Omadacycline Efficacy against Enterococcus faecalis Isolated in China: In Vitro Activity, Heteroresistance, and Resistance Mechanisms

This study aimed to evaluate the in vitro antimicrobial activity, heteroresistance emergence, and resistance mechanism of omadacycline (OMC) in clinical Enterococcus faecalis isolates from China. A total of 276 isolates were collected retrospectively in China from 2011 to 2015. The MICs of OMC, doxycycline (DOX), and minocycline (MIN) against E. faecalis were determined by broth microdilution. Tetracycline (TET)-specific resistance genes and multilocus sequence typing (MLST) of the isolates were investigated using PCR.

through natural or acquired resistance mechanisms (4,5). Recently, several reports have shown that the increased incidence of multidrug-resistant enterococci, including vancomycin (VAN)-resistant and linezolid (LZD)-resistant strains, has limited our treatment choices, and controlling multidrug-resistant E. faecalis infections has become a critical need in clinics (4)(5)(6).
Omadacycline (OMC) has been recently developed and is a first-in-class aminomethylcycline antibiotic with broad-spectrum activity against Gram-positive and Gramnegative aerobic, anaerobic, and atypical pathogens, including Staphylococcus, Enterococcus, Legionella, and Chlamydia species (7)(8)(9)(10). Oral and intravenous OMC formulas have been evaluated in phase III clinical trials and have shown excellent efficacy in the treatment of acute skin and soft tissue infections and community-acquired pneumonia (10)(11)(12)(13). OMC differs from earlier tetracycline (TET) derivatives-including doxycycline (DOX), minocycline (MIN), and expanded-spectrum glycylcycline antibiotics such as tigecycline (TGC)-owing to two major structural modifications. Recent reports have demonstrated the in vitro active antibacterial potency of OMC with MICs of Յ0.25 mg/ liter against various Gram-positive microbes, including enterococci, methicillin-resistant Staphylococcus aureus, and multidrug-resistant Streptococcus pneumoniae (12)(13)(14). However, the antimicrobial activity of OMC against clinical isolates of E. faecalis from China has not been established.
Both OMC and TGC represent new derivatives of the TET class of antimicrobial drugs. These drugs are regarded as last-resort antimicrobial treatments for infections by difficult-to-treat bacteria, such as multidrug-resistant E. faecalis. TGC resistance has been linked to genetic mutations that affect the 30S ribosomal subunit of TET binding sites-including mutations that affect the genes encoding 16SrRNA (four copies) and ribosomal protein S10 -as well as the overexpression of genes that encode efflux pump regulators, such as soxS, marA, ramA, and robA (15)(16)(17)(18). Moreover, overexpression of tet(M) and tet (K) has been shown to increase TGC MIC values and favor the evolution of resistance in Enterococcus faecium (19).
Heteroresistance, which refers to the presence of subpopulations of bacterial cells with higher levels of antibiotic resistance than those of the surrounding populations in the same culture, can lead to the evolution of antibiotic resistance and treatment failure (20). Subpopulation analysis of OMC heteroresistance in E. faecalis can provide information about the risk of antibacterial resistance under antibiotic pressure. The incidence of OMC heteroresistance and its mechanisms in E. faecalis remain unclear. The extent to which TET/TGC resistance mechanisms may contribute to OMC heteroresistance and resistance evolution needs to be determined.
In the present study, we examined the in vitro antimicrobial activity of OMC against clinical E. faecalis isolates collected from patients in China. The clonality and OMC susceptibility with respect to the sequence type (ST) were analyzed. Furthermore, population analysis profiling (PAP), molecular sequencing techniques, and in vitro functional tests were performed to explore the incidence and underlying mechanism of OMC heteroresistance in clinical E. faecalis isolates. Particular attention was given to mutations that affect the 30S ribosome unit in E. faecalis strains with OMC-induced resistance.

