An outbreak of blaOXA-51-like- and blaOXA-66-positive Acinetobacter baumannii ST208 in the emergency intensive care unit.

A series of clinical isolates of drug-resistant (DR) Acinetobacter baumannii with diverse drug susceptibility was detected from eight patients in the emergency intensive care unit of Tokai University Hospital. The initial isolate was obtained in March 2010 (A. baumannii Tokai strain 1); subsequently, seven isolates were obtained from patients (A. baumannii Tokai strains 2-8) and one isolate was obtained from an air-fluidized bed used by five of the patients during the 3 months from August to November 2011. The isolates were classified into three types of antimicrobial drug resistance patterns (RRR, SRR and SSR) according to their susceptibility (S) or resistance (R) to imipenem, amikacin and ciprofloxacin, respectively. Genotyping of these isolates by multilocus sequence typing revealed one sequence type, ST208, whilst that by a DiversiLab analysis revealed two subtypes. All the isolates were positive for blaOXA-51-like and blaOXA-66, as assessed by PCR and DNA sequencing. A. baumannii Tokai strains 1-8 and 10 (RRR, SRR and SSR) had quinolone resistance-associated mutations in gyrA/parC, as revealed by DNA sequencing. The ISAba1 upstream of blaOXA-51-like and aminoglycoside resistance-associated gene, armA, were detected in A. baumannii Tokai strains 1-7 and 10 (RRR and SRR) as assessed by PCR. Among the genes encoding resistance-nodulation-division family pumps (adeB, adeG and adeJ) and outer-membrane porins (oprD and carO), overexpression of adeB and adeJ and suppression of oprD and carO were seen in isolates of A. baumannii Tokai strain 2 (RRR), as assessed by real-time PCR. Thus, the molecular characterization of a series of isolates of DR A. baumannii revealed the outbreak of ST208 and diverse antimicrobial drug susceptibilities, which almost correlated with differential gene alterations responsible for each type of drug resistance.

Introduction

Acinetobacter baumannii is emerging as a nosocomial pathogen, particularly in intensive care units, including burn care units (Bayram ; Guzek ; Ohashi ). Hospitalized patients at a greater risk of Acinetobacter infections are those particularly ill on a ventilator, those with a prolonged hospital stay, those who have open wounds and those with invasive devices, such as urinary catheters (Wendt ; Wisplinghoff ; Ho ; Howard ; Zheng ).

A stepwise evolution in the acquisition of multidrug resistance in clinical isolates of Pseudomonas aeruginosa and A. baumannii has been reported (Higgins ; Asai ). Elucidation of the mechanism(s) underlying the drug resistance is important to prevent such resistance.

We identified a series of clinical isolates of drug-resistant (DR) A. baumannii with a diverse pattern of drug resistance from eight patients in the emergency intensive care unit (EICU) of Tokai University Hospital from March 2010 to November 2011. In order to elucidate the diversity on drug resistance in the same sequence type, we studied the molecular characteristics of the clinical isolates and the relationship with the resistance pattern.

Methods

Patients and clinical specimens.

The patients studied were admitted to the EICU (57 beds, including three beds in the severe burn care unit) of Tokai University Hospital (total 804 beds), because of serious burns, traffic injuries or cerebral haemorrhage. During this period, all patients, except two with cerebral haemorrhage, were treated by the systemic administration of antibiotics, including β-lactams, fluoroquinolones, aminoglycosides and glycopeptides, for infections of wounds and/or the respiratory tract.

Routine microbial examinations were performed on a weekly basis on clinical specimens from the patients’ sputum, urine (via catheter), venous blood, wounds, etc.

Bacteriological surveillance of environmental surfaces was performed for the medical equipment shared by the patients in order to seek a possible reservoir of the pathogen, such as an air-fluidized bed. A drug-sensitive strain of A. baumannii (A. baumannii Tokai strain 9; resistance pattern SSS, see below) was used as a control. The epidemiological investigation was performed to elucidate the possible transmission route using information on the clinical care of the patients and detection of A. baumannii. The study was approved by the Review Board of Tokai University (13R-036).

Growth conditions and antibiotic susceptibility testing.

Bacteria were cultured at 37 °C in Luria–Bertani broth (Kyokuto Pharmaceutical Industrial). The criteria for multidrug-resistant (MDR) A. baumannii was resistance to imipenem (IPM; MIC >16 µg ml−1), amikacin (AMK; MIC >32 µg ml−1) and ciprofloxacin (CPFX; MIC >4 µg ml−1), and DR A. baumannii was defined as being resistant to one or two of the drugs according to the Japanese National Guideline Concerning the Prevention of Infections and Medical Care for Patients with Infections.

