Twenty-seven-nucleotide repeat insertion in the rplV gene confers specific resistance to macrolide antibiotics in Staphylococcus aureus

Macrolide antibiotics are used for treatment of soft-tissue infection caused by Staphylococcus aureus in humans. However, infections with S. aureus are increasingly difficult to treat owing to the emergence and rapid spread of multiple-drug resistant S. aureus. Resistance to macrolide in S. aureus is mostly due to the modification of 23 S rRNA by methylases encoded by erm genes. Here, we have identified that a 27-nucleotide repeat sequence insertion in the rplV gene induced a specific resistance to macrolide antibiotics. An erythromycin-resistant strain, 8325ER+, was screened by resistance to erythromycin from the macrolide-sensitive strain 8325-4. Comparative genome sequencing analysis showed that 8325ER+ contained a 27-nt repeat sequence insertion in the rplV gene that encodes the ribosomal protein L22, when compared to its parent strain. The 27-nt repeat sequence led to an insertion of 9 amino acids in L22, which had been identified to reduce the sensitivity to erythromycin and other macrolide antibiotics. Moreover, we show that the ectopic expression of the mutated rplV gene containing the 27-nt repeat sequence insertion in several susceptible strains specifically conferred resistance to macrolide antibiotics. Our findings present a potential mechanism of resistance to macrolide antibiotics in S. aureus.


INTRODUCTION
Staphylococcus aureus is the leading Gram-positive bacterium that can cause infections in humans worldwide, including mild skin infections, bacteremia, sepsis, and endocarditis [1][2][3]. Over the last century, infections with S. aureus have become increasingly difficult to treat owing to the emergence and rapid spread of multiple-drug resistant S. aureus [4][5][6].
Macrolides, which consist of a 14-to 16-membered lactone ring with different appended sugars and comprise a key group of inhibitors of bacterial translation, are ribosome-targeting antibiotics used to treat infections caused by Staphylococcus species [7,8]. Erythromycin, azithromycin, and clarithromycin are members of the macrolide antibiotics, a large group of antibacterial agents that include natural or newer semi-synthetic compounds [9,10]. Their inhibitory activity depends on binding to a site near the ribosomal nascent peptide exit tunnel, which starts at the peptidyl transferase center and spans the body of the large ribosomal subunit, thereby halting translation of a particular subset of nascent peptides [11][12][13].
Resistance to macrolide may be mediated by three primary mechanisms: a) modification of ribosomes, Research Paper www.oncotarget.com such as dimethylation of a unique adenine residue in the 23S ribosomal RNA (rRNA), A2085 in S. aureus (corresponding to E. coli A2058), which is located in the macrolide-binding site in the nascent peptide exit tunnel, by the erythromycin resistance methyltransferase encoded by the erm genes [14][15][16]; b) activated efflux systems, involving a member of the ATP-binding cassette (ABC) family of transporters encoded by the macrolide-streptogramins resistance A (msrA) gene, keeping intracellular antibiotic concentration at a subtoxic level and conferring inducible resistance to erythromycin and type B streptogramins in staphylococci [17,18]; and c) production of antibioticinactivating enzymes, such as phosphorylase, a macrolide phosphotransferase C (encoded by mphC in staphylococci) that inactivates antibiotics [19,20]. According to other studies, mutations in Escherichia coli rplV and rplD genes coding for ribosomal proteins L22 and L4, respectively, can also confer resistance to macrolide antibiotics [21,22]. A mutant change in rplV was also observed in antibiotic-resistant S. aureus [23].
Here, we screened a resistant strain obtained by culturing the sensitive S. aureus strain 8325-4 in the presence of erythromycin. A 27-nt repeat sequence insertion in the rplV (rplV indel ) gene was identified in this erythromycin-resistant strain, which induced specific resistance to macrolides.