RESULTS
In vitro antimicrobial activity of OMC against clinical E. faecalis isolates. The clinical E. faecalis strains were isolated from various infective sample sources, including urine (48.6%), wound secretions (17.0%), blood (11.2%), and bile (7.3%), among others (Fig. S1). Moreover, the OMC MIC data of those isolates were obtained and are summarized in Table 1 Table 1). The distribution of TET-specific resistance genes in clinical E. faecalis isolates is shown in Table 2. The OMC MIC 50/90 values of 0.5/1 mg/liter were obtained in E. faecalis isolates with at least one of the TET-specific resistance genes, which included tet(M), tet(K), and tet(L), suggesting that the presence of these genes did not affect OMC sensitivity in E. faecalis. In contrast, DOX and MIN MICs in E. faecalis isolates harboring tet(M), tet(K), or tet(L) genes were 16-to 32-fold higher than those of strains without these genes. Four TET-specific resistance genes, namely tet(O), tet(S), tet(W), and tet(U), were not found in any E. faecalis isolates (  Table 4). The OMC MICs of resistant subpopulations were decreased by 4-to 8-fold after 10 passages on antibiotic-free medium, which suggested that these subpopulations lacked stable mutations conferring OMC resistance and could be reversed to the susceptible phenotype after the removal of antibiotic pressure (Table S1). Moreover, the OMC MICs for these resistant subpopulations could be significantly reduced to Յ0.03 mg/liter and 0.5 mg/liter with the addition of the efflux pump inhibitors (EPIs) carbonyl cyanide m-chlorophenylhydrazone (CCCP) and Phe-Arg-␤-naphthylamide (Pa␤N), respectively. This indicates that EPIs could potentiate OMC activity in E. faecalis ( Table 4).
Differentially expressed genes between the heteroresistant parental isolate and its resistant subpopulation. Because the heteroresistance mechanism in this study was elusive, we performed RNA sequencing (RNA-Seq) to compare the transcription of unique genes between the heteroresistant parental strain EF16C185 (EF16) and its resistant subpopulation EF16C185-RS (EF16-O2). Before RNA sequencing, the growth curves of EF16 and EF16-O2 were determined, and the growth rates of these two strains  Omadacycline Resistance in E. faecalis Antimicrobial Agents and Chemotherapy showed no significant difference (Fig. S3). A total of 693 differentially expressed genes (DEGs), including 378 upregulated and 315 downregulated genes, were identified in EF16-O2 compared to EF16 in RNA sequencing analysis ( Fig. 1a and Tables S2 and S3).
The OMC MICs in resistant subpopulations of heteroresistant isolates could be reversed to sensitive levels by adding an EPI, supporting the notion that membrane proteins or efflux pumps may participate in the initial upregulation of OMC MIC values during the progression of heteroresistance. Twelve upregulated DEGs, which showed potential associations with heteroresistance, were validated by quantitative real-time (qRT-PCR) in EF16 and EF16-O2 (Table 5). To further validate their impact on OMC susceptibility, the expression levels of those 12 DEGs were investigated in the remaining five heteroresistant parental isolates, their resistant subpopulations, and strains with OMC-induced resistance ( Fig. 2 and Fig. S4). Overexpression of six candidate genes, namely, OG1RF_RS00630, OG1RF_RS12140, OG1RF_RS02205, OG1RF_RS06145,  OG1RF_RS06880, and OG1RF_RS11485, was correlated with the occurrence of OMC heteroresistance or resistance (Fig. 2). Overexpression of OG1RF_RS00630 in E. faecalis contributes to the occurrence of OMC heteroresistance. To confirm the roles of the candidate genes (OG1RF_RS12140, OG1RF_RS02205, OG1RF_RS06145, OG1RF_RS11485, OG1RF_RS06880, and OG1RF_ RS00630) in OMC heteroresistance and resistance in E. faecalis, the overexpression of these genes in OMC-sensitive isolates was conducted. The overexpression plasmids pRS12140, pRS02205, pRS06145, pRS11485, pRS06880, and pRS00630 were constructed and transformed into OMC-sensitive isolates (three per gene), showing low expression level of the target gene (Table S4 and Table S5). Stable overexpression of the candidate  genes in the transformed strains was confirmed by qRT-PCR (Fig. S5). Overexpression of the six candidate genes did not induce OMC MIC elevation in the absence of antibiotic pressure. However, PAP showed that overexpression of OG1RF_RS00630 led to the occurrence of the OMC-heteroresistant phenotype in all three OMC-sensitive E. faecalis isolates, compared with negative findings in the parental isolate controls (Table 6). Phylogenetic and homology analyses showed that E. faecalis OG1RF_RS00630 encodes a bone morphogenetic protein (BMP) family ATP-binding cassette (ABC) transporter substrate-binding protein that is ubiquitous in a variety of bacterial species and shares high identity (over 60%) with those of other Gram-positive bacteria ( Fig. S6 and Table  S6). Our data indicate that the product of this gene may contribute to OMC heteroresistance in E. faecalis.