Biochemical identification and susceptibility testing of the isolates was performed to obtain MIC values using a microdilution method (CLSI, 2009) and the MicroScan WalkAway-96 SI system (Siemens Japan). The antibiotics used in this study were obtained as follows: AMK and IPM were from Banyu Pharmaceutical, aztreonam (AZT) was from Eizai, ceftazidime (CAZ) was from Glaxo SmithKline, cefepime (CFPM) was from Bristol-Myers Squibb, CPFX was from Bayer HealthCare, cefozopran (CZOP) was from Takeda Pharmaceutical, doripenem (DRPM) was from Shionogi, fosfomycin (FOM) was from Meiji Seika, gentamicin (GM) was from Schering-Plough, levofloxacin (LVFX) was from Daiichi-Sankyo Pharmaceutical, meropenem (MEPM) was from Daiichi-Sumitomo Pharmaceutical, minocycline (MINO) was from Wyeth & Takeda Pharmaceutical, piperacillin (PIPC) was from Taisho Toyama Pharmaceutical and tobramycin (TOB) was from Towa Pharmaceutical.

Molecular typing.

DNA templates were extracted using a ZR-Duet DNA/RNA MiniPrep kit (Zymo Research). Multilocus sequence typing (MLST) was performed as described previously (Bartual ; Fu ). MLST sequences were uploaded into the A. baumannii MLST Sequence Type Database (http://pubmlst.org/abaumannii/) to determine the alleles and sequence types. A. baumannii isolates were screened for gene homology by a repetitive-element-based PCR (rep-PCR) DiversiLab Microbial Typing System (Sysmex bioMérieux), which amplified the regions between the non-coding repetitive sequences in bacterial genomes, as described previously (Carretto ; Higgins ). The annealing temperature of the PCR amplification used in this study was 55 °C for gltA, gyrB, recA and cpn60, and 50 °C for gdhB, gpi and rpoD. The amplification products were purified with a DNA purification kit (Qiagen). The DNA sequencing was performed using an ABI3500xL Genetic Analyzer (Applied Biosystems).

Evaluation of the mechanisms of resistance

Screening for metallo-β-lactamase (MBL).

A. baumannii isolates were screened for the production of MBL by a double-disc synergy test with discs containing sodium mercaptoacetic acid as described previously (Arakawa ).

PCR assay for β-lactamase and armA.

The following resistance genes were examined by PCR: blaIMP-1, blaVIM, blaOXA-23-like, blaOXA-24-like, blaOXA-51-like, blaOXA-58-like and ISAba1, as described previously (Turton ; Woodford ). The armA gene, which encodes 16S rRNA methylases and confers high resistance to aminoglycosides, was screened by PCR using primers that were described previously (Yamane ).

Sequencing of OXA-type β-lactamase, and gyrA and parC.

Sequencing of OXA-type β-lactamase was performed as described previously (Endo ). The quinolone resistance-determining regions of gyrA and parC were amplified and analysed as described previously (Liu ). DNA sequencing of the amplified DNA products was performed using an ABI3500xL Genetic Analyzer (Applied Biosystems).

Quantitative real-time (qRT)-PCR).

RNA templates were extracted by a ZR-Duet DNA/RNA MiniPrep kit (Zymo Research). The expression levels of three different genes encoding resistance–nodulation–division (RND) family pumps (adeB, adeG and adeJ) and two different genes encoding outer-membrane porins (oprD and carO) were analysed by qRT-PCR using a StepOnePlus Real-Time PCR System (Applied Biosystems) (Peleg ; Fernando & Kumar, 2012; Zander ). The primers used for the analysis are listed in Table 1. The housekeeping gene 16S rRNA was used as a control (Coyne ; Srinivasan ; Hou ). Reactions (20 µl) were set up using 400 nM primers and 2 µl cDNA template (diluted 1 : 10) with SYBR Premix Ex Taq II (Tli RNaseH Plus) and ROX plus (Takara Bio). The data analysis was carried out using StepOne software. The expression of each target gene was normalized based on the level of the 16S rRNA mRNA gene and was expressed as a relative rate compared with that in the susceptible isolate of each pair (the expression of A. baumannii Tokai strain 9 was taken as 1.0). Experiments were conducted at least three times independently and all reactions were carried out in triplicate.

