High prevalence of oxacillinases in clinical multidrug-resistant Acinetobacter baumannii isolates from the Tshwane region, South Africa – an update

Background Acinetobacter baumannii is an important hospital-acquired pathogen in healthcare facilities that frequently causes bacteraemia and ventilator-associated pneumonia in intensive care units. Acinetobacter baumannii can be isolated from various sites in the hospital environment like medical equipment, bed linen, medical personnel and indwelling catheters. It is difficult to treat A. baumannii infections because of their highly resistant antimicrobial profiles. The purpose of this study was to determine the prevalence of β-lactamase genes in multidrug-resistant (MDR) clinical A. baumannii isolates using Multiplex-PCR (M-PCR) assays. Methods One hundred MDR A. baumannii isolates were collected from the diagnostic division of the Department of Medical Microbiology after routine analysis of the submitted specimens. All collected isolates were identified and tested for susceptibility using the VITEK 2® system (bioMérieux, France). Six isolates were excluded from this study because the isolates were incorrectly identified as A. baumannii with the VITEK 2® system (bioMérieux, France). Molecular tests, namely M-PCR assays, pulsed-field gel electrophoresis (PFGE) and multilocus sequence typing (MLST) were performed. MLST analyses were performed on representative isolates from the four major pulsotypes (≥5 isolates with 80 % similarity) and selective isolates from each minor pulsotype. Results All the A. baumannii isolates showed 100 % resistance to ampicillin, amoxicillin, cefuroxime, cefuroximine axetil, cefoxitin, cefotaxime and nitrofurantoin. Seven percent of the isolates were resistant to amikacin. Two percent of the isolates were classified as having intermediate susceptibility to tigecycline. A. baumannii isolates showed an antibiotic resistance profile of 67 % and higher to antibiotics, such as ceftazidime, cefepime, imipenem, meropenem, gentamicin, ciprofloxacin and trimethoprim/sulfamethoxazole. None of the isolates were resistant to colistin. The M-PCR assays showed that 99 % of the isolates contained the OXA-51 gene and 77 % contained the OXA-23 gene. None of the isolates contained the GES, GIM, IMP, KPC, NDM, OXA-24, OXA-58, PER, SIM, SPM, VEB and VIM genes. Representative A. baumannii isolates were grouped into five existing sequence types (ST): ST106, ST258, ST339, ST502, ST758 and ST848. Isolates belonging to the pan-European clonal lineages I and II (EUI and EUII) were identified. Conclusion The high prevalence of MDR A. baumannii isolates has a severe impact on available treatment choices and this in return impacts on treatment outcomes in the studied healthcare facilities. The most dominant ST among the collected isolates was ST758, member of the EUI group. The presence of the OXA-23 gene was not restricted to a specific ST. Continuous research and surveillance is necessary to monitor the circulating β-lactamase genes in clinical settings to guide infection control policies in order to try and curb the spread of this bacterium.


