Staphylococcus aureus with an erm-mediated constitutive macrolide-lincosamide-streptogramin B resistance phenotype has reduced susceptibility to the new ketolide, solithromycin

Background Solithromycin, the fourth generation of ketolides, has been demonstrated potent antibacterial effect against commonly-isolated gram-positive strains. However, Staphylococcus aureus (S. aureus) strains with a higher solithromycin MIC have already been emerged, the mechanism of which is unknown. Methods Antimicrobial susceptibility test was performed on 266 strains of S. aureus. The antibiotic resistance phenotype of erm-positive strain was determined by D-zone test. Spontaneous mutation frequency analysis was performed to compare the risk levels for solithromycin resistance among different strains. Efflux pumps and mutational analysis of ribosomal fragments as well as erm(B) gene domains were detected. Quantitative reverse transcription polymerase chain reaction was conducted to compare the transcriptional expression of the erm gene between the constitutive macrolide-lincosamide-streptogramin B (cMLSB)- and inducible MLSB (iMLSB)-phenotypes. Results In the erm-positive S. aureus strains, the minimum inhibitory concentration (MIC)50/90 of solithromycin (2/> 16 mg/L) was significantly higher than that in the erm-negative strains (0.125/0.25 mg/L). Of note, the MIC50 value of the strains with iMLSB (0.25 mg/L) was significantly lower than that of the strains with cMLSB (4 mg/L). A comparison among strains demonstrated that the median mutational frequency in isolates with cMLSB (> 1.2 × 10− 4) was approximately > 57-fold and > 3333-fold higher than that in iMLSB strains (2.1 × 10− 6) and in erythromycin-sensitive strains (3.6 × 10− 8), respectively. The differential antibiotic in vitro activity against strains between cMLSB and iMLSB could not be explained by efflux pump carriers or genetic mutations in the test genes. The expression of the erm genes in strains with cMLSB did not differ from that in strains with iMLSB. Conclusions The reduced susceptibility to solithromycin by S. aureus was associated with the cMLSB resistance phenotype mediated by erm. Electronic supplementary material The online version of this article (10.1186/s12879-019-3779-8) contains supplementary material, which is available to authorized users.


Background
The emergence and rapid transmission of antibiotic-resistance genes among Staphylococcus aureus (S. aureus) strains pose serious public health challenges worldwide [1]. The erythromycin ribosome methylase (erm) genes encode proteins that methylate adenine residues A2058/ 2059 in the peptidyl transferase region of 23S rRNA domain V, and are responsible for macrolide, lincosamide, and streptogramin B (MLSB) antibiotic resistance [2]. In some regional reports, the frequency of methicillin-resistant S. aureus (MRSA) with macrolide resistance was over 90% while methicillin-sensitive S. aureus (MSSA) rose gradually to over 40% [3]. The rapid transmission and broad antibiotic resistance spectrum of erm have greatly limited the clinical utility of traditional macrolides such as erythromycin and azithromycin.
The Erm gene-mediated resistant strains exhibit two antimicrobial resistance phenotypes, constitutive MLSB (cMLSB) and inducible MLSB (iMLSB). These phenotypes can be distinguished by D-zone test and are due to different molecular regulatory mechanisms. The translation initiation of erm in the iMLSB strain is inhibited due to the sequestration of its mRNA ribosome-binding site by the leader peptide as well as dependent on inducers like erythromycin binding to the leader peptide, by which constrains its role and releases the mRNA ribosome-binding site. In contrast, cMLSB strains allow for direct and timely inducer independent translation of transcripts, in that variations in the leader peptide gene sequence abolish its translation [4,5]. Through this mechanism, the cMLSB phenotype confers resistance to macrolides, lincosamides, and streptogramin B, while the iMLSB phenotype is resistant to macrolides and streptogramin B but sensitive to lincosamides.
The novel agent, solithromycin, has been reported to be effective against erythromycin-resistant strains and have a formidable antibacterial effect with an extensive antibacterial spectrum. The minimum inhibitory concentrations (MICs) of solithromycin in drug-resistant or multi-drug-resistant strains of clinical isolates (methicillin-resistant S. aureus and macrolide-lincosamide-streptogramin B-resistant Streptococcus pneumoniae) are comparable or generally lower than those of other regularly used antibiotics, such as linezolid or vancomycin. The antibacterial spectrum of solithromycin includes Haemophilus influenzae, Moraxella catarrhalis, betahemolytic streptococci, Legionella, Bordetella pertussis, Chlamydophila pneumoniae, and S. aureus. However, S. aureus strains with a solithromycin MIC over 32 mg/L have already emerged, the mechanism of which is unknown [6][7][8].
In order to address this unknown mechanism, we conducted antimicrobial susceptibility testing of solithromycin against erm-positive and -negative strains, analysed the differences in MIC distribution among strains, and explored the potential basis of solithromycin resistance.

