Gene 534 (2014) 313–319

Contents lists available at ScienceDirect

Gene journal homepage: www.elsevier.com/locate/gene

Suppression of stop codon UGA in acrB can contribute to antibiotic resistance in Klebsiella pneumoniae ATCC10031 Motoyasu Onishi a,1, Minako Mizusawa a,1, Tomofusa Tsuchiya b, Teruo Kuroda a, Wakano Ogawa a,⁎ a b

Department of Microbiology, Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama University, Tsushima, Kita-ku, Okayama 700-8530, Japan Department of Microbiology, College of Pharmaceutical Sciences, Ritsumeikan University, Noji-higashi, Kusatsu, Shiga 525-8577, Japan

a r t i c l e

i n f o

Article history: Accepted 11 October 2013 Available online 26 October 2013 Keywords: Klebsiella pneumoniae Drug resistance Plasmid Drug efflux pump Nonsense mutation

a b s t r a c t We previously reported that Klebsiella pneumoniae MGH78578 exhibited higher resistance against various antimicrobials than K. pneumoniae ATCC10031. In this study, we showed that the plasmid, pKPN5, in K. pneumoniae MGH78578 played an important role in resistance against aminoglycosides, ampicillin, tetracycline, and chloramphenicol, while genome-derived β-lactamases and drug efflux pumps appeared to be more important in resistance to cloxacillin. acrAB, encoding a potent multidrug efflux pump, was cloned from K. pneumoniae MGH78578 and ATCC10031, to investigate reasons for the high drug resistance of K. pneumoniae MGH78578, and the results revealed that AcrAB from K. pneumoniae ATCC10031 conferred weaker drug resistance than AcrAB from K. pneumoniae MGH78578. DNA sequencing revealed that acrB from K. pneumoniae ATCC10031 carried the nonsense mutation, UGA, which was not found in acrB from K. pneumoniae MGH78578. However, acrB from K. pneumoniae ATCC10031 conferred slightly elevated resistant levels to several antimicrobials. The intact length of AcrB was detected in K. pneumoniae ATCC10031 by Western blot analysis, even though its quantity was small. Therefore, the stop codon UGA in acrB was thought to be overcome to some extent in this strain. We artificially introduced the nonsense mutation, UGA to the cat gene on pACYC184, and the plasmid also elevated the MIC of chloramphenicol in K. pneumoniae ATCC10031. These results suggest that a mechanism to overcome the nonsense mutation in acrB sustained resistance against a few β-lactams, dyes, and cholic acid in K. pneumoniae ATCC10031. © 2013 Elsevier B.V. All rights reserved.

1. Introduction We previously reported that Klebsiella pneumoniae ATCC10031 was more sensitive to various antimicrobial chemicals than K. pneumoniae MGH78578 (Ogawa et al., 2005). K. pneumoniae MGH78578 was declared as a clinical isolate and its whole genome sequence was clarified (accession NC_009648). This strain was shown to possess genes on plasmids for resistance to antibiotics such as chloramphenicol and aminoglycosides. Meanwhile, the origin of K. pneumoniae ATCC10031 remains obscure (information on the ATCC homepage and by personal communications with the ATCC). The genome sequence of this strain has not yet been investigated; therefore, it is difficult to presume the factors related to drug resistance in this strain.

Abbreviations: SDS, sodium dodecyl sulfate; TPPCl, tetraphenylphosphonium; CCCP, carbonyl cyanide m-chlorophenylhydrazone; RND, resistance–nodulation–cell division; RT-PCR, reverse transcription polymerase chain reaction; MIC, minimum inhibitory concentration; PCR, polymerase chain reaction; AcrBKp, AcrB from K. pneumoniae; AcrBEc, AcrB from E. coli. ⁎ Corresponding author at: Department of Molecular Microbiology, Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama University, Tsushima, Kita-ku, Okayama, 700-8530, Japan. Tel./fax: +81 86 251 7958. E-mail address: [email protected] (W. Ogawa). 1 These two authors contributed equally to this article. 0378-1119/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gene.2013.10.029

We believed that it was unreasonable to account for the large difference in drug resistant levels between K. pneumoniae ATCC10031 and MGH78578 by plasmids only because the resistant levels to detergents, dyes, and quaternary ammonium compounds were also markedly different between these two strains. Multidrug efflux pumps have been associated with fundamental resistance to various antimicrobial chemicals in Gram-negative bacteria (Hornsey et al., 2010; Magnet et al., 2001; Pumbwe et al., 2006). Multidrug efflux pumps belonging to RND family in Escherichia coli and Pseudomonas aeruginosa have been shown to play an important role in intrinsic resistance to antimicrobials (Morita et al., 2001a; Morita et al., 2001b; Yoneyama et al., 1997). The AcrAB–TolC system in the RND family was shown to be the most important for intrinsic drug resistance in E. coli, and resistance to various antibiotic chemicals was decreased in AcrAB–TolC defective mutants of this bacterium (Fralick, 1996; Ma et al., 1995). K. pneumoniae, which is a member of Enterobacteriaceae, the same as E. coli, generally possesses the AcrAB–KocC system on the genome (e.g. AcrA: YP_001334126.1 in K. pneumoniae MGH78578, ZP_14288389.1 in K. pneumoniae DSM30104, YP_002240036.1 in K. pneumoniae 342, YP_005225489.1 in K. pneumoniae HS11286, AcrB: YP_001334125.1 in K. pneumoniae MGH78578, ZP_14288388.1 in K. pneumoniae DSM30104, YP_002240037.1

314

M. Onishi et al. / Gene 534 (2014) 313–319

transformation, and the plasmid solution may have also contained the other four plasmids. However, only pKPN5 possesses chloramphenicol resistant genes, and K. pneumoniae ATCC10031 transformed with pKPN5 was selected on L-agar plates containing 20 μg/ml chloramphenicol.

in K. pneumoniae 342, YP_005225488.1 in K. pneumoniae HS11286, KocC: YP_001337075.1 in K. pneumoniae MGH78578, ZP_14290986.1 in K. pneumoniae DSM30104, YP_002236547.1 in K. pneumoniae 342, YP_005228876.1 in K. pneumoniae HS11286). The primary structures of these proteins from K. pneumoniae exhibited high similarity to the corresponding proteins from E. coli. In our investigation, these proteins were able to recognize almost the same variety of chemicals as their substrates, and AcrAB–KocC was thought to be a crucial system for drug resistance in K. pneumoniae (Li et al., 2008). We previously reported that ethidium efflux activity in K. pneumoniae ATCC10031 was markedly weaker than that in K. pneumoniae MGH78578 (Ogawa et al., 2005). We anticipated the difference in drug resistant levels between K. pneumoniae ATCC10031 and MGH78578 to be accountable for the difference in the mRNA expression of acrAB. However, acrA and acrB mRNA expression levels were only two-fold higher in K. pneumoniae MGH78578 than in K. pneumoniae ATCC10031, which adds to the difficulty in explaining the different drug resistant levels between these two strains (Li et al., 2008). Here, we described the involvement of drug resistant factors on plasmids and the genome in K. pneumoniae, including AcrAB, to explain the difference in drug resistant levels between the two K. pneumoniae strains.

Plasmid elimination from K. pneumoniae MGH78578 was performed using Zorzópulos's method with a few modifications (Zorzópulos et al., 1984). K. pneumoniae MGH78578 was cultured in Lbroth containing 1.3 mg/ml acridine orange and 0.5 mM carbonyl cyanide m-chlorophenylhydrazone (CCCP) at 42 °C until stationary phase of growth. A total of 100 μl of the cell culture was spread on L-agar plates and incubated at 37 °C. We then selected cells unable to grow on L-agar plates containing 100 μg/ml kanamycin. We investigated whether pKPN5 was detected in the candidates by the preparation of plasmids. The presence of pKPN5 was investigated by PCR to amplify the qacEΔ1 region on pKPN5. One of the strains that lost pKPN5 was named K. pneumoniae GS3. K. pneumoniae GS3 was then investigated by PCR with a respective set of primers to determine whether pKPN3, pKPN4, pKPN6, and pKPN7 were present or not.

2. Materials and methods

2.3. Preparation of everted membrane vesicles

2.1. Strains and plasmids

Cells were cultured until the late exponential phase in L broth at 37 °C. The harvested cells were washed twice with cold FPV buffer (10 mM Tris–HCl (pH 7.2), 140 mM choline chloride, 5 mM MgSO4, 6 mM β-mercaptoethanol, and 10% glycerol) and were suspended in the same buffer. Everted membrane vesicles were prepared by passing cells through a French press as described previously (Kuroda et al., 1994). Membrane vesicles were suspended in the FPV buffer and an equal volume of glycerol was added to the vesicle suspension. The membrane mixture was frozen in liquid nitrogen and stored at − 80 °C until use. Protein concentrations were determined by the method of Lowry et al. (1951).

The strains and plasmids used in this study were listed in Table 1. pKPN5 was extracted from K. pneumoniae MGH78578 for the Table 1 Bacterial strains and plasmids used in this study. Strains and plasmids K. pneumoniae MGH78578

ATCC10031 GS3 SKY2

Reference

Multidrug-resistant strain, used for the genome project, possessing five plasmids (pKPN3, pKPN4, pKPN5, pKPN6, and pKPN7) ATCC collection, parental strain of SKY2 K. pneumoniae strain prepared by removing This study pKPN4 and pKPN5 from MGH78578 Deletion mutant of acrABKp from ATCC10031 Ogawa et al. (2012)

E. coli TG1 KAM32

Parental strain of KAM32 and KAM33 Deletion mutant of acrB and ydhE

KAM33

Deletion mutant of acrAB and ydhE

Plasmids pKPN5 pSTV28 pKAC30A

Plasmid found in K. pneumoniae MGH78578, accession NC_009651 Vector plasmid

pSTV28 derivative, carrying acrB from K. pneumoniae ATCC10031 pKAC28A pSTV28 derivative, carrying acrAB from K. pneumoniae ATCC10031 pKAC28M pSTV28 derivative, carrying acrAB from K. pneumoniae MGH78578 pKBU28 pSTV28 derivative, carrying the acrB anterior part from K. pneumoniae ATCC10031 pUC18 Vector plasmid pKBD19 pUC18 derivative, carrying the acrB posterior part from K. pneumoniae ATCC10031 pACYC184 Vector plasmid pACYC184cat(op85) pACYC184 carrying the nonsense mutation in cat

Matsuo et al. (2007) Matsuo et al. (2007)

Takara Bio, Inc. This study This study This study This work

This study

This study

2.2. Construction of K. pneumoniae GS3

2.4. Western blot analysis Proteins in membrane vesicles were transferred to a nitrocellulose membrane (Advantec Toyo Roshi Kaisha, Ltd., Tokyo Japan) after SDSpolyacrylamide gel electrophoresis. To detect AcrB from K. pneumoniae (hereinafter AcrB from K. pneumoniae was called AcrBKp and AcrB from E. coli was called AcrBEc), the anti-AcrBEc antibody was used as a primary antibody because AcrBKp from K. pneumoniae and AcrB from E. coli show high similarity with each other, and the anti-AcrB antibody to E. coli can also detect AcrBKp. This antibody was a generous gift from Dr. A. Yamaguchi (Institute of Scientific and Industrial Research, Osaka University). The secondary antibody, goat anti-rabbit IgG with horseradish peroxidase, was purchased from Thermo Fisher Scientific K.K. The ECL Western blotting detection system (GE Healthcare UK Ltd., Buckinghamshire, England) was used for the detection. The quantification of AcrBKp expression was achieved using ImageJ 1.43u (National Institutes of Health, USA). 2.5. Site-directed mutagenesis in the cat gene Nucleotide substitution was performed by the method of Tomic et al. with a few modifications (Tomic et al., 1990). A mutation-introduced fragment was amplified by Primer2′ BbsI and Primer2 ScaI with pACYC184 as a template. Another fragment was also amplified using Primer1 AvaI and Primer1′ BbsI with pACYC184 as a template. These fragments were then ligated after the BbsI treatment. The ligated fragment was used as a template for PCR with the primers of Primer1 EcoRI and Primer2 ScaI. The amplified fragment was treated with EcoRI and ScaI, and ligated with pACYC184, digested with EcoRI and

M. Onishi et al. / Gene 534 (2014) 313–319

ScaI to replace the corresponding region with the fragment introduced nonsense mutation. The introduced mutation in the cat gene was confirmed by sequencing and the resultant plasmid was named pACYC184cat(op85) (Fig. 3). Sequences of the primers were listed in Table 2.

Table 3 MICs of the various antimicrobial agents of the K. pneumoniae strains. Antimicrobial agent

2.6. MIC determination The minimum inhibitory concentrations (MICs) of antimicrobial agents were determined as described previously (Ogawa et al., 2012). The same experiment was repeated at least five times and showed the most reproducible values in Tables 3, 4, 5, and 6. 2.7. Plasmid construction Almost the same strategy was applied to construct pKAC28M, pKAC28A, pKAC30A, and pKBU28. A DNA fragment of acrAB gene was amplified using the primers, acrAB Fw and acrAB Re with the genome DNA from K. pneumoniae MGH78578 or ATCC10031 as a template. A set of primers, acrB Fw Sal and acrAB Re, was used for the construction of pKAC30A, and a set of primers, acrB Fw Sal and acrB UP Re Hind, was used for pKBU28. Amplified fragments were digested with SalI and Hind III, and ligated with pSTV28 (Takara Bio, Inc., Otsu, Japan) digested with the same set of restriction enzymes.

Table 2 Primers used in this study. Primers

Sequences

Purpose

pKPN3new Fw

5′-tatggcggatattcgggggc-3′

pKPN3new Re

5′-gctgctgcatcagcctgtcg-3′

pKPN4 Fw

5′-cgttgctggccgtacatttg-3′

pKPN4 Re

5′-ccaagataagcctgcctagc-3′

pKPN5 Fw

5′-gcgtgcataataagccctac-3′

pKPN5 Re

5′-agattcagaatgccgaacac-3′

pKPN6 Fw

5′-ctccatctcctgactcttcg-3′

pKPN6 Re

5′-cctcgatagtccaaccatcc-3′

pKPN7 Fw

5′-agtacacacccctttcctgc-3′

pKPN7 Re

5′-ttgcagacaccccagcattc-3′

Primer1 AvaI

5′-ccaacgctgcccgagatgc-3′

Primer1 EcoRI

5′-ctttattcacattcttgcccg-3′

Primer1′ BbsI

5′-gaagactcaccagctcaccg-3′

Primer2 ScaI

5′-cgtttaagggcaccaataactg-3′

Primer2′ BbsI

5′-gaagacgctggtgatatgagatag-3′

acrAB Fw

5′ggttgtcgacgtatgtaccatagcatgacc3′ 5′agacaagcttccacagccggagaaatagag3′ 5′gcaggtcgaccgccatcagaacaaaccaag3′ 5′cggaaagcttgtcatgccatacccaccacg3′ 5′tatcaagcttctgttcgtccgtctaccgag-3′

Detection of pKPN3 in MGH78578 Detection of pKPN3 in MGH78578 Detection of pKPN4 in MGH78578 Detection of pKPN4 in MGH78578 Detection of pKPN5 in MGH78578 Detection of pKPN5 in MGH78578 Detection of pKPN6 in MGH78578 Detection of pKPN6 in MGH78578 Detection of pKPN7 in MGH78578 Detection of pKPN7 in MGH78578 Introduction for the mutation on pACYC184 Introduction for the mutation on pACYC184 Introduction for the mutation on pACYC184 Introduction for the mutation on pACYC184 Introduction for the mutation on pACYC184 Gene cloning of acrAB

acrAB Re

acrB Fw Sal

acrB UP Re Hind

acrB DOWN Fw Hind(pKBD19)

Gene cloning of acrAB and acrB Gene cloning of acrB

Construction of pKBU28

Construction of pKBD19

315

Cloxacillin Ampicillin Cefazolin Ceftriaxone Imipenem Norfloxacin Erythromycin Kanamycin Tetracycline Chloramphenicol Novobiocin Acriflavine Benzalkonium Cl Ethidium Br Rhodamine 6G TPPCla a

MIC (μg/ml) K. pneumoniae ATCC10031 ATCC10031/ pKPN5

SKY2

MGH78578 GS3

16 32 2 0.03 0.125 0.03 8 1 1 1 4 4 4 32 32 64

4 32 2 0.015 0.25 0.03 4 1 1 1 4 4 4 8 8 32

N1024 N1024 N16 16 0.25 4 512 512 512 1024 512 256 32 1024 N1024 N1024

128 1024 4 0.03 0.25 0.06 8 512 128 128 2 8 4 64 32 128

N1024 128 2 0.125 0.125 4 256 2 4 8 256 256 32 1024 N1024 N1024

TPPCl: tetraphenylphosphonium chloride.

The primers, acrB DOWN Fw Hind(pKBD19) and acrAB Re were used to amplify the latter half of acrBKp to construct pKBD19. The amplified fragment was digested with Hind III, and ligated with pUC18 digested with the same enzyme. A plasmid whose part of acrBKp was connected in the right direction to the lac promoter on the pUC18 was selected. E. coli KAM32 was used as the cloning host. No nucleotide substitutions in the cloned genes (or fragments) were confirmed by sequencing. The sequences of the primers were listed in Table 2.

3. Results 3.1. Contribution of plasmids to the multidrug-resistant phenotype in K. pneumoniae MGH78578 K. pneumoniae MGH78578 was previously reported to possess five plasmids (pKPN3 [accession NC_009649], pKPN4 [accession NC_009650], pKPN5 [accession NC_009651], pKPN6 [accession

Table 4 MICs of various antimicrobial agents in host cells with acrAB from K. pneumoniae MGH78578 or ATCC10031. Antimicrobial agent

MIC (μg/ml) Host: E. coli KAM33 pSTV28 pKAC28M

pKAC28A

Host: K. pneumoniae SKY2 pSTV28

pKAC28M

pKAC28A

Cloxacillin 4 512 64 8 1024 8 Norfloxacin 0.03 0.25 0.03 0.03 0.125 0.03 Erythromycin 4 512 32 16 512 16 Kanamycin 1 1 1 1 1 1 Tetracycline 0.5 4 0.5 0.5 2 2 Novobiocin 1 512 16 4 128 4 Acriflavine 8 128 16 4 128 8 Benzalkonium Cl 4 32 8 4 8 4 Hoechst 33342 0.5 8 1 1 N16 2 Ethidium Br 4 1024 128 32 512 128 Rhodamine 6G 8 N1024 128 16 1024 32 SDSa 64 N1024 N1024 128 N1024 256 b TPPCl 8 1024 512 32 N1024 256 Cholate 5000 N40,000 20,000 5000 20,000 5000 a b

SDS: sodium dodecyl sulfate. TPPCl: tetraphenylphosphonium chloride.

316

M. Onishi et al. / Gene 534 (2014) 313–319

Table 5 MICs of various antimicrobial agents in cells carrying the anterior part, posterior part, or both respective parts of acrB. Antimicrobial agent

MIC (μg/ml) Host: E. coli KAM32

Norfloxacin Erythromycin Tetracycline Novobiocin Acriflavine Benzalkonium Cl Hoechst 33342 Ethidium Br Rhodamine 6G SDSa TPPClb a b

pSTV28

pKBU28

pSTV28

pKBU28

pUC18

pUC18

pKBD19

pKBD19

0.015 4 0.5 4 4 4 1 4 8 256 4

0.015 2 0.5 4 2 4 0.5 2 8 128 4

0.015 4 0.5 4 4 4 1 4 8 128 4

0.015 4 0.5 4 4 4 1 4 8 256 4

pKAC30A

0.015 8 0.5 16 8 4 1 32 64 N1024 64

SDS: sodium dodecyl sulfate. TPPCl: tetraphenylphosphonium chloride.

NC_009652], and pKPN7 [accession NC_009653]). Of the five, pKPN5 was estimated to possess many antibiotic-resistant genes (e.g. four aminoglycoside-resistant genes [str: putative APH(3″) streptomycin phosphotransferase (YP_001338789.1), KPN_pKPN5p08180: streptomycin resistance protein B (YP_001338790.1), aphA: aminoglycoside 3′phosphotransferase (YP_001338804.1), and KPN_pKPN5p08204: aminoglycoside adenylyltransferase (YP_001338814.1)], two chloramphenicolresistant genes [cmlA: chloramphenicol efflux protein (YP_001338812.1) and cmlA2: chloramphenicol acetyltransferase (YP_001338844.1)], a tetracycline efflux gene [tetD: tetracycline efflux protein, class D (YP_001338866.1)], two drug efflux pump genes [qacEΔ1: multidrug efflux pump (YP_001338811.1) and KPN_pKPN5p08252: sugE-like multidrug efflux pump (YP_001338862.1)], and a β-lactamase gene [bla: putative TEM-1 β-lactamase (YP_001338868.1)]) (Bissonnette et al., 1991; Braus et al., 1984; Paulsen et al., 1993; Salverda et al., 2010; Son et al., 2003; Wright, 1999). pKPN4 was presumed to possess both β-lactamase and aminoglycoside-resistant genes (two β-lactamase genes [KPN_pKPN4p07050: putative βlactamase SHV-12 (YP_001338660.1) and KPN_pKPN4p07062: β-lactamase (YP_001338671.1)] and two aminoglycoside-resistant genes [KPN_pKPN4p07058: putative aminoglycoside N(6′)-acetyltransferase (YP_001338668.1) and KPN_pKPN4p07058: streptomycin 3′-adenyltransferase (YP_001338669.1)]); however, genes possibly related to drug resistance were not found in the other three plasmids. On the other hand, whether K. pneumoniae ATCC10031 possesses plasmids is unclear because the genome sequence has not been investigated. To investigate the contribution of pKPN5 to the drug-resistant phenotype, we transformed K. pneumoniae ATCC10031 with pKPN5 using an increase in resistance to chloramphenicol, and measured the minimum inhibitory concentrations of various antimicrobial agents.

Table 6 MICs of the antibiotics of K. pneumoniae ATCC10031 possessing a plasmid with a nonsense mutation on the cat gene. Antimicrobial agent

MIC (μg/ml) Host: K. pneumoniae ATCC10031

Chloramphenicol Tetracycline Ampicillin

No plasmid

pACYC184

pACYC184cat(op85)

1 1 32

128 32 32

16 32 32

The MICs of cloxacillin, ampicillin, chloramphenicol, kanamycin, and tetracycline were significantly elevated in K. pneumoniae ATCC10031/ pKPN5 (Table 3). Resistant levels to kanamycin reached to the same level as K. pneumoniae MGH78578, and this result indicated that pKPN5 was critically important in aminoglycoside resistance in K. pneumoniae MGH78578. Cloxacillin resistance was also eight-fold higher in K. pneumoniae ATCC10031/pKPN5. However, pKPN5 was not effective for resistance to cephems and imipenem. Ethidium bromide (ethidium Br) was reported as a substrate of QacEΔ1, a multidrug efflux pump (Paulsen et al., 1993). Acriflavine was also reported to be recognized as a substrate of QacEΔ1 found in Staphylococcus aureus (Kazama et al., 1998). However, the MICs of these chemicals were only two-fold higher in K. pneumoniae ATCC10031/pKPN5. Therefore, pKPN5 was unlikely to contribute to an increase in resistance to these chemicals. We then removed pKPN5 from K. pneumoniae MGH78578 and measured the MICs of antimicrobials on the resultant strain, K. pneumoniae GS3 (Table 3). When the existence of the other plasmids was investigated, we found that the elimination of pKPN5 also caused the simultaneous loss of pKPN4 (Fig. 1S). The genome project of K. pneumoniae MGH78578 revealed that pKPN4 possessed four drug resistant genes. The MIC of kanamycin decreased from 512 μg/ml to 2 μg/ml in K. pneumoniae GS3 and this value was similar to that of K. pneumoniae ATCC10031. The MICs of tetracycline and chloramphenicol in the GS3 strain were 128-fold lower than those of K. pneumoniae MGH78578, and the MICs of ampicillin and cefazolin decreased from N1024 μg/ml to 128 μg/ml and from N16 μg/ml to 2 μg/ml, respectively. The MIC of cloxacillin remained unchanged in the investigated range (1024–0.5 μg/ml) in K. pneumoniae GS3, while cloxacillin resistance in K. pneumoniae ATCC10031/pKPN5 increased. The decreased resistance to ceftriaxone in K. pneumoniae GS3 may have been caused by loss of pKPN4 because pKPN4 possesses β-lactamase (SHV-12). From these results, we concluded that pKPN5 and pKPN4, especially pKPN5 played a prominent role in the multidrug-resistant phenotype in K. pneumoniae MGH78578. On the other hand, it also appeared that the difference in resistant levels to several chemicals between K. pneumoniae MGH78578 and ATCC10031 was unable to only be attributed to the retention of plasmids. 3.2. Measurement of MIC in acrAB-transformed cells We previously compared the mRNA expression levels of acrABKp between K. pneumoniae MGH78578 and ATCC10031 using RT-PCR (Li et al., 2008). The mRNA expression of acrABKp in K. pneumoniae MGH78578 was only two-fold higher than that in K. pneumoniae ATCC10031, although K. pneumoniae MGH78578 showed very high resistance to a wide variety of chemicals. We previously disrupted acrABKp in K. pneumoniae ATCC10031, however, a marked decrease in drug resistance levels was not observed in the gene-disrupted strain, K. pneumoniae SKY2 (Ogawa et al., 2012). Marked MIC changes were previously observed when we constructed E. coli KAM33, an acrABdisrupted strain from E. coli TG1 (Matsuo et al., 2007). Most of the MIC values in K. pneumoniae SKY2 seemed to be comparable to those in E. coli KAM33. For these reasons, we speculated that the activity of AcrABKp in K. pneumoniae ATCC10031 was lower than that in K. pneumoniae MGH78578. We then investigated AcrABKp activity from K. pneumoniae MGH78578 and ATCC10031 (Table 4). The MICs of various chemicals were markedly increased in E. coli KAM33 transformed with pKAC28M carrying acrABKp from K. pneumoniae MGH78578, as we previously reported (Li et al., 2008). Resistance levels to cloxacillin, novobiocin, ethidium Br, rhodamine 6G, sodium dodecyl sulfate (SDS), tetraphenylphosphonium Cl (TPPCl), and cholate were also elevated in E. coli KAM33/pKAC28A possessing acrABKp from K. pneumoniae ATCC10031 within a range of eight-fold to sixty-four-fold. However, it was clear that pKAC28M derived higher drug resistance to host cells

M. Onishi et al. / Gene 534 (2014) 313–319

317

than pKAC28A. This was the same when K. pneumoniae SKY2 was used as the host (Table 4). K. pneumoniae SKY2/pKAC28M exhibited high resistance to various antibiotics. On the other hand, the MICs of ethidium Br and tetracycline were only four-fold higher and that of TPPCl was eight-fold higher in K. pneumoniae SKY2/pKAC28A than in SKY2/ pSTV28. The MICs of other chemicals were not markedly increased in K. pneumoniae SKY2/pKAC28A, although cloxacillin, erythromycin, novobiocin, acriflavine, rhodamine 6G, and SDS were good substrates for AcrABKp.

K. pneumoniae MGH78578, but T in K. pneumoniae ATCC10031 (Fig. 1). However, it was deduced that this substitution did not change the amino acid residue (Fig. 1). Another difference was found in position 1659 in acrBKp. The nucleotide was G in K. pneumoniae MGH78578, but A in K. pneumoniae ATCC10031 (Fig. 1). Using this nucleotide substitution, the codon TGG for tryptophan in K. pneumoniae MGH78578 was deduced to change into TGA, coding the stop codon in K. pneumoniae ATCC10031. Thus, the lower activity of AcrABKp from K. pneumoniae ATCC10031 was thought to be caused by the stop codon.

3.3. Sequence analysis of acrABKp from K. pneumoniae MGH78578 and ATCC10031

3.4. Suppression of the nonsense codon in acrB by read-through or suppressor tRNA

The activity of AcrABKp derived from K. pneumoniae MGH78578 and ATCC10031 was different, and we determined the DNA sequence of the cloned acrABKp genes on pKAC28M and pKAC28A. The length of acrABKp from K. pneumoniae ATCC10031 was the same as that of K. pneumoniae MGH78578. The DNA sequence of the whole cloned region on pKAC28M coincided with that of the genome database of K. pneumoniae MGH78578, and it was revealed that no mutation was introduced during amplification by PCR. No nucleotide difference was observed between acrAKp on pKAC28A and the nucleotide sequence of K. pneumoniae MGH78578. The 37 nucleotides upstream of the start codon of acrAKp were also the same as that of K. pneumoniae MGH78578. Two nucleotide differences were found in acrBKp between pKAC28A and pKAC28M. These differences were also found on the genome sequence of K. pneumoniae ATCC10031, and were not errors during PCR amplification. One of the differences was the 147th nucleotide in acrBKp from the start codon of acrBKp. The nucleotide was C in acrBKp in

Although we found the stop codon TGA in acrBKp from K. pneumoniae ATCC10031, cloned acrBKp from K. pneumoniae ATCC10031 increased drug resistance levels in the host cells, E. coli KAM33, K. pneumoniae ATCC10031, and K. pneumoniae SKY2 even though it was much weaker than acrBKp from K. pneumoniae MGH78578 (Table 4). A decrease in drug resistance levels was also observed when acrABKp in K. pneumoniae ATCC10031 was deleted (Table 3). These results suggested that AcrBKp in K. pneumoniae ATCC10031 maintained activity as a multidrug efflux pump even though a stop codon was in the ORF. We formulated two hypotheses to explain the activity of AcrBKp from K. pneumoniae ATCC10031: 1) Kaback's group showed that the E. coli cells expressing the N-terminal portion and C-terminal portion of LacY separately could transport lactose even though its transport activity was low (Bibi and Kaback, 1990). They also reported that only the Cterminal portion of LacY also exhibited transport activity (Wu et al., 1996). We expected that AcrBKp would also be active in only a part of AcrBKp or that the separately expressed N-terminal portion and Cterminal portion of AcrBKp would interact in cells, leading to an elevation in drug resistance. The other hypothesis was: 2) read-through of the stop codon has sometimes been observed (Nilsson and RydenAulin, 2003). Suppressor tRNA was also shown to overcome stop codons. Therefore, intact AcrBKp may be produced in K. pneumoniae ATCC10031. First, we cloned part of the acrBKp gene, including from the start codon to position 1659 of the stop codon (pKBU28), and separately cloned the remaining part of acrBKp encoding the C-terminal part (pKBD19) (Fig. 1). E. coli KAM32 was transformed with the plasmids. mRNA expression from each part of acrBKp was confirmed in E. coli KAM32/pKBU28/pKBD19 (data not shown). The MICs of drugs were measured in E. coli cells possessing the two plasmids, however, the MIC of any drugs was not increased in E. coli KAM32/pSTV28/pKBD19, KAM32/pKBU28/pUC18, or KAM32/pKBU28/pKBD19 (Table 5). Thus, we concluded that only the N-terminal portion or only the C-terminal portion of AcrBKp was inactive and the separately expressed Nterminal portion and C-terminal portion did not make an active AcrBKp. We then investigated the second hypothesis using Western blotting analysis. An antibody against AcrB from E. coli was used to detect AcrBKp expression because AcrBEc and AcrBKp are very similar, and the antiAcrBEc antibody was also able to detect AcrBKp.

Fig. 1. Differences in the nucleotide sequences on acrBKp between K. pneumoniae MGH78578 and ATCC10031. Two nucleotides (147th and 1659th from the start codon of acrBKp) were different in acrBKp between K. pneumoniae MGH78578 and ATCC10031. These nucleotides were shown in bold letters. Changes in the encoding amino acid residue did not occur with the substitution of C147T. However, the substitution, G1659A formed a stop codon. The gray bar shows the cloned region on the respective plasmids. pKBU28 possessed the anterior part of acrBKp up to the 1660th nucleotide, C, and pKBD19 possessed the posterior part of acrBKp from the 1660th nucleotide, C.

Fig. 2. Immunoblotting analysis of the expression of AcrB. Two micrograms of membrane protein was loaded onto Lanes 1–3 and 6. Eight micrograms of membrane protein was loaded onto Lanes 4 and 5. Membranes from K. pneumoniae MGH78578 (Lane 1), K. pneumoniae ATCC10031 (Lane 2 and 4), K. pneumoniae SKY2 (Lane 3 and 5), and E. coli TG1 as a control (Lane 6). AcrBKp was marked with small arrows.

318

M. Onishi et al. / Gene 534 (2014) 313–319

A faint band corresponding to AcrBKp appeared in Lane 4 in Fig. 2 from K. pneumoniae ATCC10031. The isometric band was not detected in the sample from K. pneumoniae SKY2 whose acrABKp was deleted (Lane 5 in Fig. 2). Therefore, we considered that whole length of AcrBKp was generated in K. pneumoniae ATCC10031. Compared to K. pneumoniae MGH78578, the amount of AcrBKp in K. pneumoniae ATCC10031 was estimated to be approximately 8% on analysis with Image J (Fig. 2). The small amount of AcrBKp was thought to be able to raise resistance against drugs to some extent. We constructed pACYC184cat(op85) possessing an artificial nonsense mutation (UGG → UGA) in the chloramphenicol resistant gene (cat) (Fig. 3). The MIC of chloramphenicol in K. pneumoniae ATCC10031 transformed with pACYC184cat(op85) was sixteen-fold higher than that in K. pneumoniae ATCC10031 (Table 6). The MICs of tetracycline were the same between cells carrying pACYC184 or pACYC184cat(op85) though the tetracycline resistant gene also existed on the plasmids. The increased chloramphenicol-resistant level was lower than that in cells possessing an intact cat gene, but this result suggests that an active CAT protein was produced from the cat gene on pACYC184cat(op85) in spite of the nonsense mutation. The stop codon UGA was previously shown to not always terminate translation in E. coli, and UGA was suppressed by normal tryptophanyl-tRNATrp in E. coli, though the frequency was low (Engelberg-Kulka, 1981; Hirsh and Gold, 1971). Such previous reports and our results suggest that read-through of the UAG stop codon may have occurred in K. pneumoniae ATCC10031 at a certain degree of frequency or that K. pneumoniae ATCC10031 harbored suppressor tRNA for UAG. 4. Discussion In the present study, we identified a single nucleotide substitution (G → A) in acrBKp in K. pneumoniae ATCC10031, relative to acrBKp from K. pneumoniae MGH78578, and also showed that the nucleotide substitution accordingly produced the stop codon UGA in acrBKp in K. pneumoniae ATCC10031. We previously reported that ethidium efflux activity in K. pneumoniae ATCC10031 was markedly lower than that in K. pneumoniae MGH78578 (Ogawa et al., 2005). This weak efflux activity in K. pneumoniae ATCC10031 may be caused by acrBKp with the nonsense mutation and also be related to the low resistance levels observed against various antibiotics in this strain. On the other hand, cloned acrBKp with the nonsense codon UAG elevated drug-resistant levels of the host cells to some extent and we detected the intact length of AcrBKp in cells harboring acrBKp with a nonsense codon. We concluded that this nonsense codon may have been overcome by read-through of the stop codon or suppressor tRNA, and the resultant AcrBKp increased drug-resistant levels of the cell even though the expression of AcrBKp was not high.

Interestingly, the low activity of AcrABKp from K. pneumoniae ATCC10031 was also observed in E. coli KAM33 (Table 4). In addition, the MICs of cloxacillin, erythromycin, novobiocin, ethidium Br, rhodamine 6G, SDS, TPPCl, and cholate in E. coli KAM33/pKAC28A were markedly higher than those in K. pneumoniae SKY2/pKAC28A. For this reason, the copy number of the plasmid was thought to be different between E. coli KAM33 and K. pneumoniae SKY2 because pSTV28 must be originally produced as a vector for use in E. coli. Besides, E. coli TG1, the parental strain of KAM33, is an amber suppressor strain, and this suppressor may suppress the nonsense codon in acrBKp from K. pneumoniae ATCC10031. From the results obtained from the immunoblotting analysis, we estimated that approximately 7% of AcrBKp was produced in K. pneumoniae ATCC10031, relative to K. pneumoniae MGH78578. Multiple factors (including the amount of suppressor tRNA, release factors, and codon context effect) were reported to generally determine suppression efficiency (Bulmer, 1987; Kopelowitz et al., 1992; Shimizu et al., 2001). However, our estimation was very similar to a previous report investigating the codon context effect (Kopelowitz et al., 1992). In this study, we demonstrated that pKPN5 in K. pneumoniae MGH78578 played an important role in resistance to aminoglycosides, tetracycline, ampicillin, and chloramphenicol, whereas the loss of pKPN4 and pKPN5 from K. pneumoniae MGH78578 did not reduce the MIC of cloxacillin. These results suggest that only intrinsic factors derived from genes on the genome were able to confer high resistance to cloxacillin because we could not find β-lactamase genes on pKPN3, pKPN6, and pKPN7. It is unclear whether K. pneumoniae ATCC10031 possesses the same β-lactamase genes on the genome as K. pneumoniae MGH78578 because the genome sequence of K. pneumoniae ATCC10031 has not been investigated. However, the introduction of pKPN5 to K. pneumoniae ATCC10031 did not strongly affect cloxacillin resistance. bla on pKPN5 encodes TEM-1 type β-lactamase, which is also known as a penicillinase. The contribution of β-lactamase from pKPN5 may be low because cloxacillin is a penicillinase-tolerant β-lactamase. Genomederived factors (e.g. multidrug efflux pumps and β-lactamase) were more important for cloxacillin resistance than pKPN5. The results in Table 4 showed that AcrABKp significantly contributed to cloxacillin resistance. Generally, β-lactams are not very good substrates for multidrug efflux pumps; however, the contribution of AcrABKp to resistance to some β-lactams may be more than we had previously estimated. 5. Conclusion We showed that the plasmid, pKPN5, in K. pneumoniae MGH78578 played an important role in resistance against aminoglycosides, ampicillin, tetracycline, and chloramphenicol, while genome-derived βlactamases and drug efflux pumps were thought to be more important in resistance to several antimicrobial agents including cloxacillin. acrAB, which is a genome derived gene and encoding a potent multidrug efflux pump, was cloned from K. pneumoniae MGH78578 and ATCC10031. We then revealed that the AcrAB activity from K. pneumoniae ATCC10031 was weaker than that from K. pneumoniae MGH78578, and we identified a nonsense mutation on acrB from K. pneumoniae ATCC10031. However, we also showed that AcrB from K. pneumoniae ATCC10031 still possessed the activity to elevate resistance to antimicrobial agents. We confirmed that the nonsense mutation on acrB was overcome in K. pneumoniae ATCC10031 though the efficiency was not high. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.gene.2013.10.029. Conflict of interest

Fig. 3. The artificially introduced nucleotide substitution in pACYC184.A nucleotide substitution (G to A) was introduced at the position 255th in cat on pACYC184, and formed the nonsense codon, UGA in cat.

The authors declare that there is no conflict of interest relevant to this manuscript.

M. Onishi et al. / Gene 534 (2014) 313–319

Acknowledgments We would like to thank Dr. Akihito Yamaguchi and Dr. Kunihiko Nishino of Institute of Scientific and Industrial Research, Osaka University for providing the anti-AcrB antibody. English proofreading was performed by Medical English Service (Sakyo-ku, Kyoto). This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan. References Bibi, E., Kaback, H.R., 1990. In vivo expression of the lacY gene in two segments leads to functional lac permease. Proc. Natl. Acad. Sci. U. S. A. 87, 4325–4329. Bissonnette, L., Champetier, S., Buisson, J.P., Roy, P.H., 1991. Characterization of the nonenzymatic chloramphenicol resistance (cmlA) gene of the In4 integron of Tn1696: similarity of the product to transmembrane transport proteins. J. Bacteriol. 173, 4493–4502. Braus, G., Argast, M., Beck, C.F., 1984. Identification of additional genes on transposon Tn10: tetC and tetD. J. Bacteriol. 160, 504–509. Bulmer, M., 1987. Coevolution of codon usage and transfer RNA abundance. Nature 325, 728–730. Engelberg-Kulka, H., 1981. UGA suppression by normal tRNA Trp in Escherichia coli: codon context effects. Nucleic Acids Res. 9, 983–991. Fralick, J.A., 1996. Evidence that TolC is required for functioning of the Mar/AcrAB efflux pump of Escherichia coli. J. Bacteriol. 178, 5803–5805. Hirsh, D., Gold, L., 1971. Translation of the UGA triplet in vitro by tryptophan transfer RNA's. J. Mol. Biol. 58, 459–468. Hornsey, M., Ellington, M.J., Doumith, M., Hudson, S., Livermore, D.M., Woodford, N., 2010. Tigecycline resistance in Serratia marcescens associated with up-regulation of the SdeXY-HasF efflux system also active against ciprofloxacin and cefpirome. J. Antimicrob. Chemother. 65, 479–482. Kazama, H., Hamashima, H., Sasatsu, M., Arai, T., 1998. Distribution of the antisepticresistance gene qacEΔ1 in gram-positive bacteria. FEMS Microbiol. Lett. 165, 295–299. Kopelowitz, J., Hampe, C., Goldman, R., Reches, M., Engelberg-Kulka, H., 1992. Influence of codon context on UGA suppression and readthrough. J. Mol. Biol. 225, 261–269. Kuroda, T., Shimamoto, T., Inaba, K., Tsuda, M., Tsuchiya, T., 1994. Properties and sequence of the NhaA Na+/H+ antiporter of Vibrio parahaemolyticus. J. Biochem. 116, 1030–1038. Li, D.W., et al., 2008. Properties and expression of a multidrug efflux pump AcrAB–KocC from Klebsiella pneumoniae. Biol. Pharm. Bull. 31, 577–582. Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265–275. Ma, D., Cook, D.N., Alberti, M., Pon, N.G., Nikaido, H., Hearst, J.E., 1995. Genes acrA and acrB encode a stress-induced efflux system of Escherichia coli. Mol. Microbiol. 16, 45–55.

319

Magnet, S., Courvalin, P., Lambert, T., 2001. Resistance–nodulation–cell division-type efflux pump involved in aminoglycoside resistance in Acinetobacter baumannii strain BM4454. Antimicrob. Agents Chemother. 45, 3375–3380. Matsuo, T., et al., 2007. VmeAB, an RND-type multidrug efflux transporter in Vibrio parahaemolyticus. Microbiology 153, 4129–4137. Morita, Y., Kimura, N., Mima, T., Mizushima, T., Tsuchiya, T., 2001a. Roles of MexXY- and MexAB-multidrug efflux pumps in intrinsic multidrug resistance of Pseudomonas aeruginosa PAO1. J. Gen. Appl. Microbiol. 47, 27–32. Morita, Y., Komori, Y., Mima, T., Kuroda, T., Mizushima, T., Tsuchiya, T., 2001b. Construction of a series of mutants lacking all of the four major mex operons for multidrug efflux pumps or possessing each one of the operons from Pseudomonas aeruginosa PAO1: MexCD-OprJ is an inducible pump. FEMS Microbiol. Lett. 202, 139–143. Nilsson, M., Ryden-Aulin, M., 2003. Glutamine is incorporated at the nonsense codons UAG and UAA in a suppressor-free Escherichia coli strain. Biochim. Biophys. Acta 1627, 1–6. Ogawa, W., et al., 2005. Multidrug resistance in Klebsiella pneumoniae MGH78578 and cloning of genes responsible for the resistance. Biol. Pharm. Bull. 28, 1505–1508. Ogawa, W., Onishi, M., Ni, R., Tsuchiya, T., Kuroda, T., 2012. Functional study of the novel multidrug efflux pump KexD from Klebsiella pneumoniae. Gene 498, 177–182. Paulsen, I.T., et al., 1993. The 3′ conserved segment of integrons contains a gene associated with multidrug resistance to antiseptics and disinfectants. Antimicrob. Agents Chemother. 37, 761–768. Pumbwe, L., Ueda, O., Yoshimura, F., Chang, A., Smith, R.L., Wexler, H.M., 2006. Bacteroides fragilis BmeABC efflux systems additively confer intrinsic antimicrobial resistance. J. Antimicrob. Chemother. 58, 37–46. Salverda, M.L., De Visser, J.A., Barlow, M., 2010. Natural evolution of TEM-1 β-lactamase: experimental reconstruction and clinical relevance. FEMS Microbiol. Rev. 34, 1015–1036. Shimizu, Y., et al., 2001. Cell-free translation reconstituted with purified components. Nat. Biotechnol. 19, 751–755. Son, M.S., Del Castilho, C., Duncalf, K.A., Carney, D., Weiner, J.H., Turner, R.J., 2003. Mutagenesis of SugE, a small multidrug resistance protein. Biochem. Biophys. Res. Commun. 312, 914–921. Tomic, M., Sunjevaric, I., Savtchenko, E.S., Blumenberg, M., 1990. A rapid and simple method for introducing specific mutations into any position of DNA leaving all other positions unaltered. Nucleic Acids Res. 18, 1656. Wright, G.D., 1999. Aminoglycoside-modifying enzymes. Curr. Opin. Microbiol. 2, 499–503. Wu, J., Sun, J., Kaback, H.R., 1996. Purification and functional characterization of the Cterminal half of the lactose permease of Escherichia coli. Biochemistry 35, 5213–5219. Yoneyama, H., Ocaktan, A., Tsuda, M., Nakae, T., 1997. The role of mex-gene products in antibiotic extrusion in Pseudomonas aeruginosa. Biochem. Biophys. Res. Commun. 233, 611–618. Zorzópulos, J., Trevisan, A.R., Denoya, C.D., 1984. Deletions in Klebsiella pneumoniae R plasmids induced by growth in the presence of acridine orange at high temperature. Antimicrob. Agents Chemother. 25, 659–661.