance trait, as seen for a silent aadA gene being expressed after transfer from E. coli to Hafnia alvei [34]. Imperceptible spreading of LIN silent resistance genes combined with insufficient phenotypic assessment may thus lead to the prescription of non-functional therapeutics, exposing patients to an unnecessary risk for a severe disease outcome. We, therefore, want to emphasize the importance of a combined assessment of genotypic and phenotypic resistance, especially for invasive isolates presenting with a LIN MIC of 4 mg/L. Obtaining the most accurate laboratory results is a prerequisite for the prevention of resistance development or imperceptible resistance gene transmission.

4. Materials and Methods

Study material. We included all Enterococcus spp. isolates that were sent to the NRC between 2019 and 2021 and demonstrated a linezolid MIC of 4 mg/L using the reference method BMD.
Antimicrobial susceptibility testing. Isolates were cultivated on blood agar followed by BMD using an in-house and accredited procedure and by applying EUCAST clinical breakpoints for resistance determination (EUCAST v11). Here, LIN resistance is defined as MIC > 4 mg/L, whereas an MIC ≤ 4 mg/L is considered susceptible. In comparison, CLSI clinical breakpoints for LIN are as follows: MIC ≤ 2 mg/L “susceptible”, MIC = 4 mg/L “intermediate”, and MIC ≥ 8 mg/L “resistant”. Linezolid insusceptibility was additionally investigated by ETEST® and VITEK2 (bioMérieux, Nuertingen, Germany) according to the manufacturer’s instructions and by cultivation on CHROMagarTM LIN-R (CHROMagarTM, Paris, France) for 24 h and 48 h.
Determination of putative resistance mechanism(s). All isolates with an MIC of 4 mg/L in BMD were subjected to DNA extraction and multiplex-PCR in order to screen for cfr, cfr(B), optrA, and poxtA resistance determinants as described previously [35]. Furthermore, G2576T 23S rRNA gene mutations were determined by an amplification-restriction-based procedure [36].
Data analysis. Results were collectively analyzed using R Studio V4.1.2 and MS Excel Office 2019. For calculating very major error (VME, false-susceptible isolates) and major error (ME, false-resistant isolates), we made the following assumption: the presence of any of the three resistance genes cfr/optrA/poxtA and/or the detection of a G2576T mutations could potentially result in phenotypic resistance; thus, these isolates were considered resistant. VME and ME for the different AST methods are expressed as percentage of false-susceptible and false-resistant isolates of all isolates investigated.
Linezolid selection experiments. We included seven E. faecium and five E. faecalis either harboring a G2576T ribosomal mutation OR a putative resistance gene (cfr, optrA, poxtA) OR none of the above (negative control), and two positive controls with LIN MICs > 16 mg/L from the NRC strain collection (one E. faecium, one E. faecalis), both positive for the G2576T mutation and negative for transferable resistance genes as determined by our inhouse multiplex-PCR (Supplementary Table S1). As for our routine BMD, strains were cultivated first for 3 h in nutrient broth and diluted 1:50 in 0.85% sodium chloride thereafter. Of this solution, 20 µL was used to inoculate a 96-well plate, each well containing 180 µL BHI (1st experiment) or MH broth (2nd experiment) and the respective LIN concentration (see in the following). At step 1, we used sub-inhibitory concentrations of LIN = 0.5 mg/L and LIN = 1 mg/L. Technical duplicates were used throughout. Plates were incubated overnight at 37 °C and growth was visually inspected after 24 h. If growth was detected at the highest concentration (step 1 = 1 mg/L), 20 µL of one of the duplicated cells was transferred to a new 96-well selective plate (step 2). At step 2, we increased the selective pressure by 2-fold but also included the highest concentration from step 1 for recovery (concentrations at step 2 equal 1 mg/L and 2 mg/L, respectively). A growth control without LIN selective pressure was carried along. This procedure was followed for 6 days until LIN = 16 mg/L was reached as the highest concentration. Strains that stopped growing at a certain concentration were stored at −20 °C in 50% glycerol using the previous concentration/well where growth was detected. To assess the final LIN MIC reached, a standard BMD assay was carried out with LIN concentrations ranging from 0.125–32 mg/L. The entire selection experiment was repeated one more time.

