Overexpression of OqxAB and MacAB efflux pumps contributes to eravacycline resistance and heteroresistance in clinical isolates of Klebsiella pneumoniae

This study investigated the characteristics and mechanisms of eravacycline resistance and heteroresistance in clinical Klebsiella pneumoniae isolates. A total of 393 clinical K. pneumoniae isolates were collected and subjected to eravacycline and tigecycline MIC determinations using the agar dilution method. Eravacycline heteroresistance was assessed by a population analysis profile (PAP). The expression levels of efflux pumps and their regulators were determined by quantitative reverse-transcription PCR (qRT-PCR). This study identified 67 eravacycline-nonsusceptible isolates; among the extended-spectrum β-lactamase (ESBL)-positive isolates, eravacycline-nonsusceptible isolates were detected more frequently than tigecycline-nonsusceptible isolates (21.7% vs. 9.4%, p = 0.001). The study sample was observed to include 20 K. pneumoniae isolates with eravacycline heteroresistance. Compared to the reference strain, oqxA or oqxB overexpression was observed in nine eravacycline-nonsusceptible isolates (range, 35.64–309.02-fold) and 13 eravacycline-heteroresistant isolates (8.42–296.34-fold). The overexpression of macA or macB was detected in 12 eravacycline-heteroresistant isolates (3.23–28.35-fold). Overexpression of the efflux pump regulator gene ramA was observed in 11 eravacycline-nonsusceptible isolates (3.33–94.05-fold) and 18 eravacycline-heteroresistant isolates (3.89–571.70-fold). The eravacycline MICs were increased by one–fourfold by overexpression of oqxAB or macAB in three eravacycline-sensitive isolates. In conclusion, the overexpression of OqxAB and MacAB efflux pumps and the transcriptional regulator RamA were suggested to be involved in K. pneumoniae eravacycline resistance and heteroresistance.


