Abstract
Linezolid has been reported to restore erythromycin susceptibility in erythromycin-resistant Staphylococcus aureus. This phenomenon has not been reported in enterococci and the mechanisms involved therein are still unknown. The purpose of this study was to investigate the mechanisms involved and the effect of combining linezolid with erythromycin on erythromycin-resistant enterococci. Checkerboard techniques were used to determine drug interactions, and 12 of 14 isolates showed a synergistic effect between erythromycin and linezolid (fractional inhibitory concentration <0.5). We observed that the erm(B) gene, which encodes a dimethyltransferase responsible for erythromycin resistance, was expressed from transposon Tn1545 in the tested erythromycin-resistant enterococci. After exposure to linezolid, erm(B)-mediated rRNA dimethylation at A2071 could not be detected, and the erm(B) gene was lost following acquisition of erythromycin susceptibility. Thus, in conclusion, linezolid combined with erythromycin exerts a synergistic effect against erythromycin-resistant enterococci. Linezolid treatment suppressed erm(B)-mediated rRNA dimethylation at A2071, which could lead to loss of the erm(B) gene.
Introduction
Enterococci are one of the most common Gram-positive bacteria causing nosocomial infections 1 that are difficult to treat because of intrinsic resistance to many antimicrobial agents as well as a high ability to acquire and disseminate resistance determinants. Multidrug-resistant enterococci, especially vancomycin-resistant enterococci, often cause serious and incurable infections.2,3 Therefore, new therapeutic strategies, including combination therapies, are urgently needed.
Linezolid is the first member of the class of oxazolidinones, which alters bacterial protein synthesis by binding in the A site of the peptidyl transferase center. In addition, linezolid can also prevent the assembly of the functional 70S initiation complex, preventing the initiation phase of translation. 4 Linezolid is considered as one of the last lines of defense against multidrug-resistant Gram-positive bacteria. 5
However, linezolid causes a variety of adverse reactions, such as hematotoxicity and neurotoxicity, and these adverse reactions may occur frequently in patients with long-term use. 6 Antibiotic combinations may be used in clinical practice to extend the antimicrobial spectrum and minimize toxicity or resistance and even exert synergistic antimicrobial effects. 7 Linezolid showed synergistic effects with several antimicrobial agents, including with erythromycin, against penicillin-intermediate Streptococcus pneumoniae strains.8,9
Erythromycin is a 14-membered macrolide antibiotic that is mainly used to treat Gram-positive bacterial infections. 10 As enterococcal resistance to erythromycin is widespread, erythromycin is not used to treat enterococcal infections. 11 In our previous study, erythromycin-resistant enterococci become sensitive to erythromycin after exposure to linezolid.
The target sites of linezolid and erythromycin are close, with linezolid binding to the ribosome in the A-site cleft and erythromycin binding in the nascent peptide exit tunnel.12,13 Interactions may exist when the two drugs are present at the same time. In this study, we applied the checkerboard method to test the activities of linezolid plus erythromycin against enterococci.
It has been reported that for erythromycin-resistant methicillin-resistant Staphylococcus aureus (MRSA), which developed linezolid resistance during clinical treatment, erythromycin susceptibility was restored while retaining resistance to linezolid. 14 This suggests that the mechanisms of linezolid resistance may be incompatible with erythromycin resistance. This phenomenon has not been reported in enterococci.
Considering that linezolid is increasingly used in the treatment of enterococcal infections, especially treatment of vancomycin-resistant enterococcal infections, we investigated the mechanism of loss of erythromycin resistance in enterococci. High-level erythromycin resistance to enterococci is usually attributed to the presence of the erm(B) gene, 15 which gives rise to N6-dimethylation at A2058 of the 23S rRNA, in the Escherichia coli numbering scheme, which corresponds to nucleotide 2071 in enterococci. 16
The erm(B) gene is easily disseminated through conjugative plasmids and conjugative transposons 17 and is frequently reported on the highly mobile conjugative transposon Tn1545 in enterococci. 18 Therefore, we considered that the erm(B) gene loss might play an important role in restoration of erythromycin susceptibility. To test this hypothesis, the presence of erm(B), Tn1545, and dimethylation was assayed in enterococci.