5. Conclusions

Linezolid resistance genes and mutations can be observed in phenotypically linezolid-susceptible enterococcal isolates suggesting a certain degree of under-representation of their prevalence in clinical isolates. Different antimicrobial susceptibility testing methods vary in their ability to correctly identify putative resistant isolates based on their genotypic resistance profile. In our data set, the reference method, broth microdilution, revealed a substantial number of Enterococcus spp. isolates with a LIN MIC of 4 mg/L (= susceptible when applying EUCAST breakpoints) despite the presence of LIN resistance mutations and/or genes. According to CLSI, isolates with a LIN MIC of 4 mg/L are categorized as “intermediate”, which may better indicate a high degree of uncertainty in the classification of these enterococcal isolates. The widespread prevalence of 23S rDNA linezolid resistance mutations and/or possession of LIN mobile resistance determinants in phenotypically susceptible enterococcal isolates may promote the imperceptible spread and trigger a rapid development towards resistance under LIN selective pressure potentially leading to treatment failure.

Supplementary Materials

The following are available online at https://www.mdpi.com/article/10.3390/antibiotics13010101/s1, Table S1: E. faecium (n = 7) and E. faecalis (n = 5) isolates for LIN selective pressure experiments from the study set of LIN MIC = 4 mg/L (in BMD) strains collected by the NRC, 2019–2021.

Author Contributions

Conceptualization, J.K.B. and G.W.; methodology, J.K.B., M.A.F., C.F., F.F. and C.M.-C.; software, J.K.B., F.F., C.F. and C.M.-C.; validation, C.F., F.F. and C.M.-C.; formal analysis, J.K.B., C.F., F.F. and C.M.-C.; investigation, J.K.B., M.A.F. and G.W.; resources, G.W.; data curation, J.K.B., M.A.F. and C.M.-C.; writing—original draft preparation, J.K.B.; writing—review and editing, J.K.B., F.F., C.M.-C. and G.W.; visualization, J.K.B.; supervision, J.K.B., M.A.F. and G.W.; project administration, J.K.B. and M.A.F.; funding acquisition, G.W. All authors have read and agreed to the published version of the manuscript.

Funding

The work of the NRC was funded by a grant (FKZ 1369-234) from the Federal Ministry of Health, Germany (relevant period, 2018–2022).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data presented in this study are available on request from the corresponding author.