Introduction
Concerns regarding the Gram-negative pathogen Klebsiella pneumoniae, a member of the Enterobacteriaceae family, are growing worldwide due to the increasing incidence of severe infections, antibiotic-resistant strains, and reduced treatment efficacy 1 . Carbapenem-resistant Enterobacteriaceae are an emergent global health threat because carbapenems had previously been effective for eliminating multidrug-resistant Gram-negative bacterial infections 2 . In particular, increases in carbapenemresistant K. pneumoniae (CR-Kp) frequencies worldwide are resulting in K. pneumoniae infections that are very difficult to treat and are thus associated with higher mortality rates 3 .
Tigecycline, the original member of the glycylcycline group of antibiotics, has been shown to have antimicrobial activity against CR-Kp in vitro, and thus, this antibiotic may be a last resort therapeutic option against CR-Kp infections 4 . However, cases of tigecycline-nonsusceptible K. pneumoniae (TNSKP) have emerged in hospitals with wide clinical application of tigecycline [5][6][7] . In recent years, TNSKP has been reported to occur in patients without prior exposure to tigecycline [8][9][10] . The mechanisms underlying tigecycline resistance are complex and not yet well understood. The overexpression of the efflux pumps AcrAB and OqxAB has been shown to play a crucial role in tigecycline resistance in K. pneumoniae 9,11 . Meanwhile, mutations in ramR, acrR, and rpsJ genes have also been reported to contribute to K. pneumoniae resistance to tigecycline 9,12 .
Eravacycline, previously known as TP-434, is a novel fluorocycline antibiotic with broad-spectrum activity against Gram-positive and Gram-negative aerobic and anaerobic pathogens in vitro 13 . Especially noteworthy is the observation that eravacycline has efficacy against several critical antimicrobial-resistant pathogens, such as methicillin-resistant Staphylococcus aureus, vancomycinresistant enterococci, extended-spectrum β-lactamase (ESBL), carbapenemase-producing Enterobacteriaceae, and multidrug-resistant Acinetobacter baumannii [14][15][16] . Indeed, eravacycline has been reported to be two-fourfold more effective than tigecycline against common clinical Gram-positive and Gram-negative aerobic bacterial species 13 .
Like tigecycline, eravacycline can overcome most prevalent tetracycline-resistant mechanisms. Notably, Grossman et al. found that eravacycline efficacy was only slightly or undetectably affected by common tetracycline resistance factors, including efflux pumps (tetA, tetB, and tetK) and ribosomal protection protein (tetM) variants in Escherichia coli. Meanwhile, the antibacterial potencies of tigecycline and eravacycline were reduced by 4-to 16-fold in two nonisogenic Propionibacterium acnes isolates harboring a 16S rRNA gene (G1058C) mutation compared with that of the wild-type control strain 15 . The MIC for eravacycline in adeB-hyperexpressing A. baumannii was shown to be reduced by eightfold by disrupting the gene adeB 17 . However, the traits and mechanisms of eravacycline resistance among clinical K. pneumoniae isolates, especially TNSKP isolates, have yet to be clarified. Thus, the aim of the present study was to explore the characteristics and mechanisms of eravacycline resistance among clinical K. pneumoniae isolates.
Effects of efflux pump inhibitor (EPI) and ribosomal protein gene mutations on K. pneumoniae MICs To investigate the mechanisms of eravacycline resistance in K. pneumonia, 37 clinical K. pneumoniae isolates were selected for further study, including the tigecyclineor eravacycline-susceptible or nonsusceptible isolates. To explore the differences between eravacycline and tigecycline resistance, these 37 isolates were divided into three groups, those that had MICs of tigecycline that were <, =, or > the MICs of eravacycline. The effects of the EPI Phe-Arg-β-naphthylamide (PAβN) on tigecycline and eravacycline MICs are reported in Table 2. Notably, among 25 tigecycline-and/or eravacycline-nonsusceptible isolates (MIC ≥4 mg/L), six isolates showed a 16-fold decrease, 10 showed an eightfold decrease, eight showed a fourfold decrease, and one isolate showed a twofold decrease in tigecycline and/or eravacycline MICs in the presence of PAβN (50 mg/L). However, among 12 tigecycline-and eravacycline-susceptible isolates (MIC ≤ 2 mg/L), three showed a fourfold decrease and six showed a twofold decrease in tigecycline and/or eravacycline MICs in the presence of PAβN (50 mg/L).
Mutations in acrR were observed in seven of the above 37 clinical K. pneumoniae isolates and in five of the 25 tigecycline-and/or eravacycline-nonsusceptible isolates ( Table 2). Mutations in ramR were detected in five of these 37 clinical K. pneumoniae isolates and in three of the 25 tigecycline-and/or eravacycline-nonsusceptible isolates. Three isolates were observed with both ramR and acrR mutations, two of which were tigecycline-and/or eravacycline-nonsusceptible and had MICs comparable to other nonsusceptible isolates. None of the isolates in this study had rpsJ mutations. Furthermore, we detected these ribosomal protein gene mutations in 47 tigecycline-and eravacycline-sensitive clinical K. pneumonia isolates. Mutations in acrR or ramR were also observed in these eravacycline-sensitive isolates, but were only detected in four isolates and were different from those mutations detected in eravacyclinenonsusceptible isolates (Table S1).

Mutations in OqxAB and MacAB efflux pumps
Mutations in oqxAB and macAB genes in the assayed clinical K. pneumoniae isolates overexpressing oqxAB or macAB (overexpressed >threefold greater than the ATCC 13883 reference strain) were identified. Mutations in OqxAB and MacAB efflux pumps were rarely reported from previous studies. In this study, only two eravacycline-heteroresistant isolates with oqxAB mutations (EKP82-2 and EKP119-1) and one isolate with a macB mutation (EKP82-2) were identified, which were not present among those eravacycline-resistant isolates (Table S2 and S3).
OqxAB and MacAB efflux pumps associated with eravacycline resistance were tested in an overexpression experiment To confirm the roles of OqxAB and MacAB efflux pumps in eravacycline resistance of K. pneumoniae, the overexpression of oqxAB and macAB in three eravacycline-sensitive isolates was conducted. As Fig. 4 shows the expression levels of oqxA increased 6.77-7.63fold and that of macA increased 4.76-5.65-fold, following a 0.2% arabinose (Ara) induction. Interestingly, the eravacycline MICs of these three eravacycline-sensitive isolates increased one-fourfold following the 0.2% Ara induction, relative to controls not exposed to Ara (Table 4).

