Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

First Emergence of acrAB and oqxAB Mediated Tigecycline Resistance in Clinical Isolates of Klebsiella pneumoniae Pre-Dating the Use of Tigecycline in a Chinese Hospital

  • Xue Zhong,

    Affiliations Department of Pharmaceutics, School of Pharmacy, Shenyang Pharmaceutical University, Shenyang, Liaoning, PR China, Department of Pharmacy, Beijing Hospital, Beijing, PR China

  • Hongtao Xu,

    Affiliation Department of Clinical Laboratory, Beijing Hospital, Beijing, PR China

  • Dongke Chen,

    Affiliation Department of Clinical Laboratory, Beijing Hospital, Beijing, PR China

  • Haijian Zhou,

    Affiliations State Key Laboratory for Infectious Disease Prevention and Control, National Institute for Communicable Disease Control and Prevention, Chinese Centre for Disease Control and Prevention, Beijing, PR China, Collaborative Innovation Centre for Diagnosis and Treatment of Infectious Diseases, Hangzhou, PR China

  • Xin Hu,

    Affiliation Department of Pharmacy, Beijing Hospital, Beijing, PR China

  • Gang Cheng

    Affiliation Department of Pharmaceutics, School of Pharmacy, Shenyang Pharmaceutical University, Shenyang, Liaoning, PR China

First Emergence of acrAB and oqxAB Mediated Tigecycline Resistance in Clinical Isolates of Klebsiella pneumoniae Pre-Dating the Use of Tigecycline in a Chinese Hospital

  • Xue Zhong, 
  • Hongtao Xu, 
  • Dongke Chen, 
  • Haijian Zhou, 
  • Xin Hu, 
  • Gang Cheng


Tigecycline is one of the few therapeutic options for treating infections caused by some multi-drug resistant pathogens, such as Klebsiella pneumoniae. However, tigecycline-resistant K. pneumoniae has been discovered recently in China. From 2009 to 2013, nine tigecycline-resistant K. pneumoniae isolates were identified in our hospital. Six of nine strains were identified before using tigecycline. To investigate the efflux-mediated resistance mechanisms of K. pneumoniae, the expression of efflux pump genes (acrA, acrB, tolC, oqxA and oqxB) and pump regulators (acrR, marA, soxS, rarA, rob and ramA) were examined by real-time RT-PCR. Molecular typing of the tigecycline resistant strains was performed. ST11 was the predominant clone of K. pneumoniae strains, while ST1414 and ST1415 were novel STs. Efflux pump inhibitor (EPI)-carbonyl cyanide chlorophenylhydrazone (CCCP) was able to reverse the resistance patterns of 5 resistant K. pneumoniae strains. In comparison with strain A111, a tigecycline-susceptible strain (negative control), we found that the expression levels of efflux pump genes and pump regulators were higher in a majority of resistant strains. Higher expression levels of regulators rarA (2.41-fold, 9.55-fold, 28.44-fold and 18.31-fold, respectively) and pump gene oqxB (3.87-fold, 31.96-fold, 50.61-fold and 29.45-fold, respectively) were observed in four tigecycline resistant strains (A363, A361, A368, A373, respectively). Increased expression of acrB was associated with ramA and marA expression. To our knowledge, studies on tigecycline resistance mechanism in K. pneumoniae are limited especially in China. In our study, we found that both efflux pump AcrAB-TolC and OqxAB contributed to tigecycline resistance in K. pneumoniae isolates.