Table 1.

Primer sequences for qRT-PCR

Primer	Direction	Primer sequence (5'→3′)	Product size (bp)	Reference
adeB	Forward	aatactgccgccaataccag	106	Fernando & Kumar (2012)
	Reverse	ggattatggcgactgaagga
adeG	Forward	atcgcgtagtcaccagaacc	92	Fernando & Kumar (2012)
	Reverse	cgtaactatgcggtgctcaa
adeJ	Forward	catcggctgaaacagttgaa	109	Fernando & Kumar (2012)
	Reverse	gcctgaccattaccagcact
oprD	Forward	ccagctcagttgctcaatca	134	This study
	Reverse	catttggtttccagcgtttt
carO	Forward	ggttataacggcggtgacat	115	This study
	Reverse	ccaaggacgaatttcagcat
16S rRNA	Forward	cgtaagggccatgatgactt	150	Fernando & Kumar (2012)
	Reverse	cagctcgtgtcgtgagatgg

Results

Bacterial strains and antibiotic susceptibility

The characteristics of the A. baumannii Tokai strains are shown in Table 2. In March 2010, a DR A. baumannii Tokai strain 1 was detected initially from the wound of a patient with a severe burn injury. After 1.5 years, during a period of 3 months from August to November 2011, another seven clinical isolates of DR A. baumannii strains from patients (A. baumannii Tokai strains 2–8) were obtained. The DR A. baumannii Tokai strains were classified into three types according to their susceptibility to three drugs (IPM, AMK and CPFX) as RRR, SRR or SSR (R, resistant; S, susceptible; Tables 2 and 3). They were obtained from sputum, wounds and bile drains.

Table 2.

Cases and A. baumannii Tokai strains

One hundred and fifty nurses and five nurse-aids worked in the EICU and Burn centre, and they were not fixed as a team. ER, Critical care and emergency medicine; NR, neurosurgery; OP, orthopaedics; R, resistant; S, susceptible.

Strain/disease	Ward	Day detected after hospitalization	Source	Susceptibility pattern of IPM, AMK and CPFX*	Doctor team	Use of air-fluidized bed
1. 74 % total body surface area burn	Burn centre	31 (3 March 2010)	Sputum	IPM-S, AMK-R, CPFX-R (SRR)	ER-a	Yes
2. 85 % total body surface area burn (Ohashi et al., 2013)	Burn centre	9 (29 August 2011)	Wound	IPM-R, AMK-R, CPFX-R (RRR)	ER-b	Yes
3. 40 % total body surface area burn (Ohashi et al., 2013)	Burn centre	33 (19 September 2011)	Wound	IPM-R, AMK-R, CPFX-R (RRR)	ER-c	No
4. 70.5 % total body surface area burn (Ohashi et al., 2013)	Burn centre	7 (19 September 2011)	Wound	IPM-R, AMK-R, CPFX-R (RRR)	ER-c	Yes
5. Traffic injury	EICU	44 (23 September 2011)	Bile drain	IPM-S, AMK-R, CPFX-R (SRR)	ER-d	Yes
6. Traffic injury	EICU	13 (26 October 2011)	Wound	IPM-R, AMK-R, CPFX-R (RRR)	ER-d	Yes
7. Iliopsoas muscle abscess	EICU	45 (15 November 2011)	Sputum	IPM-S, AMK-R, CPFX-R (SRR)	OP	No
8. Subcortical haemorrhage	EICU	240 (15 November 2011)	Sputum	IPM-S, AMK-S, CPFX-R (SSR)	ER-d	No
9. Subarachnoid haemorrhage	EICU	50 (17 November 2011)	Sputum	IPM-S, AMK-S, CPFX-S (SSS)	NS	No
10. Air-fluidized bed	EICU	(20 November 2011)	Beads	IPM-S, AMK-R, CPFX-R (SRR)	ER-a, ER-b, ER-c, ER-d

Table 3.