8325-4 is screened in vitro
To explore the mechanism underlying the resistance to macrolide in S. aureus, we cultured wildtype S. aureus 8325-4 in BHI medium while continuously doubling the concentration of erythromycin ( Figure 1A). An isolate with acquired resistance to erythromycin was screened and named 8325 ER+ . It was able to grow in a medium containing 80 μg/mL of erythromycin, and the survival rates of 8325 ER+ strain in different concentration of erythromycin were significantly higher than the parent strain ( Figure 1B). Besides, the minimal inhibitory concentration (MIC) of erythromycin of 8325 ER+ was 160 μg/mL in a drug susceptibility test, which was interpreted as erythromycin resistant according to Clinical and Laboratory Standards Institute (CLSI) criteria [24].
In the further investigation, 8325 ER+ was cultured in BHI broth without erythromycin for 20 generations, the susceptibility to erythromycin of bacteria from different generations was determined individually. We found that the MIC of S. aureus from different generations was not altered, which suggested that resistance to erythromycin in 8325 ER+ was inheritable.

Whole-genome sequencing identifies gene mutations in 8325 ER+
To test whether high expression of known erythromycin resistance genes in 8325 ER+ contributed to the resistance to erythromycin, we extracted total RNA from the erythromycin-sensitive strain 8325-4 and the erythromycin-resistant strain 8325 ER+ (Supplementary Figure 1A). Reverse transcription-polymerase chain reaction (RT-PCR) showed that erythromycin resistancerelated genes, including 23 S rRNA adenine-specific N-methyltransferases (encoded by ermA/ermB/ ermC), mphC, and msrA were not detected in 8325 ER+ (Supplementary Figure 1B). These results suggest that another mechanism is responsible for the resistance to erythromycin of 8325 ER+ .
To investigate the potential genes involved in the resistance occurrence to erythromycin, we extracted total genomic DNA and compared the genome sequence of 8325 ER+ with that of 8325-4. Sequence analysis showed that ten genes mutated, and six of them encoded different proteins ( Table 1). Considering that five of the ten mutated genes were involved in the translation process, we decided to compare cell growth between 8325 ER+ and 8325-4. We did not see a significant difference between both strains (Supplementary Figure 1C). Further analysis showed that seven genes were identified with one or two nucleotide mutations, and only one gene, rplV, encoding ribosomal protein L22 exhibited an insertion of a 27-nt fragment (Table 1). In further investigation, we found that sequence of the insertion segment correspond to a duplication of the region 292-318 of rplV ( Figure 2A). Moreover, the insertion of the 27-nt fragment happen at nucleotide 292C or 318C in rplV gene in 8325 ER+ ( Figure 2B). Furthermore, we found that the 27-nt fragment led to a 9-amino acids insertion but did not induce a frame-shifting mutation ( Figure 2C).

Twenty-seven-nucleotide insertion in the rplV (rplV indel ) gene induces resistance to erythromycin in S. aureus
The rplV gene encodes the ribosomal 50S subunit protein L22, which is important for ribosomal 50S subunit assembly at the early stage. It is essential for the formation of the nascent peptide exit tunnel of the mature ribosome [22]. Given that mutation in the rplV gene was reported to be involved in resistance to antibiotics in E. coli and S. aureus [11,[21][22][23], we focused on investigating whether the rplV indel gene induced resistance to erythromycin. Firstly, the rplV genes were amplified by PCR from the genomes of 8325 ER+ and 8325-4. We found the band of PCR product from 8325 ER+ was bigger than that from 8325-4 ( Figure 3A). Following analysis showed that the sequence of rplV gene in 8325 ER+ containing the 27-nt insertion fragment, which was consistent with the genome sequence results ( Figure 3B).
To evaluate the role of rplV indel in raising resistance of S. aureus to erythromycin, we generated several erythromycin-susceptible S. aureus strains (8325-4 indel , RN4220 indel , and Newman indel ) with ectopic expression of rplV indel . Meanwhile, these strains transferred with the wild-type rplV gene were used as control (Supplementary Figure 2). We found that the survival rates of S. aureus strains with ectopic expression of rplV indel in different concentration of erythromycin were significantly higher than control strains ( Figure 3C). And erythromycin MICs in 8325-4 indel , RN4220 indel , and Newman indel were 8 µg/mL respectively (Table 2), which suggests that ectopic expression of rplV indel in susceptible strains induced resistance to erythromycin.