Bacterial isolates
A total of 266 strains (without repetition) were isolated from patients in the Nanshan District People's Hospital of Shenzhen, China from 2013 to 2016. These strains were recovered from respiratory tract secretions, blood, pus, and wound fluid. Strains were identified using the BD Phoenix™-100 Automated Microbiology System (Bd-bio, NJ, USA) according to the manufacturer's instructions. The presence of the erm, mecA, and femB genes was confirmed by polymerase chain reaction (PCR) as previously described [9,10]. The verified strains were stored at − 80°C in Tryptic Soy Broth (TSB) containing 40% glycerol. All procedures involving human participants were performed in accordance with the ethical standards of the Shenzhen University School of Medicine and the 1964 Helsinki declaration and its later amendments. For this type of study, formal consent is not required.

D-shaped zone of inhibition
All 236 strains carrying the erm genes were tested by the previously described disc diffusion method for phenotypic identification [11]. The erythromycin and clindamycin susceptibility test discs were purchased from Sigma Aldrich (St. Louis, MO, USA), and the erythromycin and solithromycin powders were purchased from Cempra Pharmaceuticals (Chapel Hill, USA). After 16 to 18 h of incubation, the D-zone test results were assessed by transmitted or reflected light and interpreted according to the Clinical and Laboratory Standards Institute (CLSI) guidelines [12].

Antimicrobial susceptibility testing
Antimicrobial susceptibility test of these strains was performed by the agar dilution method, following the guidelines of the CLSI. The minimum inhibitory concentration (MIC) was defined as the lowest concentration of antibiotic that completely inhibited growth of the organism in the agar plate as detected by the unaided eye [12]. The MICs for erythromycin and solithromycin were determined. MIC 50 and MIC 90 values were defined as the lowest concentration of the antibiotic at which 50 and 90% of the isolates were inhibited, respectively. CLSI breakpoints were used for MIC interpretation.

Spontaneous mutation frequency
The protocol for how to detect the spontaneous mutation frequency of solithromycin resistance is referred to a previous report [13], in an effort to compare the risk levels between the erm-positive and -negative groups as well as the iMLSB and cMLSB groups. Median frequencies were calculated for each group and were used to infer frequency ratios among different groups. Representatives of putative mutant colonies for each plate and group were reassessed for solithromycin MIC determining as described above.
Amplification and sequencing of efflux pump, ribosomal, and erm genes The roles of efflux pumps, drug binding sites and erm gene mutations in the transformation of S. aureus from solithromycin susceptibility to resistance were assessed. To do so, 60 erm(B)-positive representative strains were selected and divided into two groups, one with solithromycin MIC values ≤1 mg/L and the other with MIC values ≥4 mg/L. PCR amplification and sequencing were employed to screen for efflux and gene mutations.
Total DNA from all isolates was extracted and purified with the DNeasy Blood and Tissue Kit (QIAGEN China Co., Shanghai, China) according to the manufacturer's protocol for Gram-positive bacteria. Efflux pumps, MsrA/B and MefE/A, were detected by PCR. Macrolide binding position mutations for the six copies of S. aureus rrn operons (including 5S, 16S, and 23S rRNA genes) and ribosomal proteins [L3 (rplC), L4 (rplD), and L22 (rplV)] were amplified using previously described primers and cycling parameters [13,14]. The ermB gene coding and leader peptide regions were amplified using primers ermBF/ermBR and ermBLF/ermBLR, respectively (Additional file 1: Table S1). PCR products were sequenced (BGI, Shenzhen, China) with specific primers (Additional file 1: Table S1).

Quantitative reverse transcription-polymerase chain reaction (qRT-PCR)
To explore the impact of transcriptional expression of erm(A), erm(B), and erm(C) on solithromycin MIC values, qRT-PCR was performed for iMLSB strains with solithromycin MICs ≤1 mg/L and cMLSB strains with solithromycin MICs ≥4 mg/L.
Total bacterial RNA was extracted using the RNeasyH Mini Kit and reverse transcribed into cDNA using iScript reverse transcriptase (Bio-Rad, Hercules, CA, USA) according to the manufacturer's instructions. Subsequently, qRT-PCRs were performed using SYBR green PCR reagents (Premix EX TaqTM, Takara Biotechnology, Dalian, China) with the Mastercycler realplex system (Eppendorf AG, Hamburg, Germany). As a reference gene, 16S rRNA was used to normalize transcriptional levels. All qRT-PCRs were carried out in triplicate with at least three independent RNA samples. The primers are listed in Additional file 1: Table S1.

Statistical analysis
The data were not normally distributed so median mutational frequencies were calculated for each group. Statistical analysis was carried out using SPSS for windows (version 19.0; SPSS, Chicago, IL, USA). The difference in positive percentages (%) between groups was compared by the chi-square test. MIC values for different bacterial groups were compared using the Kruskal-Wallis test. A P value of less than 0.05 was considered statistically significant.  Table 1).