Materials and Methods
Bacterial strains, antimicrobial agents, and growth media
Seven clinical enterococcus strains used in this study (shown in Table 1) were obtained from the Clinical Microbiology Laboratory of the First Affiliated Hospital of Shantou University Medical College (Shantou, China). Linezolid (LNZ) and erythromycin (ERY) were purchased from Solarbio (Beijing, China). Mueller–Hinton broth (MHB), brain–heart infusion (BHI), and bacteriological agar were purchased from OXOID (Basingstoke, United Kingdom).
Minimum Inhibitory Concentration of Linezolid and Development of Erythromycin Resistance to Linezolid
Bold values indicate the strains restored erythromycin susceptibility.
Bacteria were cultured with 0.5 × MIC of LNZ. The MIC of LNZ and ERY was measured when the medium became visibly turbid. The newly acquired MIC of LNZ was used in a new round of induction.
ERY, erythromycin; LNZ, linezolid; MIC, minimum inhibitory concentration; VRE, vancomycin-resistant E. faecium.
Susceptibility and checkerboard testing
Susceptibility and checkerboard testing were performed by the broth microdilution method using cation-adjusted MHB. Bacteria were added to a final concentration of 1 × 105 CFU/mL and cultured at 37°C for 20 hr. Results were interpreted according to the Clinical and Laboratory Standards Institute (CLSI) guidelines of 2018. 19
The checkerboard technique was performed in 96-well plates, as described elsewhere. 20 Briefly, 50 μL of linezolid, from 8 to 1/16 × minimum inhibitory concentration (MIC) by twofold dilution, was prepared in the horizontal direction, but the concentration in row A was 16 to 1/8 × MIC. Next, 50 μL of the 16 × MIC erythromycin dilution was added to row A, and the result was that the concentration of linezolid in each column remained the same, while the concentration of erythromycin in row A was 8 × MIC.
Erythromycin was diluted to the lowest concentration of 1/16 × MIC by twofold dilution in the vertical direction. Bacteria were added to a final concentration of 1 × 105 CFU/mL and cultured at 37°C for 20 hr. The concentration of linezolid and erythromycin in the 96-well plates can be seen in Supplementary Table S1. The best combination that prevented turbidity was observed.
Determination of erythromycin susceptibility following linezolid exposure
Bacteria were cultured in BHI broth with 0.5 × MIC of linezolid at 37°C for 24 hr. Then a 10-μL aliquot of culture was transferred into fresh BHI broth with the antibiotic and the cycle was repeated. When visible growth of a microorganism was detected by the unaided eye, the MIC of linezolid and erythromycin was measured. The newly acquired MIC of linezolid was used in a new round of induction.
Taking the Enterococcus faecium 301449 strain as an example, after 42 days of linezolid exposure, the strain that was resistant to linezolid was designated as E. faecium 301449-L.
Detection of the erm(B) gene and transposon Tn1545
Bacterial DNA was extracted using the MiniBEST Bacteria Genomic DNA Extraction Kit (Cat. No.: T9763; TAKARA). PCR was performed with PrimeSTAR Max DNA polymerase (Cat. No.: R045A; TAKARA) and a PCT-200 PCR cycler (MJ Research).
The erm(B) gene was amplified by using the forward primer 5′-GAAAAGGTACTCAACCAAATA-3′ and reverse primer 5′-AGTAACGGTACTTAAATTGTTTAC-3′. Transposon Tn1545 was amplified by using the forward primer 5′-CTTAGAAGCAAACTTAAGAGTGTGT-3′ and reverse primer 5′-GGTTGAGTACCTTTTCATTCGTTAA-3′. 21
The amplified products were analyzed on 1.0% agarose gels.
Primer extension to determine dimethylation at A2071 in enterococcus 23S rRNA
A primer extension assay was performed as previously described. 22 Total RNA was extracted using the RNeasy Protect Mini Kit (Cat. No.: 74134; Qiagen). Then, 10 μL of reaction solution, containing 3 μg of total RNA, 2 pmol of 5′-FAM-labeled oligonucleotide primer, and 500 μM each of dGTP, dCTP, dTTP, and ddATP (Sigma), was heated to 65°C for 5 min and rapidly cooled on ice.