Background
Acinetobacter baumannii is an emerging opportunistic pathogen that can become resistant to multiple antimicrobial agents [1][2][3]. Over the past 20 years A. baumannii's clinical importance has been driven by its ability to obtain antimicrobial resistance factors [4,5]. Resistance is acquired through the transfer of integrons, plasmids or transposons that carry groups of genes encoding resistance to several antibiotic families [6]. This bacterium can survive under a wide range of environmental conditions and on surfaces for extended periods of time and this can lead to both endemicity and outbreaks in healthcare facilities [4,7].
This pathogen possesses different antimicrobial resistance mechanisms, which include the enzymatic inactivation of β-lactam antibiotics (cephalosporins, carbapenems and monobactams) through the production of extended spectrum β-lactamases, carbapenemases and AmpC-type (ampicillin class C) enzymes [12][13][14][15]. The β-lactamases of Gram-negative bacteria belong to Ambler classes A to D [12,13]. The inherent class D oxacillinases (OXA) of A. baumannii belong to the chromosomally encoded OXA-51-like enzyme group [16][17][18]. However, the OXA-51 enzyme has been under considerable pressure due to antibiotic use and has evolved and now plays an important role in the carbapenem resistance of A. baumannii isolates [19]. If ISAbaI is located upstream of the bla OXA genes the A. baumannii isolate will have an increased OXA-51 expression which leads to carbapenem resistance (imipenem and meropenem) [19][20][21].
Infections caused by A. baumannii are difficult to treat because of the resistance to antimicrobial agents, especially β-lactams [3,22]. There have been reports of MDR A. baumannii from hospitals all around the world (Argentina, Brazil, China, Europe and the United Kingdom) [4,23,24]. Only a few antimicrobial agents namely colistin (in combination with carbapenems) and tigecycline are available for the treatment of MDR A. baumannii [1,2,15]. However, there are now reports of colistin resistance in A. baumannii isolates, which is of great concern as colistin plays a pivotal role in MDR A. baumannii treatment [25][26][27][28]. Fortunately colistin resistance in A. baumannii is still rare [21,25].
The aim of this study was to detect the prevalence of β-lactamase genes and to determine the genetic relationship among the A. baumannii isolates circulating in our clinical setting. The prevalence of β-lactamase genes is important for the surveillance of circulating A. baumannii strains as it provides more information for the infection control policies of the hospitals.

Study design and setting
The quantitative study was aimed to determine the prevalence of β-lactamase genes in consecutively collected A. baumannii isolates obtained from clinical specimens and to determine the genetic relatedness of circulating strains. The study was conducted at the Department of Medical Microbiology, Prinshof Campus, University of Pretoria/National Health Laboratory Service (NHLS) and received ethical approval from the Research Ethics Committee, Faculty of Health Sciences, University of Pretoria (121/2013); a waiver for informed consent was provided.

Collection, culturing and storage of A. baumannii clinical isolates
One hundred consecutive clinical A. baumannii isolates were collected without prior knowledge of patient information, hospital ward, specimen type and antibiotic susceptibility status. All 100 A. baumannii isolates were collected from the diagnostic laboratory during May to July 2013. The laboratory processes specimens from a tertiary academic hospital as well as district hospitals and various clinics as part of standard care. Routine analysis was performed as requested on the form and according to sample type. The isolates were identified and tested for susceptibility using the automated VITEK® 2 system (BioMérieux, France) with the VITEK® 2 GN card and the VITEK® 2 AST-N255 card. Gender, age, type of specimen, hospital, ward and the antibiotic susceptibility profiles were collected.
One colony of an isolate (obtained from a blood agar plate) was inoculated into 5 ml tryptone soya broth (TSB) (Oxoid, UK) and incubated on a shaking incubator (Vacutec, South Africa) at 35°C for a period of 18 to 24 h. Growth was indicated by turbidity in the broth. Isolates were stored by adding 900 μl of 50% glycerol (Merck, Germany) to 900 μl of the overnight TSB culture (Oxoid, UK) in 2 ml cryotubes (Greiner Bio-One, Germany) in a one to one ratio. This was done in triplicate and the isolates were stored in a -80°C freezer (New Brunswick, USA). The quality control strains used were the multidrug resistant A. baumannii ATCC® BAA-1605 strain; Klebsiella pneumoniae ATCC 8303 and K. pneumoniae ATCC BAA-2146.