A derivative of minocycline, tigecycline, is a new class of glycylcyclines. It exhibits a broad-spectrum of activity against most of Gram-positive, Gram-negative pathogens , anaerobes, and ‘atypical’ bacteria [1]. However, tigecycline-resistance has emerged recently in different pathogens such as Acinetobacter baumannii, Klebsiella pneumoniae and Enterobacteriaceae, especially in multidrug-resistant (MDR) strains. Over the period 2007 to 2013, there have been 13 reports of tigecycline-resistant strains from USA, UK, France, Saudi Arabia, Greece, Spain, Germany and Taiwan and, among them, four were K. pneumoniae, seven were A. baumannii, one was Enterobacter hormaechei, one was Enterococcus faecalis and one was Bacteroides fragilis [2] . Also, most of them were changed from tigecycline-susceptible to tigecycline-resistant during treatment. The highest minimum inhibitory concentration (MIC) of tigecycline was 24 µg/mL [3]. Recently, emergence of non-susceptibility to tigecycline has also been reported in A. baumannii and K. pneumoniae in Mainland China [4], [5] . Here, we present nine cases of tigecycline-resistant K. pneumoniae in Mainland China. Tigecycline was approved by the US Food and Drug Administration (FDA) in 2005 and was launched in Mainland China in 2011. However, we have discovered six strains resistant to tigecycline isolated before 2011 which shows that there existed tigecycline-resistant strains before using tigecycline. This phenomenon was in accordance with tigecycline-resistant report of Rosenblum et al. which pointed out that tigecycline resistance predates its introduction [6] . We hypothesized that mechanisms of cross-resistance or other as-yet-undefined regulatory elements could be active in such strains. Currently , Veleba M et al. have shown that the factors for decreased sensitivity to tigecycline in K. pneumoniae may be contribute to overexpression of the AcrAB RND-type and newly described OqxAB efflux pumps [7]. In addition, the pumps were regulated by ramA, marA, soxS, rarA or rob of transcriptional activators [6], [8][10] . The main purpose of this study was to investigate whether similar mechanisms existed in the resistant strains isolated from Mainland China or whether there are other reasons for this tigecycline resistance.

Materials and Methods

Bacterial strains

Two hundred and fifty-four carbapenem-resistant Enterobacteriaceae isolates were isolated from clinical samples of Beijing Hospital between 2009 and 2013. (belong to sample secondary use, free of informed consent and ethic). Identification of the strains was performed by using the VITEK-2 and susceptibility testing was carried out by Kirby-bauer, broth microdilution and E-test methods then ten K. pneumoniae isolates were selected to study the efflux mechanism. These consisted of nine K. pneumoniae isolates resistant to tigcycline and one K. pneumoniae isolate, A111, susceptible to tigcycline used as a negative control. Escherichia coli ATCC 25922 and K. pneumoniae ATCC 11296 were used as reference strains. The strains used in this study are shown in Table 1.

Tigecycline susceptibility testing

Tigecycline susceptibility testing was carried out using the following three different methods: Firstly, the Kirby-bauer method was used as a tigecycline susceptibility test. Then, the minimum inhibitory concentrations (MICs) of tigecycline were determined by standard broth microdilution tests and the E-test (BioMerieux), according to the Clinical and Laboratory Standards Institute (CLSI) recommendation and manufacturers' instructions. The MICs for the strains were interpreted in accordance with FDA guidelines for tigecycline, MIC≦2 µg/mL and ≧8 µg/mL were classified as susceptible and resistant respectively. E. coli ATCC 25922 was used as the quality control strain.

Molecular typing by MLST

In order to check the clonality of the selected isolates,the ten isolates were typed by Multilocus sequence typing (MLST). MLST with seven housekeeping genes (gapA, infB, mdh, pgi, phoE, rpoB and tonB) was performed in accordance with the protocol described on the K. pneumoniae MLST database. Alleles and sequence types (STs) of strains were assigned by using tools on the K. pneumoniae MLST database. BioNumerics version 5.10 software (Applied Maths, Kortrijk, Belgium) was used to create the minimum spanning tree [11] and in this, the founder ST was defined as the ST with the greatest number of single-locus variants. Types were represented by circles and the size of a circle indicated the number of strains with this particular type.

Efflux mechanism

The efflux pump inhibitor (EPI) carbonyl cyanide-chlorophenylhydrazone (CCCP; Sigma) was used to investigate the activity of the efflux pump in the tigecycline-resistant and tigecycline-susceptible K. pneumoniae isolates. The MICs of tigecycline were determined by the broth microdilution method in the presence and absence of CCCP (fixed concentration of 16 µg/mL). If the MIC values decreased 4-fold or greater in the presence of EPIs, this was defined as a significant inhibition effect [12] .