Discussion
In previous studies, researchers observed that eravacycline had a more powerful antibacterial effect than tigecycline against clinically common aerobic bacterial species 13,15 . However, the results of this study showed that eravacycline-nonsusceptible clinical K. pneumoniae isolates were encountered more frequently than tigecyclinenonsusceptible isolates, especially among ESBL-positive isolates. The occurrence of antibiotic resistance varies regionally. Until now, there have been no reports of eravacycline susceptibility in K. pneumoniae from China. We believe that the K. pneumoniae isolates from different regions provide an explanation for different results, but this issue needs additional study in China for confirmation. In the present study, the overexpression of several antibiotic resistance-related genes, including genes encoding proteins that constitute the resistancenodulation-cell division (RND)-type efflux pump AcrAB, the quinolone/olaquindox efflux pump OqxAB, and the transcriptional activators RamA and SoxS, was observed among the examined eravacycline-nonsusceptible isolates. Imipenem, meropenem, and colistin heteroresistances have been documented previously in carbapenemaseproducing isolates of K. pneumoniae [18][19][20][21] . To the best of our knowledge, this is the first study to clinically document K. pneumoniae isolates with eravacycline heteroresistance.
The RND-type efflux pump, AcrAB-TolC, has been shown to play an important role in Enterobacteriaceae antimicrobial resistance, especially tigecycline resistance in K. pneumoniae 4,9,22,23 . Quinolone/olaquindox efflux pump, OqxAB, was also associated recently with tigecycline resistance in K. pneumoniae 8,11 . The overexpression of AcrAB-TolC and OqxAB in clinical K. pneumoniae isolates without a history of tigecycline exposure may be due to exposure to other antimicrobials that are transported by the same efflux pumps [8][9][10][11][12] . The eravacyclineresistant and eravacycline-heteroresistant K. pneumoniae isolates assayed in this study had not been exposed to tigecycline or eravacycline, but their overexpression of AcrAB-TolC and/or OqxAB is consistent with the possibility of their exposure to other compounds transported by these pumps. Although the overexpression of both AcrAB-TolC and OqxAB have previously been implicated in tigecycline resistance in K. pneumoniae; in this study, we observed a more pronounced OqxAB overexpression relative to that of AcrAB-TolC in eravacycline-resistant and -heteroresistant isolates of K. pneumoniae. Similar to previous studies, we observed that the MICs of some eravacycline-resistant and -heteroresistant isolates were inhibited significantly by PAβN, despite the low-level expression of AcrAB-TolC or OqxAB 9 . Eravacycline resistance in these K. pneumoniae isolates may be due to efflux pumps other than AcrAB-TolC or OqxAB. The overexpression of the RND-type efflux pump AcrEF was shown to be associated with fluoroquinolone resistance in Salmonella enterica serovar Typhimurium 24 . A recent study by Zhang et al. 25 observed that AcrEF upregulation contributes to quinolone resistance development in E. coli. Two eravacycline-resistant K. pneumoniae isolates (EKP195 and EKP56) that overexpressed AcrF in the present study provide the first report of the involvement of AcrF in K. pneumoniae tetracycline resistance.
This study provides the first evidence implicating the overexpression of the periplasmic adapter MacAB in eravacycline resistance and heteroresistance in clinical K. pneumoniae isolates. The MacAB-TolC pump assembly, which is now the best-studied bacterial ABC drug exporter, was demonstrated in E. coli to be a cell envelope-spanning transmembrane transporter that actively extrudes the substrates, including macrolide antibiotics and polypeptide virulence factors 26,27 . Our observation of high levels of macA expression in 12 eravacycline-heteroresistant isolates indicates that the MacAB-TolC multidrug efflux pump may, like OqxAB, play an important role in eravacycline heteroresistance in K. pneumoniae. Previous studies have indicated that tigecycline resistance in K. pneumonia is primarily due to the overexpression of AcrAB efflux pump. However, the results of this study showed that the OqxAB and MacAB efflux pumps play a more important role than AcrAB in the development of eravacycline resistance in K. pneumonia. The reason for this difference may be that the C-7 and C-9 of the tetracycline core D-ring of eravacycline are notably different from those in tigecycline, although the two antibacterials belong to the tetracycline family 13 .
The expression of AcrAB-TolC efflux pump can be modulated by several transcriptional regulators, including those encoded by ramA, marA, rarA, soxS, and robA 4,22,23 . In the present study, we observed the overexpression of transcriptional regulator ramA in our eravacyclineresistant and -heteroresistant K. pneumoniae isolates without the overexpression of AcrAB-TolC efflux pumpencoding genes. Interestingly, the overexpression of OqxAB and MacAB-TolC efflux pump-encoding genes was observed in eravacycline-resistant and -heteroresistant K. pneumoniae isolates. The efflux pump OqxAB is activated by the transcriptional regulators RamA and RarA 8,28 . The observed overexpression of OqxAB in this study may have been related to the upregulated expression of RamA, rather than being attributable to RarA or other transcriptional regulators. The ramR gene is located upstream of ramA, and ramR is a negative regulator of ramA. Thus, diverse ramR mutations can lead to the upregulation of ramA expression and subsequently contribute to reduced antibiotic susceptibility 9,28 . In this  Wild + pZP1137-macAB + 0.2%Ara 0.5 2 0.5 Ara arabinose study, the eravacycline-heteroresistant isolates with higher ramA expression than those in eravacyclineresistant isolates may be due to the heteroresistant isolates having more ramR mutations. However, among some eravacycline-resistant (EKP195, EKP154, CRKP21, and EKP217) and -heteroresistant (EKP162-1, EKP82-2, EKP100-1, EKP66-1, EKP57-1, EKP55-1, EKP229-1, EKP17-1, and EKP129-1) isolates, we observed elevated ramA expression in the absence of any ramR mutations. Thus, the mechanisms responsible for high ramA expression may due to other unknown transcriptional regulators in those isolates and require further investigation. Nonetheless, the above-reported efflux pumps cannot explain the observed eravacycline resistance observed in four isolates (EKP192, CRKP6, CRKP21, and EKP56), whose MICs were inhibited significantly by PAβN despite the low-level expression of efflux pump genes. Therefore, the underlying mechanisms for these eravacyclineresistant isolates need to be further studied. The role of the MacAB-TolC multidrug efflux pump in K. pneumoniae resistance to eravacycline also requires further study.
In conclusion, the results of this study implicate the overexpression of OqxAB, but not AcrAB-TolC, as well as the overexpression of the transcriptional regulator RamA in K. pneumoniae eravacycline resistance and heteroresistance. Moreover, the present data suggest that the MacAB-TolC multidrug efflux pump may also participate in eravacycline resistance and heteroresistance in K. pneumoniae.