Acknowledgments

We would like to thank the excellent technical support given by Christine von Bornstädt, Birgit Pasemann, Petra Vilbrandt, and Annette Weller for routine laboratory analyses at the NRC.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Chen, Y.; Ji, J.; Ying, C.; Liu, Z.; Yang, Q.; Kong, H.; Xiao, Y.; Blood Bacterial Resistant Investigation Collaborative System (BRICS) Study Group. Blood bacterial resistant investigation collaborative system (BRICS) report: A national surveillance in China from 2014 to 2019. Antimicrob. Resist. Infect. Control 2022, 11, 17. [Google Scholar] [CrossRef] [PubMed]
  2. Heininger, A.; Zimmermann, S.; Bootsveld, C.; Boutin, S.; Nurjadi, D. Low prevalence of combined linezolid- and vancomycin-resistant Enterococcus faecium from hospital admission screening in an endemic region in Germany. J. Glob. Antimicrob. Resist. 2020, 22, 646–650. [Google Scholar] [CrossRef] [PubMed]
  3. Markwart, R.; Willrich, N.; Eckmanns, T.; Werner, G.; Ayobami, O. Low Proportion of Linezolid and Daptomycin Resistance Among Bloodborne Vancomycin-Resistant Enterococcus faecium and Methicillin-Resistant Staphylococcus aureus Infections in Europe. Front. Microbiol. 2021, 12, 664199. [Google Scholar] [CrossRef]
  4. Coombs, G.W.; Daley, D.A.; Mowlaboccus, S.; Lee, Y.T.; Pang, S.; Australian Group on Antimicrobial Resistance. Australian Group on Antimicrobial Resistance (AGAR) Australian Enterococcal Sepsis Outcome Programme (AESOP) Annual Report 2018. Commun. Dis. Intell. (2018) 2020, 44. [Google Scholar] [CrossRef] [PubMed]
  5. Fischer, M.A.; Bender, J.K.; Kriebel, N.; Weber, R.E.; Wohlfarth, E.; Maechler, F.; Noll, I.; Abu Sin, M.; Werner, G. Eigenschaften, Häufigkeit und Verbreitung von Vancomycin-resistenten Enterokokken in Deutschland—Update. Epidemiol. Bull. 2023, 28, 14. [Google Scholar] [CrossRef]
  6. Bender, J.K.; Cattoir, V.; Hegstad, K.; Sadowy, E.; Coque, T.M.; Westh, H.; Hammerum, A.M.; Schaffer, K.; Burns, K.; Murchan, S.; et al. Update on prevalence and mechanisms of resistance to linezolid, tigecycline and daptomycin in enterococci in Europe: Towards a common nomenclature. Drug Resist. Updates 2018, 40, 25–39. [Google Scholar] [CrossRef] [PubMed]
  7. Schwarz, S.; Zhang, W.; Du, X.D.; Kruger, H.; Fessler, A.T.; Ma, S.; Zhu, Y.; Wu, C.; Shen, J.; Wang, Y. Mobile Oxazolidinone Resistance Genes in Gram-Positive and Gram-Negative Bacteria. Clin. Microbiol. Rev. 2021, 34, e0018820. [Google Scholar] [CrossRef]
  8. Bender, J.K.; Fleige, C.; Klare, I.; Fiedler, S.; Mischnik, A.; Mutters, N.T.; Dingle, K.E.; Werner, G. Detection of a cfr(B) Variant in German Enterococcus faecium Clinical Isolates and the Impact on Linezolid Resistance in Enterococcus spp. PLoS ONE 2016, 11, e0167042. [Google Scholar] [CrossRef]
  9. Lobritz, M.; Hutton-Thomas, R.; Marshall, S.; Rice, L.B. Recombination proficiency influences frequency and locus of mutational resistance to linezolid in Enterococcus faecalis. Antimicrob. Agents Chemother. 2003, 47, 3318–3320. [Google Scholar] [CrossRef]
  10. Raad, I.I.; Hanna, H.A.; Hachem, R.Y.; Dvorak, T.; Arbuckle, R.B.; Chaiban, G.; Rice, L.B. Clinical-use-associated decrease in susceptibility of vancomycin-resistant Enterococcus faecium to linezolid: A comparison with quinupristin-dalfopristin. Antimicrob. Agents Chemother. 2004, 48, 3583–3585. [Google Scholar] [CrossRef]