DISCUSSION
In the present study, OMC exhibited excellent in vitro antimicrobial activity against LZD-nonsusceptible E. faecalis, indicating potential application of OMC in the treatment of multidrug-resistant E. faecalis infections. E. faecalis maintained low OMC MICs in TET-resistant strains even as DOX and MIN MICs increased. The present findings support the efficacy of OMC against clinically infective E. faecalis in China (17,18). The MIC 50/90 of OMC in 276 E. faecalis isolates was 0.5/1.0 mg/liter in the present study, which was higher than that reported in previous studies (Յ0.125/0.125 mg/liter) (7,9,13). This difference suggests that susceptibility of E. faecalis to OMC varies across geographical regions, which may lead to regional differences in clinical perspectives. The main STs observed in our sample were ST16 and ST179, which is consistent with prior reports (21,22). Our data indicate the clonality of E. faecalis, with clustering of the 1 mg/liter OMC MIC level in the ST16 genotype. This underscores the risk of nosocomial transmission of E. faecalis strains with high OMC MICs.
The most common TET resistance mechanisms in Gram-positive and Gram-negative pathogens include ribosomal protection and efflux pump proteins. Notably, the dissemination of these TET-specific resistance factors has limited the clinical utility of TET and its derivatives, including DOX and MIN (17,18). Antimicrobial susceptibility of new TET drugs, such as TGC and eravacycline, might be slightly influenced by these two mechanisms. The overexpression of some TET-specific resistance genes, including the genes that encode ribosomal protection and efflux pump proteins, have been reported to reduce TGC susceptibility (19). Our data showed that the ribosomal protection protein gene, tet(M), and the efflux pump genes, tet(K) and tet(L), did not affect OMC susceptibility. This finding indicates that OMC might have overcome these two common resistance mechanisms. Although OMC retained low MIC values against clinical E. faecalis isolates in this study, monitoring the incidence of heteroresistance can provide information about the potential risk of OMC resistance in the future (23,24). Six OMC-heteroresistant isolates were detected among 238 clinical E. faecalis isolates with MIC values of Յ0.5 mg/liter according to PAPs. The OMC and TGC MIC values were elevated synchronously in OMC-resistant subpopulations, pointing to a cross-resistance risk for these two antibiotics. Therefore, recognition of the development of OMC resistance in E. faecalis may be facilitated by monitoring OMC heteroresistance in clinical isolates. Furthermore, the OMC MICs of resistant subpopulations could be reversed after 10 passages on antibiotic-free medium, while those of isolates with OMC-induced resistance remained stable. This indicates that the resistant subpopulations were reversible and not the result of a stable mutation, which is consistent with the findings of previous research (25).
X-ray crystallography of TET in complex with the 30S ribosomal subunit of Thermus thermophilus revealed two TET binding sites on the 30S subunit (26,27). As TGC and OMC are new TET derivatives, they are thought to target similar binding sites as TET. However, competition studies with radiolabeled TET indicate that OMC can inhibit in vitro translation at half the concentration required for TET (17), suggesting that OMC may have a 2-fold stronger affinity than TET for the ribosome. Mutations affecting TET binding sites of the 30S ribosomal subunit have been shown to confer TET and TGC resistance in several bacterial species (15)(16)(17). The mutational characteristics of OMCinduced resistance in E. faecalis in this study were consistent with the prior identification of various target-interaction related mutations in individual copies of 16S rRNA in bacteria with TGC-induced resistance (16). We detected mutations in all four 16S rRNA gene copies in E. faecalis under OMC pressure. Furthermore, strains having a greater number of genetic mutations in 16S rRNA copies tended to have greater OMC and TGC resistance. Mutations in the gene encoding the 30S ribosomal protein S10 occurred at a relatively low frequency (2/10; 20%) in our strains with OMC-induced resistance, compared to prior findings in multiple bacterial species under TGC pressure (15), indicating that the role of S10 in the evolution of OMC resistance needs to be further studied.