rplV indel contributes to specific resistance to macrolides in S. aureus
Next, we asked whether the rplV indel gene could induce resistance to other macrolides, including azithromycin and clarithromycin. First, we also found the same results with erythromycin susceptibility test. The survival rates of 8325 ER+ strain in different concentration of azithromycin and clarithromycin were significantly (ERY) of wild-type S. aureus. 8325-4 was cultured and passaged in BHI medium supplemented with various concentrations of erythromycin (initially 0.125 μg/mL, followed by two-fold increased until the concentration of erythromycin reached 64 μg/mL). Each screening step included solid and liquid BHI medium for screening. Solid medium was used for picking an isolate of S. aureus at 37° C incubator for 12 h, and liquid medium was used for enrichment of the isolate at 37° C with shaking at 220 rpm. When screening was completed, the erythromycin-resistant isolate was inoculated in BHI medium without antibiotics for 20 generations. (B) Survival rates of 8325-4 and 8325 ER+ in different concentration of erythromycin. The survival curve of wild-type 8325-4 is shown in blue and the 8325 ER+ strain in red. Values are the means of triplicate wells; error bars indicate SD. www.oncotarget.com higher than parent strain ( Figure 4A). We then determined the MIC of azithromycin and clarithromycin in 8325 ER+ as 200 µg/mL and 100 µg/mL, respectively (Table 2). In line with expectation, the ectopic expression of rplV indel in susceptible strains also induced resistance to these two antibiotics ( Figure 4B-4D, Table 2).
As the ribosomal protein L22 is essential in formation of the ribosomal polypeptide exit tunnel [22], we then determined whether rplV indel was involved in resistance to antibiotics targeting the ribosome. It was revealed that neither 8325 ER+ nor susceptible strains with ectopic expression of rplV indel were resistant to chloramphenicol and linezolid (Table 2), which target the 50 S ribosomal subunit. Similar results were also obtained in a drug susceptibility test of antibiotics that target the 30 S subunit or cell wall (Table 2). Consistent with the

DISCUSSION
Macrolides are usually used in clinical therapy for skin infections caused by S. aureus. Several mechanisms involved in S. aureus resistance to macrolides have been revealed. In this study, we revealed that a 27-nt insertion in the rplV gene induced a specific resistance to macrolides.
The 8325 ER+ resistance to macrolides was not due to the occurrence of identified erythromycin-resistant genes (ermA/ermB/ermC/mphC/msrA) but the rplV indel gene. Interestingly, the 27-nt insertion sequence is a repeat sequence of the rplV gene, but it did not induce frameshifting mutation. Sequence analysis showed that this fragment might be inserted behind 291C or 318C. The ectopic expression of the rplV indel gene in several susceptible strains specifically conferred resistance to macrolide antibiotics. As shown in Table 2, the MIC of macrolides was higher in 8325 ER+ than in 8325 indel . We think this may be majorly due to the coexistence of rplV and rplV indel . The ribosome consisting of ribosomal protein L22 (rplV) is still sensitive to macrolides. Mutations in other genes may also contribute the resistance to macrolides in 8325 ER+ . These assumptions will be investigated in the future.
In 1967, bacterial resistance to macrolides, caused by mutations in ribosomal protein, was reported [25]. In E. coli, it has been found that the deletion of M 82 K 83 R 84 increases expression of the AcrAB-TolC efflux system and results in resistance to macrolides [21,26]. Here, we reveal that a 27-nt insertion in the rplV gene confers specific resistance to macrolides in S. aureus. However, the level of msrA, a well-identified gene of the efflux system involved in resistance to macrolides in staphylococci [17,18], was not altered in 8325 ER+ compared with that of 8325-4. The resistance to macrolides induced by rplV indel may be due to the conformational changes of L22 protein induced by the 27-nt insertion, which will be investigated in the future.
Ribosome protein L4 forms part of the lining of the peptide exit tunnel with L22. Mutations in ribosome protein L4 also induce macrolides resistance in a variety of pathogenic and non-pathogenic bacteria [21,[27][28][29]. There are two amino acids mutations (G69A, T70P) identified in the L4 protein from 8325 ER+ , which may also contribute to the macrolides resistance. Although the ectopic expression of the rplV indel gene in several susceptible strains specifically conferred resistance to macrolide antibiotics, combination with the mutant L4 protein may further elevate the resistance. In this study, we tried to detect rplV indel in 84 clinical S. aureus isolates resistant to macrolides. Most of these strains harbor ermA/B/C genes (Supplementary Table 1). There was no strain identified that contained the rplV indel genes (data not shown). These results indicate that the occurrence of rplV indel in clinical isolates is rare compared with that of erm genes.
In conclusion, our findings present a 27-nt insertion in rplV that induces the specific resistance to macrolides in S. aureus.