Bacterial strains and growth conditions
From January 2010 to December 2016, 393 nonduplicate K. pneumoniae isolates from various clinically sampled infections of individuals not previously exposed to tigecycline or eravacycline were collected from inpatients at Shenzhen Nanshan People's Hospital at Shenzhen University School of Medicine in China. The strains were identified with a Phoenix 100 automated microbiology system (BD, Franklin Lakes, NJ, USA) and then, two subcultured generations were reidentified by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (IVD MALDI Biotyper, Germany). All strains were cultured in Luria-Bertani medium at 37°C. All procedures involving human participants were performed in accordance with the ethical standards of the Shenzhen University School of Medicine and with the 1964 Helsinki declaration and its later amendments. For this type of study, formal consent is not required.

Antimicrobial susceptibility testing
The ESBL production of K. pneumoniae isolates was detected using the Phoenix 100 automated microbiology system. MICs for imipenem, meropenem, tigecycline, and eravacycline were determined by the agar dilution method, according to the Clinical and Laboratory Standards Institute (CLSI) guidelines (CLSI-M100-S26). The MIC breakpoints for carbapenem were defined according to CLSI-M100-S26. Isolates observed to have elevated MICs for carbapenem were confirmed by a manual ETEST® (bioMérieux) to have a reduced susceptibility to either imipenem or/and meropenem 29 .