Real-time fluorescence quantitative PCR (RT-FQ-PCR)

The expression levels of the acrR, marA, soxS, rarA, rob, and ramA regulator genes and the efflux components acrA, acrB, tolC, oqxA and oqxB were assessed using RT-FQ-PCR. Oligonucleotide primers and probes used for RT-FQ-PCR (Table 2) were designed with Primer Express Software version 2.0 (Applied Biosystems) and purchased from Life Technologies (AB & Invitrogen). RNA was extracted using an EASYspin RNeasy Kit (Gene-Foci) then, for cDNA synthesis, a cDNA Synthesis Kit (Invitrogen) was used, according to the manufacturer's instructions. The cDNAs were quantified by real-time PCR amplification with specific primers (Table 2) using a Taqman One-Step RT-PCR master mix reagents kit (Life Tech, USA) on a Life Tech 7500 Fast Real-Time PCR System (Applied Biosystems) with an initial incubation at 95°C for 2 min, followed by 40 cycles of 10 s at 95°C, 30 s at 60°C, and 10 s at 72°C. Each sample was processed in triplicate. In all cases, a housekeeping gene (rpoB) was used to normalize the expression of the target genes for each isolate. The critical threshold cycle (CT) numbers were determined by the detection system software. The amount of target is given as 2−△CT. Expression analysis was carried out to measure the relative expression of the mRNA compared with that of K. pneumoniae A111.


Tigecycline susceptibility

254 carbapenem-resistant Enterobacteriaceae isolates included K. pneumoniae (192 isolates) , Escherichia coli (41 isolates) , Enterobacter cloaca (12 isolates) and Enterobacter aerogenes (9 isolates). Nine tigecycline-resistant K. pneumoniae isolates were obtained, a tigecycline nonsusceptibility rate of 4.69% (9/192) for carbapenem-resistant K. pneumoniae, One Enterobacter cloacae and one Enterobacter aerogenes besides K. pneumoniae also showed resistance to tigecycline but not the aim of this study. Of all the clinical carbapenem-resistant K. pneumoniae isolates, the tigecycline MIC ranged from 0.5–16 µg/ml with an MIC50 of 4 µg/ml and an MIC90 of 16 µg/ml. Nine K. pneumoniae clinical isolates were collected that required a tigecycline MIC over 8 µg/ml (table 3) . The highest MIC value of tigecycline was 16 µg/ml (A361, A363, A368, A373) . In addition, an ordinary clinical strain A111 which was an ST11 type and susceptible to tigecycline was designated a negative control. So, nine resistant strains and one susceptible strain were included in this study. The susceptibility profiles of the ten strains are presented in Table 3.

Table 3. Susceptibility profiles of nine tigecycline-resistant and one tigecycline-susceptible strain.

Molecular typing of the strains

MLST analysis of nine isolates showed four different sequence types (STs). ST11 was the predominant ST, accounting for 6 (70%) of the isolates. Other three isolates belonged to ST378, ST 1414 and ST1415, with 5, 4 and 5 loci different from ST11, respectively. ST1414 and ST1415 were novel STs identified for the first time in this study.

Efflux pump activity

The effect of the EPIs on tigecycline MICs is shown in Table 4. One strain (A103) showed an 8-fold decrease, four strains (A363R, A361, A368, A373) showed a 4-fold decrease, four strains (A352, A840, A916, A941) showed a 2-fold decrease in the tigecycline MIC in the presence of CCCP (16 µg/ml). There was no reduction in MIC observed in strain A111 under the same conditions.

Table 4. The MIC of tigecycline (TGC) in the multidrug resistant strains, with or without an efflux pump inhibitor, Carbonyl hydrogenated chlorophenyl hydrazone (CCCP).

Analysis of target pump genes and regulator expression

In our study, nine tigecycline-resistant strains were demonstrated to carry efflux pump genes and pump regulators. The relative x-fold increases in the pump genes and pump regulator levels were quantified after comparisons with K. pneumoniae A111 (Fig. 1 a-k). Compared with isolate A111, a higher expression level of regulator rarA (2.41-fold, 9.55-fold, 28.44-fold, and 18.31-fold, respectively) and pump gene oqxB (3.87-fold, 31.96-fold, 50.61-fold, 29.45-fold, respectively) was observed in four tigecycline-resistant strains (A363, A361, A368, A373, respectively). These data suggested that there was a correlation between the increased expression of oqxB and rarA. In addition, except for oqxB, the amount of acrB transcript of the resistant strains was higher in comparison with strain A111. The expression of acrB had an increased tendency with ramA and marA expression. Interestingly, the transcriptional levels of the soxS gene, acrR gene, and rob gene, appeared to be linked to acrB expression, but the role of the three genes was not significant (data shown in Fig. 1).