Bacterial strains, plasmids, and growth conditions
The strains and plasmids used in this study are listed in Supplementary origin of recovery were 22 strains from blood (26.2%), 11 strains from pus (13.1%), 14 strains from secretions (16.7%), 27 strains from sputum (32.1%), 9 strains from wound (10.7%) and one strain from abdominal fluid (1.2%). Strains were cultured using brain heart infusion (BHI) medium (BD) at 37° C for 12 h with shaking at 220 rpm. Clinical isolates and wild-type strains including 8325-4, RN4220 and Newman were cultured in antibiotic-free BHI broth, while 8325 ER+ strain was cultured in BHI broth supplemented with 50 μg/mL erythromycin, those wild-type strains transformed with the shuttle plasmid pOS1 supplemented with 25 μg/ mL chloramphenicol, E. coli strain transformed with a cloning plasmid pMD-19T supplemented with 100 μg/ mL ampicillin.

Erythromycin screen in vitro
Schematic diagram of stepwise screening for resistance to erythromycin of wild-type S. aureus was showed in Figure 1A. 8325-4 was cultured and passaged in BHI medium supplemented with various concentrations of erythromycin (initially 0.125 μg/mL, followed by twofold increased until the concentration of erythromycin reached 64 μg/mL). Each screening step included solid and liquid BHI medium for screening. Solid medium was used for picking an isolate of S. aureus at 37° C incubator for 12 h, and liquid medium was used for enrichment of the isolate at 37° C with shaking at 220 rpm. When screening was completed, the erythromycin-resistant isolate was inoculated in BHI medium without antibiotics for 20 generations.

Measurement of bacterial growth curve
Bacteria were incubated in BHI broth at 37° C with shaking at 220 rpm overnight. The concentration of bacteria was adjusted to 1 × 10 7 cfu/mL, then 1:100 inoculated in BHI broth without antibiotics at 37° C with shaking at 220 rpm for 12 hours. Growth curves of bacteria were constructed by measuring of the cell density at A600 nm at one-hour intervals for 12 hours.

Antibiotic susceptibility assay
Antibiotics were purchased from Selleck. Susceptibility to antibiotics was tested by using broth microdilution method according to the Clinical and Laboratory Standards Institute (CLSI) [24]. Briefly, antibiotics were prepared by serial two-fold dilutions in BHI broth, then various concentration of antibiotics were made in triplicate in 96-well culture dishes containing 1 × 10 5 cfu/well bacteria and incubated for 18-24 h at 37° C. Control wells were free of antibiotic. Bacteria growth was determined by reading the optical density (OD) at 630 nm. The survival rates of bacteria were calculated by the rates of OD 630 nm measurement at each concentration of antibiotic versus control wells. The MIC was determined to be the dose of antibiotic that inhibited bacteria growth by >95%.