  11. Rodriguez-Lucas, C.; Fernandez, J.; Vazquez, X.; de Toro, M.; Ladero, V.; Fuster, C.; Rodicio, R.; Rodicio, M.R. Detection of the optrA Gene Among Polyclonal Linezolid-Susceptible Isolates of Enterococcus faecalis Recovered from Community Patients. Microb. Drug Resist. 2022, 28, 773–779. [Google Scholar] [CrossRef]
  12. European Committee on Antimicrobial Susceptibility Testing. EUCAST Reading Guide for Broth Microdilution. Available online: https://www.eucast.org/fileadmin/src/media/PDFs/EUCAST_files/Disk_test_documents/2022_manuals/Reading_guide_BMD_v_4.0_2022.pdf (accessed on 26 November 2023).
  13. Dejoies, L.; Boukthir, S.; Pean de Ponfilly, G.; Le Guen, R.; Zouari, A.; Potrel, S.; Collet, A.; Auger, G.; Jacquier, H.; Fihman, V.; et al. Performance of commercial methods for linezolid susceptibility testing of Enterococcus faecium and Enterococcus faecalis. J. Antimicrob. Chemother. 2020, 75, 2587–2593. [Google Scholar] [CrossRef] [PubMed]
  14. Jones, R.N.; Moet, G.J.; Woosley, L.N.; Sader, H.S.; Fritsche, T.R. Critical Evaluation of Linezolid Susceptibility Testing using two Automated Systems (Vitek, Vitek 2). In Proceedings of the 47th ICAAC Abstracts: 47th Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, IL USA, 17–20 September 2007; American Society for Microbiology: Washington, DC, USA, 2007. [Google Scholar]
  15. Nguyen, C.T.; Bethel, C.; Pettit, N.N.; Charnot-Katsikas, A. From Etest to Vitek 2: Impact of Enterococcal Linezolid Susceptibility Testing Methodology on Time to Active Therapy. Antimicrob. Agents Chemother. 2020, 64, e00302-20. [Google Scholar] [CrossRef] [PubMed]
  16. Qi, C.; Zheng, X.; Obias, A.; Scheetz, M.H.; Malczynski, M.; Warren, J.R. Comparison of testing methods for detection of decreased linezolid susceptibility due to G2576T mutation of the 23S rRNA gene in Enterococcus faecium and Enterococcus faecalis. J. Clin. Microbiol. 2006, 44, 1098–1100. [Google Scholar] [CrossRef] [PubMed]
  17. Tenover, F.C.; Williams, P.P.; Stocker, S.; Thompson, A.; Clark, L.A.; Limbago, B.; Carey, R.B.; Poppe, S.M.; Shinabarger, D.; McGowan, J.E., Jr. Accuracy of six antimicrobial susceptibility methods for testing linezolid against staphylococci and enterococci. J. Clin. Microbiol. 2007, 45, 2917–2922. [Google Scholar] [CrossRef] [PubMed]
  18. Marshall, S.H.; Donskey, C.J.; Hutton-Thomas, R.; Salata, R.A.; Rice, L.B. Gene dosage and linezolid resistance in Enterococcus faecium and Enterococcus faecalis. Antimicrob. Agents Chemother. 2002, 46, 3334–3336. [Google Scholar] [CrossRef]
  19. Chen, H.; Wu, W.; Ni, M.; Liu, Y.; Zhang, J.; Xia, F.; He, W.; Wang, Q.; Wang, Z.; Cao, B.; et al. Linezolid-resistant clinical isolates of enterococci and Staphylococcus cohnii from a multicentre study in China: Molecular epidemiology and resistance mechanisms. Int. J. Antimicrob. Agents 2013, 42, 317–321. [Google Scholar] [CrossRef]
  20. Lee, S.M.; Huh, H.J.; Song, D.J.; Shim, H.J.; Park, K.S.; Kang, C.I.; Ki, C.S.; Lee, N.Y. Resistance mechanisms of linezolid-nonsusceptible enterococci in Korea: Low rate of 23S rRNA mutations in Enterococcus faecium. J. Med. Microbiol. 2017, 66, 1730–1735. [Google Scholar] [CrossRef]
  21. Long, K.S.; Vester, B. Resistance to linezolid caused by modifications at its binding site on the ribosome. Antimicrob. Agents Chemother. 2012, 56, 603–612. [Google Scholar] [CrossRef]