Spontaneous mutation frequencies of erm-positive and -negative strains
The erm-mediated cMLSB strains had the highest spontaneous mutation frequency, followed by the erm-mediated iMLSB strains and the erm-negative strains. The median mutation frequency of the cMLSB strains (> 1.2 × 10 − 4 ) was > 57-fold and > 3333-fold higher than that of the iMLSB strains (2.1 × 10 − 6 ) and the erm-negative strains (3.6 × 10 − 8 ), respectively. The time required for the cMLSB strains to exhibit mutants was shorter than that required for the iMLSB strains and the erm-negative strains ( Table 2). For each group, putative mutant colonies appearing on 2 × MIC solithromycin Müller-Hinton agar (MHA) plates, had reduced susceptibility to solithromycin (Additional file 1: Table S2).

Efflux detection and mutation analysis of macrolidetargeted sites and erm(B) gene
After PCR amplification and sequencing, the efflux pumps were detected negative. Meanwhile, among the 60 representatives, no site mutations, as previously reported to have a definite relationship with macrolide antibiotics resistance, were founded either in the ribosomal genes or erm(B). Some mutations detected may be just nonsense or randomized events during cell replication because they were out of any laws on their distributions (Additional file 2: Table S3).

Expression of erm(A), erm(B), and erm(C) genes by iMLSB and cMLSB strains
To our surprise, after qRT-PCR of these genes, expression of the erm-positive strains with the cMLSB phenotype appeared to be slightly lower than that of the strains with the iMLSB phenotype. In comparison with the iMLSB phenotype, the relative expression of erm(A), erm(B), erm(C) in strains with the cMLSB phenotype was 0.23-, 0.54-, and 0.38-fold, respectively (Additional file 3: Figure S1).

Discussion
Limited data shows that solithromycin has a more potent antimicrobial activity against a variety of bacteria than traditional and novel marolides such as erythromycin and telithromycin. According to previous reports, solithromycin MIC 50/90 values were 0.008/0.12 mg/L for S. pneumoniae, 0.06/0.12 mg/L for Moraxella catarrhalis, 0.015/0.03 mg/L for beta-hemolytic streptococci, 1/2 mg/L for Haemophilus influenzae, 0.06/0.06 mg/L for MSSA, and 0.06/> 32 mg/L for MRSA [7]. Solithromycin exhibits different in vitro antimicrobial activity against S. aureus. Assessing this difference is critical because the effect of solithromycin on S. aureus in China is unclear. In this study, we found several specific features of solithromycin susceptibility in S. aureus strains from China.   . 3 The number of strains with each solithromycin MIC for the iMLSB and cMLSB strains. The number of isolates with each solithromycin MIC was counted separately for erm-mediated different MLSB phenotypes, by which the difference in solithromycin antibacterial effect between iMLSB-and cMLSB-phenotypic strains was compared. MIC was determined by the agar dilution method according to the guidelines of the CLSI is different from a previous investigation, in which solithromycin resistance was predominate in MRSA strains [7]. Third, the MIC 50 values for the iMLSB strains were dramatically lower than those of the cMLSB strains, suggesting that the erm-mediated cMLSB phenotype increases solithromycin MICs and is a signature pattern for solithromycin resistance. To the best of our knowledge, no other reports have demonstrated that the cMLSB phenotype in S. aureus predicts solithromycin resistance.
Although strains with the cMLSB phenotype primarily exhibited solithromycin resistance, a few cMLSB strains did show low solithromycin MIC values. The majority of S. aureus strains with the iMLSB phenotype were sensitive to solithromycin, but whether possession of the erm gene increases the risk for solithromycin resistance under antibiotics pressure needs to be determined. The spontaneous mutation frequency is a simple and practical method to evaluate the resistance risk during antibiotic pressure [15] and is used to assess the effect of cMLSB on solithromycin resistance in S. aureus and the resistance risk for the iMLSB phenotype. The data demonstrate that harboring erm genes predicts the risk for solithromycin resistance to antibiotic stress, as solithromycin-sensitive S. aureus with cMLSB phenotype is at greater risk than with iMLSB phenotype, which were at greater risk than erythromycin-sensitive S. aureus. This conclusion is consistent with previous findings reported by Pamela McGhee et al., who demonstrated that the degree of solithromycin resistance in erm(B)-positive S. pneumoniae and Streptococcus pyogenes strains was greater than that in erythromycin-sensitive counterparts. However, this report did not demonstrate that erm gene mediated cMLSB was a signature of solithromycin resistance [16]. In summary, the solithromycin-sensitive strains of S. aureus with cMLSB have an increased risk of resistance, which is far higher than that of strains with iMLSB or with erythromycin-sensitivity. It is well known that pharmacological effects are influenced by a number of factors including antimicrobial susceptibility and desirable pharmacokinetic and pharmacodynamic parameters, such as high bioavailability. Therefore, further evaluation is necessary to determine the clinical significance of solithromycin-sensitive strains with the iMLSB phenotype (and potentially high resistance mutation frequency) during solithromycin antibiotic pressure.