Another 10-μL mixture containing 200 U PrimeScript II RTase (Cat. No.: 2690A; TAKARA) was added, then reaction mixtures were heated at 45°C for 50 min and incubated at 70°C for 15 min for primer extension. The primers were complementary to enterococcus 23S rRNA bases 2079 to 2114 (5′-CAAACACTCAATATCAAACTACAGTAAAGCTCCATG-3′). Extension products were separated on 12% polyacrylamide sequencing gels and visualized using the Tanon-5200 Multi-Imaging System (Tanon).
Detection of erm(B) expression
To determine whether the expression of the erm(B) gene changes in the presence of linezolid, enterococci were inoculated in MH medium either with or without 0.5 mg/L linezolid and then incubated at 37°C overnight. Total RNA was extracted and first-strand cDNA was synthesized from 1 μg of total RNA by using the PrimeScript RT Reagent Kit with gDNA Eraser (Cat. No.: RR047A; TAKARA) and a PCT-200 PCR cycler. After reverse transcription, 1 μL of cDNA was amplified by TB Green Premix Ex Taq™ II (Cat. No.: RR820A; TAKARA) in an ABI 7500 Real-Time PCR System (Applied Biosystems).
The erm(B) gene was amplified by using the forward primer 5′-TGAATCGAGACTTGAGTGTGCAA-3′ and reverse primer 5′-GGATTCTACAAGCGTACCTT-3′ and normalized to control 23S rDNA, which was amplified by using the forward primer 5′-AGAAATTCCAAACGAACTTG-3′ and reverse primer 5′-CAGTGCTCTACCTCCATCATT-3′. 23
Amplification was performed under the following conditions: denaturation at 95°C for 30 sec, then 40 cycles at 95°C for 5 sec and 60°C for 34 sec, followed by another dissociation stage at 95°C for 15 sec, 60°C for 1 min, and 95°C for 15 sec. Fold change was calculated using the 2−ΔΔCt method and presented as the fold change in expression in groups treated with 0.5 mg/L linezolid relative to the untreated group.
Ethics approval and informed consent
The protocol has been reviewed by the Ethics Committee (IRB) of Shantou University Medical College. As strains that were used in the clinical diagnosis were cultured from residual samples, confidentiality of patient data was preserved and compliance with the Declaration of Helsinki was ensured. As the data did not affect patient care, the exemption criteria were met.
Written informed consent from the patient was not required (ethical approval No. SUMC-2021-094).
Results
Erythromycin susceptibility is restored in erythromycin-resistant enterococci upon acquisition of resistance to linezolid
Enterococci used in this study were highly resistant to erythromycin, as judged by antimicrobial susceptibility testing. When linezolid resistance was induced in the seven tested strains, five strains lost their resistance to erythromycin (MICs were decreased to 0.125–1 mg/L) (Table 1). Enterococcus faecalis acquired a higher level of resistance to linezolid than E. faecium. It is interesting to note that E. faecalis B1232 and E. faecalis 8139 maintained resistance to erythromycin even when the linezolid concentration reached 64 mg/L.
Linezolid and erythromycin exert synergistic toxicity in enterococcus strains
The activities of linezolid plus erythromycin were tested using the checkerboard method (Table 2). When the best combination that prevented turbidity in the checkerboard test was observed, we calculated fractional inhibitory concentration (FIC) indices by the following formula: (MIC of LNZ in combination/MIC of LNZ alone)+(MIC of ERY in combination/MIC of ERY alone).
Linezolid and Erythromycin Exert Synergistic Effects on Enterococcus Strains
FIC indices of ≤0.5 are defined as synergistic interactions; FIC indices were calculated by the following formula: (MIC of LNZ in combination/MIC of LNZ alone)+(MIC of ERY in combination/MIC of ERY alone).
Strains at day 42 are strains (day 0) that are resistant to linezolid after 42 days of exposure to linezolid, refer to Table 1.
FIC, fractional inhibitory concentration.
FIC indices of ≤0.5 were defined as synergistic interactions. 8 In addition to E. faecalis 3095-L and VRE 310682-L, linezolid and erythromycin showed a synergistic effect in 12 other enterococcus strains.