  22. Mendes, R.E.; Deshpande, L.M.; Jones, R.N. Linezolid update: Stable in vitro activity following more than a decade of clinical use and summary of associated resistance mechanisms. Drug Resist. Updates 2014, 17, 1–12. [Google Scholar] [CrossRef]
  23. Liu, Y.; Wang, Y.; Schwarz, S.; Wang, S.; Chen, L.; Wu, C.; Shen, J. Investigation of a multiresistance gene cfr that fails to mediate resistance to phenicols and oxazolidinones in Enterococcus faecalis. J. Antimicrob. Chemother. 2014, 69, 892–898. [Google Scholar] [CrossRef] [PubMed]
  24. Nuesch-Inderbinen, M.; Heyvaert, L.; Treier, A.; Zurfluh, K.; Cernela, N.; Biggel, M.; Stephan, R. High occurrence of Enterococcus faecalis, Enterococcus faecium, and Vagococcus lutrae harbouring oxazolidinone resistance genes in raw meat-based diets for companion animals—A public health issue, Switzerland, September 2018 to May 2020. Eurosurveillance 2023, 28, 2200496. [Google Scholar] [CrossRef] [PubMed]
  25. Li, P.; Yang, Y.; Ding, L.; Xu, X.; Lin, D. Molecular Investigations of Linezolid Resistance in Enterococci OptrA Variants from a Hospital in Shanghai. Infect. Drug Resist. 2020, 13, 2711–2716. [Google Scholar] [CrossRef]
  26. Ha, H.T.A.; Nguyen, P.T.L.; Hung, T.T.M.; Tuan, L.A.; Thuy, B.T.; Lien, T.H.M.; Thai, P.D.; Thanh, N.H.; Bich, V.T.N.; Anh, T.H.; et al. Prevalence and Associated Factors of optrA-Positive-Enterococcus faecalis in Different Reservoirs around Farms in Vietnam. Antibiotics 2023, 12, 954. [Google Scholar] [CrossRef]
  27. Pai, M.P.; Rodvold, K.A.; Schreckenberger, P.C.; Gonzales, R.D.; Petrolatti, J.M.; Quinn, J.P. Risk factors associated with the development of infection with linezolid- and vancomycin-resistant Enterococcus faecium. Clin. Infect. Dis. 2002, 35, 1269–1272. [Google Scholar] [CrossRef] [PubMed]
  28. Olearo, F.; Both, A.; Belmar Campos, C.; Hilgarth, H.; Klupp, E.M.; Hansen, J.L.; Maurer, F.P.; Christner, M.; Aepfelbacher, M.; Rohde, H. Emergence of linezolid-resistance in vancomycin-resistant Enterococcus faecium ST117 associated with increased linezolid-consumption. Int. J. Med. Microbiol. 2021, 311, 151477. [Google Scholar] [CrossRef] [PubMed]
  29. Deekshit, V.K.; Srikumar, S. ‘To be, or not to be’-The dilemma of ‘silent’ antimicrobial resistance genes in bacteria. J. Appl. Microbiol. 2022, 133, 2902–2914. [Google Scholar] [CrossRef]
  30. Dimitriu, T.; Matthews, A.C.; Buckling, A. Increased copy number couples the evolution of plasmid horizontal transmission and plasmid-encoded antibiotic resistance. Proc. Natl. Acad. Sci. USA 2021, 118, e2107818118. [Google Scholar] [CrossRef]
  31. Cook, L.C.; Dunny, G.M. Effects of biofilm growth on plasmid copy number and expression of antibiotic resistance genes in Enterococcus faecalis. Antimicrob. Agents Chemother. 2013, 57, 1850–1856. [Google Scholar] [CrossRef]
  32. Stasiak, M.; Mackiw, E.; Kowalska, J.; Kucharek, K.; Postupolski, J. Silent Genes: Antimicrobial Resistance and Antibiotic Production. Pol. J. Microbiol. 2021, 70, 421–429. [Google Scholar] [CrossRef]
  33. Carvalho, K.R.; Carvalho-Assef, A.P.; Santos, L.G.; Pereira, M.J.; Asensi, M.D. Occurrence of blaOXA-23 gene in imipenem-susceptible Acinetobacter baumannii. Mem. Inst. Oswaldo Cruz 2011, 106, 505–506. [Google Scholar] [CrossRef] [PubMed]
  34. Zhao, S.; White, D.G.; Ge, B.; Ayers, S.; Friedman, S.; English, L.; Wagner, D.; Gaines, S.; Meng, J. Identification and characterization of integron-mediated antibiotic resistance among Shiga toxin-producing Escherichia coli isolates. Appl. Environ. Microbiol. 2001, 67, 1558–1564. [Google Scholar] [CrossRef] [PubMed]