Figure 1. (a-k) Bar chart showing levels of pump genes and pump regulators in clinical strains of K. pneumoniae.

Relative x-fold increases in the pump genes and regulator levels were quantified after comparisons with K. pneumoniae A111 (tigecycline-susceptible).


Tigecycline has been developed and has attracted much attention as the agent of last resort to treat clinical infections caused by multidrug-resistant K. pneumoniae. In 2011, tigecycline was first launched in Mainland China. However, we discovered six strains which were resistant to tigecycline before 2011, but it cannot be attributed to the use of tigecycline. We hypothesized that this resistance could be indirectly attributed to the use of other antibiotics which were also transported by the same efflux pump because tigecycline is a substrate of chromosomally encoded resistance-nodulation-division efflux pumps. A possible assumption was that these efflux systems were widely existing among K. pneumoniae. When strains exposed to some antibiotics which were substrates of efflux pumps, this might lead to overexpression of these efflux pumps. In addition to our report, other recent reports have described cross-resistance to tigecycline. Hornsey et al. and Deng et al. reported the emergence of resistance to tigecycline associated with other antibiotics rather than tigecycline [4], [13]. In addition, another study concluded that the use of carbapenems might induce resistance not only to carbapenems but also to many other antibiotics, including tigecycline [14] . These reports confirmed our assumption that our isolated tigecycline-resistant strains were likely related to overexpression of the efflux pump which caused cross-resistance.

It has been reported that antibiotic-resistance appeared to be mediated in part by active efflux systems [15], [16]. Recently some studies have demonstrated that efflux pump inhibitors (EPIs) are able to reverse resistance patterns by blocking bacterial pumps and preventing discharge of certain antibiotics [16]. In order to detect if there is an overexpression efflux in resistant K. pneumoniae strains, we used efflux pump inhibitors (EPIs) CCCP to assess the activity of the efflux pumps. In our study, CCCP restored the susceptibility in five strains (A103, A361, A363, A368, A373), indirectly proving that efflux pump overexpression contributed to tigecycline resistance. However, no significant reduction in tigecycline MICs was observed in the other four strains (A352, A840, A916, A941), indicating that CCCP was not effective. One possible explanation is that EPIs have different specificities towards various efflux pumps and interfere with them in a different manner. It may be that other EPIs, such as PAβN, NMP, reserpine or verapamil, could restore the susceptibility of these strains. Also, the results (isolates A352, A840, A916, A941) of efflux pump activity (using CCCP) as well as the expression levels of target pump genes and its regulators suggested that it might have other mechanisms other than the overexpression of efflux pumps mediating tigecycline resistance.

It has been reported that resistance to tigecycline appears to be mediated in part by the active efflux systems AcrAB-TolC [8], [17]. Also, the AcrAB-TolC pump is activated by several transcriptional regulators such as ramA, marA, soxS, rob or acrR. These pump regulators play a role in promoting tigecycline resistance due to up-regulation of the efflux pump AcrAB-TolC [9], [10], [18]. Our present study showed that the increased MIC for K. pneumoniae strains correlated with the overexpression of ramA or marA, demonstrating that ramA was not always needed to confer tigecycline-resistance, and marA was also a global activator of the acrB transporter (Fig. 1). A similar hypothesis was proposed recently, some studies have suggested that other pathways to tigecycline resistance must exist in K. pneumoniae [6], [7]. One of these studies demonstrated that a newly described OqxAB efflux pump could contribute to tigecycline resistance in K. pneumoniae although the chromosomally encoded rarA regulator lies downstream of the efflux pump OqxAB [7] . In our study, the strains A363, A361, A373, and A368 had MIC values of 16 µg/ml, and quantitative real-time PCR analyses showed that rarA overexpression was observed in conjunction with an elevated expression of the multidrug efflux pump OqxAB, confirming that raA might be involved in the regulation of OqxAB production. The results presented here further support a previous report that rarA is one of the regulator pathways that controls the expression of oqxB in K. pneumoniae [19]. However, for the strains A840 and A916, although the expression level of rarA was higher, this did not lead to overexpression of oqxB; on the contrary, for the strains A941 and A103, the amount of the oqxB transcript was higher, but the expression level of rarA was not increased in comparison with the susceptible strain A111. This data indicates that some regulators other than rarA may be able to contribute to oqxB expression. In addition to our report, one recent report has described the same results. In that study, no differences in the sequences 5′of oqxA were found between two strains, which differed 20-fold in levels of oqxB transcripts. Thus, they hypothesized the increased expression of oqxB in the two strains appears not to be due to mutation in a putative promoter, but might relate to differences in other as-yet-undefined regulatory elements in these strains.[20] We hypothesized that this was the reason why the MICs of four strains (A840, A916, A941, A103) were lower than those of the other strains (A363, A361, A368, A373). Efflux pumps AcrAB-TolC play an important role in strains with MICs of 8 µg/ml (A840, A916, A941, A103) . For the strains with an MIC of 16 µg/ml (A363, A361, A368, A373) , the AcrAB-TolC and OqxAB efflux pumps both contributed to the tigecycline-resistance. From our findings, it was clear that rarA as well as those of the pump gene oqxB were more important in the case of our selected tigecycline resistant K. pneumoniae strains. This result was especially important with regard to the previous observations showing that a high-level of tigecycline resistance was exhibited due to alternative pathways in K. pneumoniae [7] . Irrespective of the nature of the pump acting in these strains, we have clearly demonstrated that the efflux pump plays a key role in the tigecycline-resistance in K. pneumoniae. However, the mechanisms responsible for the high expression level of oqxB leading to tigecycline-resistance remain unknown and need further study.