DNA extraction of A. baumannii isolates
The genomic DNA of each of the 100 A. baumannii isolates was extracted from a single colony grown in TSB (Oxoid, UK) culture media by using the ZR Fungal/Bacterial kit [29]. The manufacturer's instructions were followed with some modifications; instead of 50 to 100 mg of cells, 1 000 μl of overnight broth was used, centrifuged (Eppendorf 5417C, Germany) and the pellet was resuspended in phosphate buffered saline (PBS) solution (pH 7.2, Gibco® PBS, Life Technologies Corporation, USA) and 600 μl of the lysis solution were used instead of the prescribed 750 μl. The extracted genomic DNA was stored at -20°C (Defy Ltd, Multimode, SA) until further analysis.
The detection of β-lactamase genes in A. baumannii by using M-PCR assays Multiplex PCR assays were performed by using the QIAGEN® M-PCR kit (QIAGEN®, Germany) for the detection of several antibiotic resistance genes in A. baumannii (Table 1). An initial denaturation step of 95°C for 15 min, followed by 30 cycles of denaturation at 94°C for 30 s, an annealing temperature dependent on the melting temperature of the primer pair (multiplex I: 52°C; multiplex II and III: 57°C and multiplex IV: 60°C) and extension at 72°C for 90 s, followed by the final extension step at 70°C for 10 min. A negative control (sterile ultrapure water) was included for all M-PCR assays. The following positive controls were used for multiplex I to IV: A. baumannii ATCC® BAA-1605™ strain; K. pneumoniae ATCC 8303 [Klebsiella pneumoniae carbapenemase (KPC) positive control] and K. pneumoniae ATCC BAA-2146 [New Delhi metallo-β-lactamase (NDM) positive control]. All isolates testing negative for the presence of the inherent OXA-51 gene were sent for identification using matrix assisted laser desorption time of flight mass spectrometry (MALDI-TOF). A ready-to-use 100 bp DNA ladder (Fermentas, Thermo-Scientific, USA) was used to determine the approximate sizes of the amplicons. Gel electrophoresis (Bio-Rad, Germany) was performed at 90 V for 100 min with 1x Tris-Borate-EDTA (TBE) buffer [40 mM Tris-HCl, 20 mM Boric acid (Sigma-Aldrich, USA) and 1 mM EDTA (Sigma-Aldrich, USA), pH 8.0]. The bands were visualised and captured using a UVP Doc It transilluminator (Ultra-violet Products Incorporated, USA).

Determination of the genetic relationship of clinical A. baumannii isolates using the PFGE assay
The Pulsenet PFGE protocol [30] was followed with some modifications. The optical density (OD) was measured at 630 nm. The absorbance was adjusted to an OD range of 1.4 to 1.6 by using the ELx800 absorbance microplate reader (Bio-TEK, USA). The plugs were incubated overnight at 51°C and after incubation the plug was transferred to a new BD falcon tube (BD Biosciences, USA). The ApaI enzyme (New England Biolabs, USA) was used and the Salmonella choleraesuis serovar Braenderup (ATCC BAA-664) was used as a marker. Gel electrophoresis was performed using the following set conditions: interval 25 linear 5; angle 120°constant; voltage 220 V linear to 200 V and ran for 24 h. After the PFGE run was completed, the 1.2 % (m/v) SeaKem® LE agarose (Lonza, USA) gel was stained in 1 L distilled water, containing 250 μl EtBr (10 μg.ml-1) for 15 min and was destained in distilled water for 30 min. The gel was viewed and documented with a UV Doc It transilluminator (Ultra-violet Products Incorporated, USA). The banding patterns were analysed using the GelCompar II software programme (Applied Maths, Belgium). A distance matrix was constructed using the Dice coefficient and a dendogramme was constructed using the unweighted pair group method with arithmetic mean (UPGMA) [31]. Pulsotype designation was based on isolates showing ≥80 % relatedness, which corresponds to the Tenover criteria of possibly related 4 to 6 band differences between isolates [32,33]. Representatives from each major PFGE pulsotype (≥5 isolates) and selective minor pulsotypes (< 5 isolates) with ≥80 % similarity were chosen for MLST analyses.
All seven housekeeping genes were individually amplified in a 50 μl reaction as described in the Bioline kit's guideline (Bioline, UK). The following programme was used on the thermal cycler (G-storm Model GS4822, Labtech, UK): an initial denaturation step of 94°C for 2 min, followed by 35 cycles of denaturation at 94°C for 30 s, an annealing temperature at 50°C for 30 s, extension at 72°C for 30 s, followed by the final extension step at 72°C for 5 min [34]. The amplicons were visualised on a 1.8 % (m/v) MetaPhor agarose gel (Lonza, USA) as described previously. All the amplicons were sequenced in both forward and reverse directions by Inqaba Biotechnical Industries, South Africa. An ABI file obtained for each sequence were analysed with the CLC Main Workbench Version 6.0 (CLCbio, USA). The sequences were assigned to corresponding allelic profiles and sequence types by using the MLST Oxford database (http://pubmlst.org/ abaumannii/).