The lack of genetic mutations in 30S ribosomal subunits of OMC-heteroresistant parental isolates and their resistant subpopulations indicates that the MIC elevation observed with the emergence of OMC heteroresistance cannot be explained by such mutations. The TGC resistance and heteroresistance in several Gram-negative bacterial species have been linked to the overexpression of several efflux pump proteins (e.g., SoxS, MarA, RamA, and RobA) (28,29). Our data showed that the addition of an EPI could restore sensitivity to OMC in resistant subpopulations of OMC-heteroresistant isolates, indicating that efflux pumps may be involved in OMC heteroresistance under antibiotic pressure. Following both RNA sequencing and qRT-PCR analyses, 12 upregulated DEGs encoding efflux pump proteins or membrane proteins were observed in the resistant subpopulations compared with the parental strains. Furthermore, in vitro recombination experiments indicated that overexpression of the candidate gene OG1RF_RS00630 favored OMC heteroresistance in E. faecalis but did not affect OMC and TGC MICs in the absence of antibiotic pressure. Phylogenetic and homology analyses indicated that the efflux protein encoded by OG1RF_RS00630 is an ATPbinding cassette family protein expressed mainly in the cellular membrane. A previous study showed that the ortholog of OG1RF_RS00630 (EF0177) from E. faecalis V583 is involved in ABC transporter-mediated ribonucleoside uptake (30). Our data demonstrate for the first time that OG1RF_RS00630 participates in the development of OMC heteroresistance in E. faecalis.
In conclusion, OMC exhibited robust antimicrobial effects against clinical E. faecalis isolates from China, with lower MIC 50/90 values than those of DOX or MIN. Furthermore, OMC showed excellent in vitro antimicrobial activity to LZD-nonsusceptible clinical E. faecalis isolates. Isolates with OMC MICs of 1 mg/liter showed ST16 clonality. Mutations in 30S ribosome units were associated with OMC resistance, and overexpression of a BMP family ABC transporter substrate-binding protein (OG1RF_RS00630) in E. faecalis appeared to favor the occurrence of OMC heteroresistance. Moreover, the emergence of heteroresistance among clinical E. faecalis isolates in China, especially those with high OMC MICs, is noteworthy as a possible harbinger of future resistance development.

MATERIALS AND METHODS
Bacterial isolates, growth conditions, and chemicals. A total of 276 nonduplicate clinical E. faecalis strains were collected retrospectively from inpatients at Shenzhen Nanshan People's Hospital (a tertiarycare teaching center hospital in China with 1,200 beds) from 1 January 2011 to 31 December 2015. Bacterial species were identified by standard methods using a Vitek 2 compact system (bioMérieux, Marcy l'Etoile, France). E. faecalis ATCC 29212 was used as the quality control strain. All procedures performed were approved by the ethical standards of Shenzhen Nanshan People's Hospital. The E. faecalis strains were cultured in tryptic soy broth (TSB; Oxoid, Basingstoke, UK) at 37°C with shaking at 220 rpm. The aminomethylcycline antibiotic OMC was purchased from MedChem Express (Princeton, NJ). DOX, MIN, and TGC were purchased from Aladdin (Shanghai, China).