MLST was performed for molecular typing in the present study. ST11 is the epidemic clone of K. pneumoniae carbapenemase 2-producing K. pneumoniae in China, and it had only one of the single-locus variants (SLVs) of ST258 which is the dominant clone in Eastern countries [21]. Other three STs had five to six loci different from ST11.


In conclusion, our study report mechanism of tigecycline-resistant K. pneumoniae in Mainland China. The results of molecular typing suggest that the predominant clone of K. pneumoniae strains was ST11. In addition, we identified two novel STs, namely ST1414 and ST1415. Efflux-mediated mechanisms, including high expression of AcrAB-TolC and OqxAB efflux pumps appears to play a key role in tigecycline resistance while the pump regulators, acrR, marA, soxS, rarA, rob and ramA, all make their individual contributions to the overexpression of AcrAB-TolC and OqxAB efflux pumps. In addition, rarA as well as the pump gene oqxB are more important in our selected high level tigecycline-resistant K. pneumoniae strains. Efflux pump inhibitors (EPIs)-CCCP was able to reverse resistance patterns in the majority of K. pneumoniae strains. Although tigecycline is a promising antibiotic for the treatment of infections caused by multidrug-resistant K. pneumoniae, the development of tigecycline resistance is an issue of great concern.


We thank the team of the curators of the Institute Pasteur MLST system (Paris, France) for importing novel alleles, profiles and or isolates at

Author Contributions

Conceived and designed the experiments: XZ GC XH. Performed the experiments: XZ HZ DC. Analyzed the data: XZ HX HZ. Contributed reagents/materials/analysis tools: DC XH. Wrote the paper: XZ GC HX.