Susceptibility patterns of A. baumannii Tokai strains

Strain	MIC (µg ml−1)
	β-Lactams								Aminoglycosides			Fluoroquinolones		Other agents
	IPM	PIPC	CAZ	CFPM	S/C	AZT	MEPM	CZOP	GM	TOB	AMK	LVFX	CPFX	MINO	FOM	S/T
1, 5, 7, 10	2	>64	>16	16	<16	16	>8	16	>8	>8	>32	4	>2	4	>16	>2
2, 3, 4, 6	>8	>64	>16	16	<16	16	>8	16	>8	>8	>32	>4	>2	≤2	>16	>2
8	≤1	≤8	≤2	<4	<16	8	≤1	4	≤1	≤1	≤4	>4	>2	≤2	>16	≤2
9	≤1	≤8	4	16	<16	8	≤1	8	8	2	8	≤0.5	1	≤2	>16	≤2

S/C, sulbactam/cefoperazone; S/T, sulfamethoxazole/trimethoprim.

As the interval between the first patient and the others was long (>18 months), the environment of the ward was suspected to be a possible reservoir of the pathogen. Based on the results of the bacteriological surveillance of environmental surfaces, A. baumannii Tokai strain 10 was isolated from the cracks of a rubber frame and a lump of beads in an air-fluidized bed that was used by five patients during their hospitalization (A. baumannii Tokai strains 1, 2 and 4–6).

Molecular typing

The molecular genotyping of isolates by a MLST analysis revealed a sequence type of ST208 for A. baumannii Tokai strains 1–8 and 10 (ST profile, gltA-gyrB-gdhB-recA-cpn60-gpi-rpoD: 1-3-3-2-2-97-3) and another type for A. baumannii Tokai strain 9 (ST profile, gltA-gyrB-gdhB-recA-cpn60-gpi-rpoD: 15-48-58-42-36-54-41). The molecular genotyping of isolates by rep-PCR showed the same pattern (>97 % similarity) as one type for eight of the isolates (A. baumannii Tokai strains 1–7 and 10) (Fig. 1), and the other isolates (A. baumannii Tokai strains 8 and 9) had different patterns (85 and <70 % similarity, respectively).

Fig. 1.

Results of the rep-PCR analysis and MLST in clinical isolates of A. baumannii. A. baumannii Tokai strains 1–7 and 10 showed identical patterns and homologous rates of identity >97 %. Two isolates, A. baumannii Tokai strains 8 and 9, had different patterns (85 and <70 % similarity, respectively). MLST analysis revealed that A. baumannii Tokai strains 1–8 and 10 were of the same sequence type (ST208).

Expression of resistance-related genes

The MBL assay of the clinically isolated A. baumannii Tokai strains revealed no apparent MBL production and all isolates showed expression of OXA-51-like carrying OXA-66 β-lactamase (Table 4). The expression of IMP-1, VIM, OXA-23-like, OXA-24-like and OXA-58-like was negative. Expression of ISAba1 and armA was found in A. baumannii Tokai strains 1–7 and 10. The DNA sequencing of gyrA and parC revealed that Ser83 (TCA) was changed to TTA (Leu) and that Ser80 (TCG) was changed to TTT (Phe) or TTG (Leu) in A. baumannii Tokai strains 1–8 and 10.

Table 4.

Expression of resistance-related genes as assessed by PCR and qRT-PCR in A. baumannii Tokai strains

Strain(s)	Susceptibility pattern	Gene expression							Mutation
		OXA-type β-lactamase					ISAba1	armA	gyrA (Ser83)	parC (Ser80)
		OXA-23-like	OXA-24-like	OXA-51-like	OXA-58-like	OXA-66
1, 5, 7, 10	SRR	−	−	+	−	+	+	+	Leu	Leu
2, 3, 4, 6	RRR	−	−	+	−	+	+	+	Leu	Leu
8	SSR	−	−	+	−	+	−	−	Leu	Phe
9	SSS	−	−	+	−	+	−	−	Ser	Ser

Our analysis of genes encoding RND pumps included an analysis of the expression of three previously characterized genes, adeB, adeG and adeJ, which encode the RND pump in the adeABC, adeFGH and adeIJK operons, respectively. The result of A. baumannii Tokai strains 1, 2, 8 and 9 as representative strains from each group with the same susceptibility pattern is shown in Table 5. Overexpression of adeB and adeJ was seen in A. baumannii Tokai strain 2. The expression of oprD was decreased in A. baumannii Tokai strains 2 and 8. Underexpression of carO was seen in isolates with A. baumannii Tokai strains 1, 2 and 8.

Table 5.

Relative expression of efflux pumps and outer-membrane porins in A. baumannii Tokai strains by qRT-PCR

The results for A. baumannii Tokai strains 1, 2, 8 and 9 are shown as a representative strain from each group with the same susceptibility pattern.