  35. Bender, J.K.; Fleige, C.; Klare, I.; Werner, G. Development of a multiplex-PCR to simultaneously detect acquired linezolid resistance genes cfr, optrA and poxtA in enterococci of clinical origin. J. Microbiol. Methods 2019, 160, 101–103. [Google Scholar] [CrossRef] [PubMed]
  36. Werner, G.; Strommenger, B.; Klare, I.; Witte, W. Molecular detection of linezolid resistance in Enterococcus faecium and Enterococcus faecalis by use of 5′ nuclease real-time PCR compared to a modified classical approach. J. Clin. Microbiol. 2004, 42, 5327–5331. [Google Scholar] [CrossRef]
Antibiotics 13 00101 g001
Figure 1. Final MIC of Enterococcus spp. isolates after exposure to increasing LIN concentrations. Selected isolates with G2576T ribosomal RNA gene mutations or linezolid resistance genes (blue bars) and without (green bars) were challenged with increasing concentrations of LIN from 0.5 mg/L to 16 mg/L. A total of four isolates reached LIN MICs above the EUCAST clinical breakpoint (>4 mg/L, red dotted line). Two LIN-resistant (E. faecium UW21148 and E. faecalis UW23555) and one LIN-susceptible isolate (E. faecalis ATCC29212) from the NRC strain collection served as positive and negative controls, respectively (grey bars). LIN MIC was determined by broth microdilution. Abbreviations: * isolate which had lost the optrA gene during the course of the selection experiments; pos., positive; neg., negative.
Figure 1. Final MIC of Enterococcus spp. isolates after exposure to increasing LIN concentrations. Selected isolates with G2576T ribosomal RNA gene mutations or linezolid resistance genes (blue bars) and without (green bars) were challenged with increasing concentrations of LIN from 0.5 mg/L to 16 mg/L. A total of four isolates reached LIN MICs above the EUCAST clinical breakpoint (>4 mg/L, red dotted line). Two LIN-resistant (E. faecium UW21148 and E. faecalis UW23555) and one LIN-susceptible isolate (E. faecalis ATCC29212) from the NRC strain collection served as positive and negative controls, respectively (grey bars). LIN MIC was determined by broth microdilution. Abbreviations: * isolate which had lost the optrA gene during the course of the selection experiments; pos., positive; neg., negative.
Antibiotics 13 00101 g001
Table 1. Distribution of acquired resistance genes cfr, optrA, poxtA, and of G2576T 23S rRNA gene mutations in E. faecium and E. faecalis isolates displaying a LIN MIC = 4 mg/L (in BMD) collected at the NRC, 2019–2021.
Table 1. Distribution of acquired resistance genes cfr, optrA, poxtA, and of G2576T 23S rRNA gene mutations in E. faecium and E. faecalis isolates displaying a LIN MIC = 4 mg/L (in BMD) collected at the NRC, 2019–2021.
E. faecium
(N = 174)		cfroptrApoxtA23S rDNA
G2576T		n%
----11364.9
---+4928.2
--+-74.0
-+--10.6
-+-+10.6
-++-10.6
+---10.6
+--+10.6
total 174100
E. faecalis
(N = 21)		cfroptrApoxtA23S rDNA
G2576T		n%
-+--1676.0
----419.0
---+15.0
total 21100
Table 2. Linezolid resistance determination by different AST methods among the set of isolates with a LIN MIC of 4 mg/L (in BMD) collected at the NRC, 2019–2021.
Table 2. Linezolid resistance determination by different AST methods among the set of isolates with a LIN MIC of 4 mg/L (in BMD) collected at the NRC, 2019–2021.
LIN MICBMDETEST®VITEK2CHROMagarTM