PAP analysis
PAPs were obtained as described previously 30 . Onehundred-microliter aliquots of a starting cell suspension (corresponding to a 0.5 McFarland standard for K. pneumoniae cultures grown on blood agar plates for 24 h at 37°C; 1.0-1.5 × 10 8 cfu/ml) were spread onto Mueller-Hinton agar plates with or without eravacycline (1, 2, 4, 6, 8, or 16 mg/l). After incubating for 24 h at 37°C, the number of colonies was counted. As CLSI Enterobacteriaceae MIC breakpoints for tigecycline and eravacycline have not yet been established, K. pneumoniae isolates with an MIC ≥4 mg/l for tigecycline or eravacycline were considered to be not susceptible based on the reference studies 17,31 . Eravacycline heteroresistance was defined as the presence of eravacycline-susceptible isolates with an eravacycline MIC of <2 mg/l, in which detectable subpopulations grew in the presence of ≥4 mg/l eravacycline, with a detection limit of 20 cfu/ml. Each analysis was performed three times.

Efflux inhibition
Efflux pump activity in tigecycline-and/or eravacyclineresistant K. pneumoniae isolates was detected using the efflux pump inhibitor (EPI) Phe-Arg-β-naphthylamide (PAβN, Sigma). Tigecycline and eravacycline MICs were determined by the agar dilution method in the presence and absence of PAβN (50 mg/l). Significant inhibition of efflux pumps was confirmed based on the MIC reduction to a quarter (or more) of the baseline values in the presence of EPI 8 .

qRT-PCR
Transcript expression levels of the efflux pump genes acrA, acrB, tolC, oqxA, oqxB, acrE, acrF, macA, and macB and their transcriptional regulator genes acrR, marA, soxS, rarA, robA, and ramA were determined by qRT-PCR, with the primers listed in Table S4. Total bacterial RNA was extracted with an RNeasy mini kit (QIAGEN GmbH, Hilden, Germany), and cDNA was synthesized with a PrimeScript RT reagent kit (TAKARA BIO INC, Shiga, Japan). Finally, qRT-PCR was performed using a SYBR Premix Ex Taq II kit (TAKARA BIO INC, Shiga, Japan) in a Mastercycler ep realplex system (Eppendorf, Hamburg, Germany), with an initial incubation at 95°C for 2 min, followed by 40 cycles of 15 s at 95°C and 60 s at 60°C. Each sample was run in triplicate. The expression of target genes was normalized relative to the 16S rRNA housekeeping gene rrsE. Threshold cycle (Ct) numbers were confirmed by the qRT-PCR system software, and data were analyzed in accordance with the 2 −ΔΔCt method. The expression levels of the target genes were compared with those of K. pneumoniae ATCC 13883 (tigecycline susceptible 9 and eravacycline susceptible [according to our sensitivity test] strain, expression = 1).
Detection of mutations in acrR, ramR, rpsJ, oqxAB, and macAB The acrR, ramR, and rpsJ codons of the S10 ribosomal protein-encoding genes, as well as those of the efflux pump-encoding genes oqxAB and macAB were amplified by PCR (primers listed in Table S4) and sequenced. Mutations in acrR, ramR, and rpsJ in K. pneumoniae isolates were identified by comparison with a reference sequence, namely, the genome of K. pneumoniae subsp. pneumoniae MGH 78578 (GenBank accession number CP000647) 9 . Mutations in oqxAB and macAB were identified by comparison with reference sequences from the NCBI database (for detailed sequences see Tables S2  and S3 footnotes).
Overexpression of OqxAB and MacAB in eravacyclinesensitive isolates of K. pneumoniae Complete oqxAB gene was amplified by PCR from the eravacycline-heteroresistant clinical isolate EKP82-2, and the macAB gene was amplified from eravacyclineheteroresistant clinical isolate EKP129-1. The PCR fragments were purified and digested with endonucleases NheI and BglII and then were inserted into the plasmid pZP1137 for gene overexpression. Correct cloning was verified by PCR and sequencing. Verified plasmid constructs were introduced into three eravacycline-sensitive K. pneumoniae clinical isolates: CRKP9, EKP130, and EKP209. All strains, plasmids, and primers used for overexpression are listed in Tables S5 and S6. The overexpression of OqxAB and MacAB was induced with 0.2% arabinose (Ara). All assays were performed at least in triplicate.

Statistical analysis
Continuous data are reported as the means ± standard deviations (SDs) and were analyzed with Student's t tests, one-way factorial analyses of variance (ANOVA), or nonparametric Mann-Whitney U tests. Categorical data are reported as numbers (with percentages) and were compared with Chi-square or Fisher's exact tests. P values <0.05 were regarded as significant. All data were analyzed in SPSS (version 17.0, Chicago, IL, USA).