Population analysis profiles. Population analysis profiling (PAP) was used as a reference method to investigate OMC heteroresistance among 238 clinical E. faecalis isolates with OMC MIC values of Յ0.5 mg/liter as described previously (23). Briefly, 50-l aliquots of cell suspension (corresponding to a 0.5 McFarland standard for Enterococcus cultures grown on blood agar plates for 24 h at 37°C; approximately 1 ϫ 10 8 CFU/ml) were spread onto Mueller-Hinton agar plates, with or without various concentrations of OMC (0.5, 1, 2, 3, 4, and 5 mg/liter). Plates were then incubated at 37°C and colonies were counted after 24 h. According to the FDA breakpoint of OMC MIC values for E. faecalis, OMC heteroresistance was defined as an OMC-susceptible isolate (MIC Յ0.5 mg/liter) with subpopulations growing in the presence of Ն1 mg/liter OMC, with a detection limit of 20 CFU/ml. Three colonies, categorized as the resistant subpopulations of each OMC-heteroresistant isolate, were selected from the 1 mg/liter OMC concentration of the PAP test, and the OMC MICs were reassessed after serial passaging on antibiotic-free medium to evaluate the stability of the heteroresistant phenotype. The OMC-resistant subpopulations or their parental strains were cultured in Mueller-Hinton broth (MHB) either supplemented with 1 mg/liter OMC or without OMC, respectively, for the subsequent experiments.
In vitro induction of OMC-resistance in E. faecalis under OMC pressure. Six OMC-heteroresistant parental isolates and the OG1RF strain were used to induce OMC-resistant isolates. These isolates were subcultured serially in MHB containing gradually increasing concentrations of OMC, with the initial concentration being MIC values followed by successive increases to 2ϫ, 4ϫ, 8ϫ, and 16ϫ MIC (31). Strains were cultured for four passages before their entry into the next concentration. Isolates from the passages of each concentration were stored at Ϫ80°C in MHB containing 40% glycerol, until further determination of genetic mutations and subsequent MIC assays.
PCR for multilocus sequence typing and detection of TET resistance and 30S ribosomal subunit genes. Total DNA samples were extracted from E. faecalis isolates in lysis buffer for microorganisms and submitted to direct PCR using the PCR mastermix (Thermo Fisher Scientific, Waltham, MA) according to the manufacturer's instructions. The multilocus sequence typing (MLST) was determined by PCR and sequence alignment as previously described (21,32). We also used PCR to detect TET-specific resistance genes, including tet(K) and tet(L), which encode efflux pump proteins; tet(M), tet(S), tet(O), and tet(W), which encode ribosomal protection proteins; and tet(U), a putative TET resistance determinant, as previously described (22). Genetic mutations in 30S ribosomal subunits, including four 16S rRNA gene copies and the 30S ribosomal protein S10, were detected by PCR amplification and sequence alignment. The primers used in this study are listed in Tables S7 to S9.
Efflux pump inhibition. The role of the efflux pump in the development of OMC heteroresistance was evaluated using the EPIs Phe-Arg-␤-naphthylamide (Pa␤N; Sigma, St. Louis, MO) and carbonyl cyanide m-chlorophenylhydrazone (CCCP; Sigma). The OMC MICs of the resistant subpopulations of heteroresistant isolates were determined by agar dilution in the presence and absence of PA␤N (50 mg/liter) or CCCP (50 mg/liter), as previously described (28). Efflux pump inhibition of antimicrobial susceptibility was considered significant if the magnitude of the MIC value was decreased at least 4-fold in the presence of EPIs, which is consistent with the methods of previous investigations (24).