RNA isolation and RT-PCR
For detecting erythromycin resistance genes in S. aureus. Total bacterial RNA was extracted from S. aureus, which were grown with shaking at 37° C using Trizol (Invitrogen) as previously described [30]. Briefly, DNase digestion of 80 μL of total RNA was performed with 10U of RNase-free DNase I (Promega) and 10 μL of the 10 × reaction buffers in a total reaction volume of 100 μL for 30 min at 37° C. For cDNA synthesis, 6 μL total RNA (≈250 ng) was incubated at 65° C for 5 min, then add 2 μL of 4 × DNA remove buffer and incubate at 37° C for 5 min, finally add 2 μL of 5 × RT Master MixII (TOYOBO) and incubate at 37° C for 15 min, 50° C for 5 min, 98° C for 5 min.

Detection of macrolides-resistance genes
Macrolides resistance genes ermA, ermB, ermC, msrA and mphC were examined in the erythromycinsensitive strain 8325-4 and the erythromycin-resistant strain 8325 ER+ with primers listed in Supplementary  Table 3. The PCR reaction mixture contained 2.5 μL of 10 × PCR reaction buffer, 0.25 μL enzyme, 0.5 μL dNTP mix, 0.3 mM of gene-specific forward and reverse primers, and 2 μL of template, made up to a final volume of 25 μL with distilled water. Cycling parameters were set as follows: initial activation step at 95° C for 5 min, denaturation at 95° C for 30 s, annealing at 55° C for 30 s, and extension at 72° C for 30 s. gyrB was used as the endogenous reference gene. The PCR products were resolved in 2% agarose gel and visualized by ultraviolet imaging.

Whole-genome sequencing of S. aureus
Bacteria were grown in BHI broth at 37° C for 12 h with shaking at 220 rpm and harvested by centrifuge at 12000 rpm for 1 min. Genomic DNA was extracted by using EasyPure ® Bacteria Genomic DNA Kit (TransGen

Analysis of the rplV gene in clinical isolates
For detecting 27-nt fragment insertion in rplV gene in clinical macrolides resistant S. aureus isolates, the rplV indel gene was detected by PCR amplification. Primers used for the rplV gene were rplV-F and rplV-R listed in Supplementary Table 3. Clinical isolates were grown on blood agar plates and incubated overnight at 37° C, bacteria DNA was prepared by suspending a fresh colony in 400 μL of sterile distilled water and heating at 100° C for 10 min and then centrifuged at 12000 rpm for 5 min. PCR amplification was carried out under the following conditions: 95° C for 5 min, followed by 34 cycles of 95° C for 30 s, 56° C for 30 s, 72° C for 30 s, and 72° C for 5 min. PCR products were resolved on a 2% agarose gel and visualized by ultraviolet imaging. The nucleotides were sequenced by Sangon Biotech, and BLAST analysis was performed using DNAMAN.
Ectopic expression of rplV indel in S. aureus strains rplV genes were amplified by PCR from the genomes of wild-type S. aureus 8325-4 and the erythromycin-resistant strain 8325 ER+ with primers rplV-F-EcoRI and rplV-R-BamHI (Supplementary Table 3). The PCR products were ligated into pMD-19T vector, the recombinant plasmids were transformed into DH5ɑ. The recombinant pMD-19T plasmid was eliminated by cutting the plasmid with the EcoRI and BamHI restriction enzymes, then digested fragments were ligated into EcoRI and BamHI-digested pOS1 vector. The recombinant plasmids were transformed into DH5ɑ, then electrotransformation into recipient strains S. aureus RN4220. The plasmid was isolated from RN4220, then electrotransformation into S. aureus 8325-4 and Newman.

Statistical analyses
Statistical tests were performed using GraphPad Prism v.5.0 (GraphPad Software Inc., San Diego, CA, United States). The Differences between survival curves were evaluated for statistical significance using the unpaired t test. All P-values of ≤ 0.05 was considered significant.