The erm(B) gene and Tn1545 are lost following acquisition of erythromycin susceptibility in enterococci
The presence of the erm(B) gene and the Tn1545 transposon in enterococcus strains is shown in Fig. 1A. Primers for Tn1545 used in this study identified a fragment that included the regulatory region and 48 bp of the 5′ end of erm(B). 21 If erm(B) and Tn1545 are both positive, it would indicate that erm(B) is carried by Tn1545. Except for E. faecium 301449, all other strains carried the erm(B) gene on Tn1545.

Dimethylation at A2071 mediated by the erm(B) gene disappeared after strains became resistant to linezolid.
Both erm(B) and Tn1545 were lost in strains that developed linezolid resistance and acquired erythromycin susceptibility simultaneously. E. faecalis B1232-L and E. faecalis 8139-L did not become sensitive to erythromycin and retained the erm(B) gene and Tn1545 following acquisition of resistance to linezolid.
Expression of the erm(B) gene in enterococci is regulated by linezolid
Real-time PCR was performed to determine the effect of linezolid on expression of the erm(B) gene. Our results showed that expression of erm(B) in enterococci was upregulated by 0.5 mg/L linezolid, except in VRE 310682. For E. faecalis, expression of erm(B) in E. faecalis 3095 was increased 324-fold, whereas in E. faecalis B1232 and E. faecalis 8139, erm(B) expression was only upregulated 39-fold and 17-fold, respectively (Fig. 2).

Fold changes in the erm(B) transcript (relative to 23S rDNA). Bars show the mean ± SEM (untreated vs. treated with 0.5 mg/L linezolid, n = 3). Significant differences (*p < 0.05 and **p < 0.01) are indicated by asterisks. SEM, standard error of the mean.
erm(B)-mediated rRNA dimethylation at A2071 is eliminated by linezolid
Dimethylation at A2071 was detected by reverse transcriptase extension from base 2079 to 2114, as described above. Dimethylation at A2071 aborts the extension of reverse transcription, and if this site is unmethylated, reverse transcription will be terminated by incorporation of ddATP at U2054. A2071 in enterococcus 23S rRNA was dimethylated, even in E. faecium 301449, which lacked the erm(B) gene (Fig. 1B).
Our results suggest that E. faecium 301449 may carry other Erm gene subtypes. When enterococci became susceptible to erythromycin, dimethylation at A2071 was barely detected, even in strains that remained resistant to erythromycin (E. faecalis B1232-L and E. faecalis 8139-L).
Discussion
Enterococcal infections are difficult to treat because of their multidrug resistance, 1 including resistance to erythromycin. Bacterial resistance to erythromycin depends on the following factors: target site alteration, drug efflux, or drug inactivation.24,25 The main reason for the high resistance of enterococci to erythromycin is that these strains carry the erm(B) gene, 15 which induces N6-dimethylation at A2071 of the enterococcus 23S rRNA.
In this study, except for E. faecium 301449, all other strains carried the erm(B) gene. Besides the erm(B) gene, there are dozens of different erm genes encoding methyltransferases, which also presumably cause dimethylation at A2071. 25 This could explain why the A2071 of E. faecium 301449 is dimethylated despite lacking the erm(B) gene (Fig. 1).
Loss of erythromycin resistance in MRSA has been reported in in vitro and in vivo studies,14,26 but this phenomenon has not been previously reported in enterococci, and the mechanisms involved remain unknown. It may provide a new reference for clinical medication, but there is little literature reporting on the mechanisms involved therein, and only few studies report that loss of erythromycin resistance is accompanied by disappearance of Erm dimethyltransferase. 26
In this study, after being exposed to linezolid, enterococci became resistant to linezolid and sensitive to erythromycin, except for E. faecalis B1232 and E. faecalis 8139. Dimethylation of A2071 in all E. faecium strains could not be detected after exposure to linezolid and paralleled the loss of the erm(B) gene. Linezolid-resistant E. faecalis isolates showed a variable response to erythromycin.