Measurement of bacterial growth curves. Overnight bacterial cultures of EF16C185 (EF16) or EF16C185-RS (EF16-O2) were diluted 1:200 into 1 ml of fresh MHB or into MHB containing 1 mg/liter OMC, respectively, of which 300 l was added into each well of a 96-well plate. Three parallel wells were used for each sample. The plates were placed in a Bioscreen C MBR (Oy Growth Curves Ab Ltd., Helsinki, Finland), and the bacteria were grown at 37°C with shaking at 220 rpm. Growth curves of the strains were determined by measuring the optical density at 600 nm (OD 600 ) at 30-min intervals over a period of 16 h. RNA sequencing. The heteroresistant parental strain EF16 or its resistant subpopulation EF16-O2 was grown overnight in 20 ml of antibiotic-free MHB or MHB containing 1 mg/liter OMC, respectively. The overnight cultures were diluted 1:100 into 50 ml of MHB under the same conditions and grown to the mid-log phase (4 h) at 37°C. The bacteria were harvested by centrifugation, and total RNA of EF16 and EF16-O2 was isolated using an RNeasy minikit (Qiagen, Hilden, Germany) according to the manufacturer's instructions (three biological replicates for each strain). The RNA quality and quantity were determined by 1.0% formaldehyde denaturing agarose gel electrophoresis and spectrophotometry in a NanoDrop ND-1000 machine, respectively. RNA sequencing was performed as previously described (33). Briefly, rRNA was removed from total RNA with a Ribo-Zero rRNA removal kit for Gram-positive bacteria (Illumina, San Diego, CA). Fragmented RNA was used as a template for PCR with random primers. To build the cDNA libraries, cDNA fragments were purified using an AMPure XP system (Beckman Coulter, Beverly, USA) to select fragments of ϳ150 to 200 bp in length. The PCR was then performed using Phusion high-fidelity DNA polymerase (Thermo Fisher Scientific, Waltham, MA), and the library quality was assessed on an Agilent Bioanalyzer 2100 system (Agilent Technologies, CA). The library preparations were then sequenced on an Illumina HiSeq X Ten platform, and reads of paired-end 150 bp (PE150) were generated. Clean data were obtained by removing rRNA reads, sequencing adapters, short fragment reads, and other low-quality reads from the raw data. The remaining reads were mapped to the reference genome of OG1RF on the NCBI website with Bowtie 2 software. When the reads were aligned, one mismatch with the reference sequence was allowed. The alignments were further processed using BEDTools software to determine the transcript expression levels and their differential expression between each two of the three samples. Differential expression of all transcripts was quantified using DEGseq software, and the fold change values were recorded. Genes with adjusted P values (Benjamini-Hochberg method) of Ͻ0.05 and at least a 2-fold difference in expression were considered DEGs.
Quantitative real-time PCR analysis. Transcriptional levels of OMC heteroresistant candidate genes were determined by qRT-PCR with the primers listed in Table S10, according to previously described methods (24). Briefly, overnight cultures of the bacterial strains were diluted 1:100 into 10 ml of MHB (cultures of the resistant subpopulations supplemented with 1 mg/liter OMC) and grown for 4 h at 37°C. Total bacterial RNA was extracted using an RNeasy minikit (Qiagen, Hilden, Germany), and the extracted RNA was reverse transcribed into cDNA using a PrimeScript real-time (RT) reagent kit (TaKaRa Bio, Inc., Shiga, Japan). qRT-PCR was performed with the SYBR Premix Ex Taq II kit (TaKaRa Bio, Inc., Shiga, Japan) in a Mastercycler ep realplex system (Eppendorf, Hamburg, Germany). The internal control gene recA was used to normalize the expression of each candidate gene. The threshold cycle (C T ) numbers were determined by the detection system software, and the data were analyzed based on the 2 ϪΔΔCT method. Expression levels of the target genes were compared with those of the E. faecalis OG1RF strain or the heteroresistant parental strain (expression ϭ 1). All qRT-PCRs were carried out in triplicate with three independent RNA samples.
Gene overexpression. Full-length candidate genes, which included OG1RF_RS12140, OG1RF_ RS02205, OG1RF_RS06145, OG1RF_RS11485, OG1RF_RS06880, and OG1RF_RS00630, were amplified from the total DNA extracted from OG1RF isolates and then integrated separately into the pIB166 vector (the inserted gene was controlled by a P23 promoter for overexpression) (34). The positive clones were screened by chloramphenicol and verified by PCR and sequencing. The overexpression plasmids (pRS12140, pRS02205, pRS06145, pRS11485, pRS06880, and pRS00630) were then transformed separately into three OMC-sensitive isolates (Table S5) and further identified by PCR and sequencing. The pIB166 vector was transferred into the same isolates as a control. Subsequently, overnight cultures of the derivative and parental strains were diluted 1:100 into antibiotic-free MHB and grown for 4 h at 37°C, and transcriptional levels of candidate genes were measured by qRT-PCR as described above. The OMC MIC values for these derivatives were determined, and heteroresistance was evaluated by PAP determination under OMC pressure as described above. The primers for plasmid construction are listed in Table S11.
Statistical analysis. Continuous data were analyzed with the Student's t test and one-way factorial analysis of variance (ANOVAs) in the SPSS software package (version 17.0; Chicago, IL). P values of Ͻ0.05 were regarded as statistically significant.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 1.6 MB.