Strain	Susceptibility pattern	Relative expression
		Efflux pump (ratio)			Outer-membrane porin
		adeB	adeG	adeJ	oprD	carO
1	SRR	0.91	0.46	0.94	1.23	0.02
2	RRR	2.28	1.02	2.41	0.49	0.01
8	SSR	0.10	0.81	0.84	0.88	0.003
9	SSS	1.00	1.00	1.00	1.00	1.00

Discussion

We investigated a series of clinical isolates of DR A. baumannii ST208 in the EICU of Tokai University Hospital. In order to elucidate the diversity of the drug resistance patterns in the same sequence type in these isolates, we studied the molecular characteristics of these isolates and their relationship with the resistance pattern.

A. baumannii Tokai strains 1–7 and 10 were positive for OXA-51-like and OXA-66 β-lactamase and ISAba1. A. baumannii strains with resistance to AMK (A. baumannii Tokai strains 1–7 and 10) were positive for armA. These results are consistent with the idea that ISAba1 regulates the expression of OXA-51-like carrying OXA-66 β-lactamase and that armA is related to aminoglycoside resistance. The five isolates (A. baumannii Tokai strains 1, 2 and 4–6) could have been derived from the same source and/or transmitted horizontally, because the same air-fluidized bed had been used by those patients. Among them, A. baumannii Tokai strains 2, 4 and 6 showed multidrug resistance (RRR). These patients were treated with carbapenem (MEPM or DRPM) prior to sampling for at least 1 week, which may have played a role in the overexpression of adeB and adeJ in A. baumannii Tokai strain 2. A. baumannii Tokai strain 10 was detected from the cracks of the rubber frame and a lump of beads in an air-fluidized bed, even though the bed had been cleaned and disinfected every time after use. Although a few nosocomial outbreaks of A. baumannii ST2 have been reported (Suzuki ; Yamada & Suwabe, 2013), an outbreak of A. baumannii ST208 has not been reported previously in Japan.

As the pattern of the rep-PCR and sequence type of MLST in the eight isolates was the same as that in the initial case, it was suggested that the strain survived for 1.5 years in the environmental reservoir. As infection control procedures, careful attention to environmental cleaning and disinfection in order to reduce the risk of transmission is suggested. A. baumannii Tokai strain 8 (SSR) was also ST208, but had a different pattern as shown by rep-PCR. During the transmission from the same original organism, the presence of a transposon or the insertion of a different plasmid might have led to the different pattern. During the period of an outbreak, A. baumannii with different drug susceptibility patterns appeared depending on the various resistance mechanisms.

Nine isolates (A. baumannii Tokai strains 1–8 and 10) had resistance to CPFX, which can be explained by the mutations of gyrA and parC. Another major factor contributing to the resistance of this organism was the overexpression of the RND pumps (Fernando & Kumar, 2012; Amin ; Zander ). Our analysis of genes encoding RND pumps included the expression of three previously characterized genes, adeB, adeG and adeJ, which encode the RND pumps in the adeABC, adeFGH and adeIJK operons, respectively. Efflux pumps such as AdeABC have been reported to be involved in multidrug resistance (Vila ; Hou ). In our study, A. baumannii Tokai strain 2 (RRR) showed overexpression of adeB and adeJ. A. baumannii Tokai strain 8 showed better sensitivity to some β-lactams (CAZ, CFPM and CZOP) than that of A. baumannii Tokai strain 9 (SSS). This phenomenon might be associated with underexpression of adeB. Two pumps, such as adeB and adeJ, have been related to the acquisition of multidrug resistance. As for porins, the overexpression of genes encoding RND pumps and the downregulation of genes encoding porins is known to be common in clinical isolates of Acinetobacter spp. (Fernando et al., 2013). Our findings also suggest that the underexpression of carO in combination with or without oprD does not result in resistance to carbapenem in A. baumannii Tokai strains 1 and 8 (SRR and SSR). This observation is consistent with previous findings showing that a decrease in porins among Acinetobacter strains is not associated with resistance to carbapenems in the presence of β-lactamases (Rumbo ; Singh ).

In conclusion, we demonstrated that drug resistance is associated with the expression of ISAba1 and armA, and mutations in gyrA and parC, and that the overexpression of adeB and adeJ plays a role in the multidrug resistance of A. baumannii Tokai strain ST208.