Unlike E. faecalis 3095-L, E. faecalis B1232-L and E. faecalis 8139-L remained resistant to erythromycin and the erm(B) gene expression also remained positive, although dimethylation was almost undetectable. The expression of erm(B) in E. faecalis was upregulated by 0.5 mg/L linezolid. In particular, erm(B) expression in E. faecalis 3095 was remarkably increased 324-fold, while the upregulated erm(B) expression levels in E. faecalis B1232 and 8139 were only 39-fold and 16-fold, respectively (Fig. 2).
Gupta et al have shown that erythromycin resistance methyltransferase expression reduces Staphylococcus aureus fitness, demonstrating that expression of Erm has an adverse effect on cell growth. 10 The burden of erm(B)-containing elements on in vitro fitness has been observed for Clostridium difficile. 27 This evidence indicates that E. faecalis 3095 may have a higher fitness cost due to erm(B) expression in the presence of linezolid compared with E. faecalis B1232 and 8139.
The erm(B) gene is carried by the conjugative transposon Tn1545, which is reported to be highly mobile. 18 The mobility of transposon Tn1545 may promote loss of the erm(B) gene. We postulate that a linezolid-mediated loss of A2071 dimethylation occurs whereby strains selectively lose the erm(B) gene to reduce their fitness cost, and because the fitness cost of erm(B) expression is low in E. faecalis B1232 and 8139 (compared with E. faecalis 3095), the erm(B) gene is retained.
The reason why dimethylation cannot be detected in the presence of the erm(B) gene is unclear. Induced linezolid resistance is usually mediated by alterations in the 23S rRNA sequence. DNA sequencing of 23S rRNA showed that there were G2576T mutations in E. faecalis B1232-L and 3095-L, and no mutation was detected in other strains (Supplementary Table S2). The other resistance mechanism of linezolid may be the reason why demethylation cannot be detected, which needs to be further studied.
Some antibiotic combinations may provide a synergistic effect in the treatment of enterococcal infections. Linezolid inhibits the formation of the initiation complex for protein synthesis, whereas erythromycin inhibits peptide elongation.4,28 Such a combination could double block the synthesis of proteins.
Sweeney et al have previously shown the ability of linezolid to combine with other antimicrobial agents and enhance cytotoxicity against S. aureus, E. faecalis, E. faecium, and S. pneumoniae. 8 In that study, linezolid plus erythromycin showed an indifferent effect on S. aureus, but a synergistic effect on penicillin-intermediate S. pneumoniae strains. The enterococcus strains used in our study were all highly resistant to erythromycin, but we found that the synergistic effect of linezolid plus erythromycin still applies.
Most importantly, there is still a significant synergistic effect on E. faecalis B1232-L and E. faecalis 8139-L, which are resistant to both erythromycin and linezolid. However, a limitation to this study is that synergism still needs to be measured by time-kill assays, followed by in vivo experiments. According to Miller et al, exposure to low levels of erythromycin can delay development of linezolid resistance in S. aureus. 29
Therefore, a combination of linezolid and erythromycin may still play a role in controlling linezolid resistance.
Conclusions
In conclusion, this study has two novel findings. The first is that linezolid and erythromycin have a synergistic effect on erythromycin-resistant enterococci following development of resistance to linezolid. The second finding is that erm(B)-mediated rRNA demethylation can disappear upon acquiring resistance to linezolid, which may be due to loss of the erm(B) gene as a result of fitness cost.
Footnotes
Authors' Contributions
W.C. was involved in conceptualization, data curation, methodology, software, funding acquisition, writing—original draft, and writing—review and editing. Y.H. was involved in resources, investigation, and writing—review and editing. X.J. was involved in supervision, methodology, and writing—review and editing. J.Y. was involved in validation, data curation, and writing—review and editing. Y.L. was involved in investigation, methodology, and writing—review and editing. F.Y. was involved in conceptualization, visualization, project administration, supervision, funding acquisition, and writing—review and editing.
Disclosure Statement
No competing financial interests exist.
Funding Information
This work was supported by the National Natural Science Foundation of China (Nos. 81273562 and Nos. 82104235) and Guangdong Basic and Applied Basic Research Foundation (Nos. 2018A0303130159).
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