  1. 1. Stein GE, Babinchak T (2013) Tigecycline: an update. Diagn Microbiol Infect Dis 75:331–336.
  2. 2. Sun Y, Cai Y, Liu X, Bai N, Liang B, et al. (2013) The emergence of clinical resistance to tigecycline. Int J Antimicrob Agents 41:110–116.
  3. 3. Reid GE, Grim SA, Aldeza CA, Janda WM, Clark NM (2007) Rapid development of Acinetobacter baumannii resistance to tigecycline. Pharmacotherapy 27:1198–1201.
  4. 4. Deng M, Zhu MH, Li JJ, Bi S, Sheng ZK, et al. (2014) Molecular epidemiology and mechanisms of tigecycline resistance in clinical isolates of Acinetobacter baumannii from a Chinese university hospital. Antimicrob Agents Chemother 58:297–303.
  5. 5. Sheng ZK, Hu F, Wang W, Guo Q, Chen Z, et al. (2014) Mechanisms of Tigecycline Resistance among Klebsiella pneumoniae Clinical Isolates. Antimicrob Agents Chemother 58:6982–6985.
  6. 6. Rosenblum R, Khan E, Gonzalez G, Hasan R, Schneiders T (2011) Genetic regulation of the ramA locus and its expression in clinical isolates of Klebsiella pneumoniae. Int J Antimicrob Agents 38:39–45.
  7. 7. Veleba M, Schneiders T (2012) Tigecycline resistance can occur independently of the ramA gene in Klebsiella pneumoniae. Antimicrob Agents Chemother 56:4466–4467.
  8. 8. Ruzin A, Visalli MA, Keeney D, Bradford PA (2005) Influence of transcriptional activator RamA on expression of multidrug efflux pump AcrAB and tigecycline susceptibility in Klebsiella pneumoniae. Antimicrob Agents Chemother 49:1017–1022.
  9. 9. Bratu S, Landman D, George A, Salvani J, Quale J (2009) Correlation of the expression of acrB and the regulatory genes marA, soxS and ramA with antimicrobial resistance in clinical isolates of Klebsiella pneumoniae endemic to New York City. J Antimicrob Chemother 64:278–283.
  10. 10. Perez A, Poza M, Aranda J, Latasa C, Medrano FJ, et al. (2012) Effect of transcriptional activators SoxS, RobA, and RamA on expression of multidrug efflux pump AcrAB-TolC in Enterobacter cloacae. Antimicrob Agents Chemother 56:6256–6266.
  11. 11. Feil EJ, Li BC, Aanensen DM, Hanage WP, Spratt BG (2004) eBURST: inferring patterns of evolutionary descent among clusters of related bacterial genotypes from multilocus sequence typing data. J Bacteriol 186:1518–1530.
  12. 12. Deng M, Zhu MH, Li JJ, Bi S, Sheng ZK, et al. (2013) Molecular epidemiology and mechanisms of tigecycline resistance in clinical isolates of Acinetobacter baumannii from a Chinese university hospital. Antimicrob Agents Chemother.
  13. 13. Hornsey M, Ellington MJ, Doumith M, Scott G, Livermore DM, et al. (2010) Emergence of AcrAB-mediated tigecycline resistance in a clinical isolate of Enterobacter cloacae during ciprofloxacin treatment. Int J Antimicrob Agents 35:478–481.
  14. 14. Kuo HY, Chang KC, Kuo JW, Yueh HW, Liou ML (2012) Imipenem: a potent inducer of multidrug resistance in Acinetobacter baumannii. Int J Antimicrob Agents 39:33–38.
  15. 15. Pages JM, Lavigne JP, Leflon-Guibout V, Marcon E, Bert F, et al. (2009) Efflux pump, the masked side of beta-lactam resistance in Klebsiella pneumoniae clinical isolates. PLoS One 4:e4817.
  16. 16. Hasdemir UO, Chevalier J, Nordmann P, Pages JM (2004) Detection and prevalence of active drug efflux mechanism in various multidrug-resistant Klebsiella pneumoniae strains from Turkey. J Clin Microbiol 42:2701–2706.
  17. 17. Ruzin A, Immermann FW, Bradford PA (2008) Real-time PCR and statistical analyses of acrAB and ramA expression in clinical isolates of Klebsiella pneumoniae. Antimicrob Agents Chemother 52:3430–3432.
  18. 18. Roy S, Datta S, Viswanathan R, Singh AK, Basu S (2013) Tigecycline susceptibility in Klebsiella pneumoniae and Escherichia coli causing neonatal septicaemia (2007-10) and role of an efflux pump in tigecycline non-susceptibility. J Antimicrob Chemother 68:1036–1042.
  19. 19. Veleba M, Higgins PG, Gonzalez G, Seifert H, Schneiders T (2012) Characterization of RarA, a novel AraC family multidrug resistance regulator in Klebsiella pneumoniae. Antimicrob Agents Chemother 56:4450–4458.
  20. 20. Kim HB, Wang M, Park CH, Kim EC, Jacoby GA, et al. (2009) oqxAB encoding a multidrug efflux pump in human clinical isolates of Enterobacteriaceae. Antimicrob Agents Chemother 53:3582–3584.
  21. 21. Qi Y, Wei Z, Ji S, Du X, Shen P, et al. (2011) ST11, the dominant clone of KPC-producing Klebsiella pneumoniae in China. J Antimicrob Chemother 66:307–312.