LIN-R 24 h 1		CHROMagarTM
LIN-R 48 h 1
n%n%n%n%n%
0.75-13 6.7%-
1-44 23%4 2.1%
1.5-35 18%-
2-33 17%121 62%
3-31 16%-
4195100%26 13%13 6.7%12062%11358%
>4-13 6.7%57 29%75 38%8242%
1 no MIC can be inferred from CHROMagarTM LIN-R plates, therefore, growth on plates was considered resistant (>4 mg/L), and no growth as susceptible (=4 mg/L). “-”, no data available for the respective concentrations (BMD, VITEK2).
Table 3. Major error and very major error of different AST methods to correctly identify linezolid-resistant and linezolid-susceptible Enterococcus spp. isolates among the set of isolates with a LIN MIC of 4 mg/L (in BMD) collected at the NRC, 2019–2021.
Table 3. Major error and very major error of different AST methods to correctly identify linezolid-resistant and linezolid-susceptible Enterococcus spp. isolates among the set of isolates with a LIN MIC of 4 mg/L (in BMD) collected at the NRC, 2019–2021.
BMDETEST®VITEK2CHROMagarTM
LIN-R 24 h 1		CHROMagarTM
LIN-R 48 h 1
n%n%n%n%n%
ME 1n.a.0 0%4 2.1%6 3.1%84.1%
VME 278 40%65 33%25 13%9 4.6%4 2.1%
1 Major error (false-resistant, phenotypically resistant but genotypically susceptible isolates); 2 Very major error (false-susceptible, phenotypically susceptible but genotypically resistant isolates). Strains containing either a G2576T ribosomal mutation and/or a resistance gene (cfr, optrA, poxtA) were considered genotypically resistant. False-resistant and false-susceptible isolates were defined under the study assumption that genotypically resistant isolates, with G2576T mutations or transferable resistance genes, should result in a resistant phenotype. Abbreviations: n.a., not applicable (for BMD, no phenotypically resistant isolates were included in the study).
Table 4. PPV of different AST methods for genotype–phenotype correlation according to the potential LIN resistance trait of the study isolates with a LIN MIC of 4 mg/L (in BMD) collected at the NRC, 2019–2021.
Table 4. PPV of different AST methods for genotype–phenotype correlation according to the potential LIN resistance trait of the study isolates with a LIN MIC of 4 mg/L (in BMD) collected at the NRC, 2019–2021.
AST MethodAST ResultcfroptrApoxtAG2576T
nPPV%nPPV%nPPV%nPPV%
ETEST®Susceptible10%1511.8%70%3922%
Resistant0 2 0 11
VITEK2Susceptible0100%758.8%442.9%3672%
Resistant1 10 3 14
CHROMagarTM LIN-R 24 hSusceptible10%194.1%0100%786%
Resistant0 16 7 43
CHROMagarTM LIN-R 48 hSusceptible0100%194.1%0100%394%
Resistant1 16 7 47
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Bender, J.K.; Fleige, C.; Funk, F.; Moretó-Castellsagué, C.; Fischer, M.A.; Werner, G. Linezolid Resistance Genes and Mutations among Linezolid-Susceptible Enterococcus spp.—A Loose Cannon? Antibiotics 2024, 13, 101. https://doi.org/10.3390/antibiotics13010101

AMA Style

Bender JK, Fleige C, Funk F, Moretó-Castellsagué C, Fischer MA, Werner G. Linezolid Resistance Genes and Mutations among Linezolid-Susceptible Enterococcus spp.—A Loose Cannon? Antibiotics. 2024; 13(1):101. https://doi.org/10.3390/antibiotics13010101

Chicago/Turabian Style

Bender, Jennifer K., Carola Fleige, Finn Funk, Clara Moretó-Castellsagué, Martin A. Fischer, and Guido Werner. 2024. "Linezolid Resistance Genes and Mutations among Linezolid-Susceptible Enterococcus spp.—A Loose Cannon?" Antibiotics 13, no. 1: 101. https://doi.org/10.3390/antibiotics13010101

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop