PDF
ePub
Citation
Print
INTRODUCTION
Acinetobacter baumannii is an opportunistic pathogen that causes hospital-acquired infections, especially in intensive care units [123]. The incidence and worldwide dissemination of carbapenem-resistant A. baumannii isolates named global clone (GC)1 and GC2 have increased [456]. Carbapenem resistance in A. baumannii has mainly been ascribed to degradation by carbapenem-hydrolyzing oxacillinases including OXA-23, OXA-24, OXA-58, OXA-51, OXA-143, and their variants [27].
Large DNA sequences (6-200 kb) are termed genomic islands when they are acquired from other organisms through plasmids, transposons, and other mobile genetic elements [89]. Among the genomic islands of A. baumannii, multiple antibiotic resistance regions that have inserted into the target ATPase gene (comM) are termed resistance islands (AbaR) [10]. There are several known AbaRs, which differ according to the clone type [1112]. The AbaR3-type resistance islands are the major resistance islands of GC1 isolates, and the AbaR4-type islands are associated with GC2 isolates [131415]. Some AbaR4-type islands of A. baumannii harbor transposons such as Tn2009, Tn2006, and Tn6008 carrying the blaOXA-23 gene [16].
Because in vitro approaches to investigate genomic islands are time-consuming and expensive, several software tools for identifying putative genomic islands have been developed. Some useful software tools and resources for identification of antimicrobial resistance genes have also been reported, such as ResFinder from the Center for Genomic Epidemiology and the Comprehensive Antibiotic Research Database (CARD) [1718]. Recently, an island prediction software tool (Genomic Island Prediction software [GIPSy]) was introduced for the prediction of genomic islands in bacteria [9]. Here, we investigated the whole genome sequence (WGS) of a carbapenem-resistant A. baumannii isolate belonging to GC2 and predicted the resistance islands using software tools.
METHODS
1. Bacterial strains and antimicrobial susceptibility tests
The A. baumannii YU-R612 strain was isolated from the sputum of a 61-yr-old man with sepsis in Severance Hospital, Seoul, Korea. The species was identified by conventional methods and the VITEK 32 GN system (bioMérieux, Marcy l'Etoile, France). Antimicrobial susceptibility testing was performed by using the VITEK II N211 system (bioMérieux). The YU-R612 strain was resistant to multiple antibiotics including piperacillin (minimum inhibitory concentration [MIC] ≥128 µg/mL), ceftazidime (≥64 µg/mL), cefotaxime (≥64 µg/mL), cefepime (≥64 µg/mL), imipenem (≥16 µg/mL), meropenem (≥16 µg/mL), ampicillin/sulbactam (≥32 µg/mL), ciprofloxacin (≥4 µg/mL), and gentamicin (≥16 µg/mL); however, it was susceptible to minocycline (≤1 µg/mL) and colistin (≤0.5 µg/mL). Modified Hodge tests and EDTA plus sodium mercaptoacetic acid (SMA) double-potentiation tests were carried out to screen carbapenemase and metallo-β-lactamase activity, respectively. Both tests showed negative results, which suggested the absence of metallo-β-lactamase genes in this strain.
2. Whole genome sequencing and PCR experiments
The WGS of the YU-R612 strain was obtained by using the PacBio RS II sequencing system (Pacific Biosciences Inc. Menlo Park, CA, USA) in a commercial laboratory (Macrogen, Seoul, Korea). PCR was performed by using in-house primers (Table 1).
3. Identification of the resistome and resistance islands
Antimicrobial resistance genes were analyzed by using ResFinder, the resources from the Center for Genomic Epidemiology (www.genomicepidemiology.org). A ResFinder threshold of ID=98% was selected. Genomic islands were determined by using resources from Genomic Island Prediction software (GIPSy) as previously described [9].
RESULTS
The A. baumannii strain YU-R612 genome consisted of a circular chromosome (ca. 4,075 kb) and two plasmids (ca. 74 kb and 5 kb). WGS analysis showed that the sequence type (ST) under the Oxford scheme was ST191 (1-3-3-2-2-94-3) and under the Institute Pasteur Multi Locus Sequence Typing (MLST) scheme was ST2, which was consistent with the assignment of this strain to GC2. ResFinder analysis showed that the YU-R612 strain possessed the following resistance genes on its chromosome: four β-lactamase genes, namely blaADC-30 (identity, 100%), blaOXA-66 (100%), blaOXA-23 (100%), and blaTEM-1 (100%); armA (100%), aadA1 (100%), and aacA4 (99.6%) as aminoglycoside resistance-encoding genes; aac(6')Ib-cr (99.2%) for fluoroquinolone resistance; msr(E) (100%) for macrolide, lincosamide, and streptogramin B resistance; catB8 (100%) for phenicol resistance; and sul1 (100%) for sulfonamide resistance.
Six putative resistant islands (PRIs) were determined on the chromosome of the YU-R612 strain by using the GIPSy tool (Fig. 1). PRI1 possessed two copies of Tn2009 carrying the blaOXA-23 gene (Fig. 2A). PRI5 carried two copies of a class I integron including 5' conserved sequence (CS) and 3'CS, and their gene cassettes possessed sul1 and armA resistance genes (Fig. 2B). PRI2, PRI3, PRI4, and PRI6 carried several hypothetical proteins that could not be further annotated. Antimicrobial resistance genes detected by ResFinder were not found on PRI2, PRI3, PRI4, or PRI6.
Interestingly, in addition to the PRIs, a resistance island was identified that was inserted in the comM gene on the YU-R612 chromosome. The resistance island was 10.6 kb in size and very similar to AbaR4-type resistance islands, especially Tn6022, which consists of the tniC-tniA-tniB-tinE-orf-uspA-sul1 genes bracketed by a partial comM gene (Fig. 2C). There were two plasmids in the YU-R612 strain, which did not harbor PRIs or antimicrobial resistance genes.
This draft genome sequencing project has been deposited at the GenBank nucleotide database (GenBank nos. CP014215, CP014216, and CP014217) under BioProject ID PRJNA309091.
DISCUSSION
Strain YU-R612 is a multiple antibiotic-resistant isolate recovered from a patient with pneumonia. Resistance to multiple antibiotics correlated with the presence of antimicrobial resistance genes detected by ResFinder: blaOXA-23 for resistance to imipenem and meropenem, armA for gentamicin, and aac(6')Ib-cr for ciprofloxacin, which are major resistance genes for A. baumannii in Korea [1920].
Under the Oxford scheme, the YU-R612 strain was shown to be ST191, indicating that it is a member of the GC2, which accounts for the majority of isolates recovered in Korea [19]. GC2, which includes clonal complex 92 in the MLST Oxford scheme, has also been identified among carbapenem-resistant A. baumannii strains in Asian and European countries [19]. Among the resistance islands of A. baumannii, AbaR4-type resistance islands are common in GC2 isolates [12]. Clonal dissemination and diversification of GC92 and the predominance of AbaR4-type resistance islands among these isolates have been reported in Korea [15].
AbaR4-type islands that shared a core structure similar to AbaR4 appeared to be distributed in GC2 isolates via integration into a distinct genomic site at comM. The backbone of AbaR4 comprised five open reading frames constituting the so-called transposition module with two other genes encoding the universal stress protein (uspA) and sulfate permease (sul1) [21]. In this study, a 10.6-kb AbaR4-type resistance island was detected in the YU-R612 strain. While some AbaR4-type resistance islands possess transposons such as Tn2006 and Tn1213 carrying antibiotic resistance genes such as blaOXA-23, the YU-Y612 strain did not carry any transposons or antibiotic resistance genes on its AbaR4-type-like resistance island. However, the blaOXA-23 gene on the Tn2009 of YU-R612 strain was detected in PRI1.
In this study, six PRIs were identified on the chromosome of the YU-R612 strain by using the GIPSy tool. Among them, two PRIs (PRI1 and PRI5) possessed transposons and class I integrons that carried important resistance genes: blaOXA-23, armA, and sul1. The duplication of Tn2009 in PRI1 and class I integrons in PRI5 was confirmed by using PCR to supplement a limitation of deep sequencing for the replication.
The GIPSy software facilitated the prediction of resistance islands and resistance genes in the WGS of this strain. Although some PRIs detected by software tools are different from known resistance islands, this software may be useful to identify novel genomic islands carrying clusters of resistance genes. Software tools provide a convenient way of identifying PRIs as well as acquired antimicrobial resistance genes, and they may simplify the complex steps in analyzing the WGS of bacteria in clinical microbiology laboratories.
In conclusion, the prediction of resistance islands in the WGS of a carbapenem-resistant A. baumannii strain belonging to GC2 in Korea revealed PRIs on the chromosome that carried Tn2009 and class I integrons. The predicted resistance islands using software tools helped in the analysis of the WGS in this strain.
XML Download