Data analysis
One hundred consecutive A. baumannii isolates were collected, which were identified by the automated VITEK® 2 system (BioMérieux, France). Six isolates were excluded from this study because the isolates were incorrectly identified as A. baumannii by the VITEK® 2 system. These isolates were identified by MALDI-TOF analysis as: A. pittii; K. pneumoniae; Pseudomonas putida (2x); Staphylococcus aureus and S. epidermidis.
The results are reported on 94 clinical A. baumannii isolates. The proportions of A. baumannii isolates with Ambler classes A, B and D resistance genes were determined for the 94 isolates. The results were reported as proportions or percentages along with 95 % confidence  intervals for the given sample size with an accuracy of within 10 %. The antimicrobial susceptibility and genotypic results obtained in this study were compared to a study done in 2008 in the same clinical setting [5]. The antimicrobial susceptibility profiles were also compared to the surveillance data from 12 sentinel public hospitals in South Africa (SA) [35].

Collection of A. baumannii clinical isolates
Fifty-eight of the isolates were collected from male patients and 36 isolates were collected from female patients. Forty-three percent of the isolates were collected from diverse sites, 4 % from urine, 41 % from sputum and 12 % from blood cultures. The mean age of the patients was 39 years (ranging from 4 days to 77 years). A. baumannii was isolated from various wards but was the most prevalent in the medical and pulmonology ICU (12 %). Age and sex is not a cofounder and hence the odds ratio for area need not be adjusted. The crude odds ratio is 1.61 with a 95 % confidence interval (0.60; 4.31) and is not significant (p = 0.345).
The 94 A. baumannii isolates showed an increase in antimicrobial resistance profiles compared to the study done by Kock et al. [5] Figure 1 shows the overall percentage of resistance in 2008 and 2013 in the Pretoria region. Data from the communicable diseases surveillance bulletin [35] showed that the A. baumannii isolates collected in 2013 in public hospitals in SA have resistance towards carbapenems and cephalosporins (imipenem 77 %; cefepime 80 % and ceftazidime 73 %).
The results of the M-PCR assay I revealed the presence of two resistant profile types of A. baumannii that were circulating in the studied clinical setting. Type 1 was positive for both OXA-51 and OXA-23 (77 %) and type 2 was only positive for OXA-51 (22 %).

Discussion
The increase in resistance profiles of A. baumannii can be due to the misuse of antibiotics in clinical settings, or because of patients who do not complete the prescribed antibiotic course. The bla OXA genes have an impact on treatment because it can already be resistant to the new inhibitors, such as β-lactamase inhibitors, or it can rapidly gain resistance [19].
The ability of the OXA β-lactamases to confer carbapenem resistance has already had a huge impact on the ability to treat Gram-negative pathogens and the current situation suggests that the problem is only set to increase [19]. The advantage of using molecular tests, such as the M-PCR assays used in this study, is that resistance genes circulating in clinical settings can be determined as this is not possible with phenotypic tests. Through molecular methods this study has indicated that 99 % of the A. baumannii isolates contained the OXA-51 gene, compared to Kock et al. [5] where only 81% out of the 97 isolates contained this gene. In the Kock et al. [5] study, the species identification obtained by the VITEK® 2 system was not confirmed by an alternative method, such as MALDI-TOF MS and thus possibly includes the other species of the A. baumannii-calcoaceticus complex. The results obtained in the current study shows the establishment of the species, A. baumannii in our clinical setting. The OXA-51 cluster is chromosomally encoded and is inherent to all A. baumannii isolates [4,16,19,23,35]. Seven isolates (of which six isolates were excluded) did not have the inherent OXA-51 gene. These isolates were further analysed using MALDI-TOF, which identified the seven isolates as: A. baumannii; A. pittii; K. pneumoniae; Pseudomonas putida (2x); Staphylococcus aureus and S. epidermidis. Studies have shown that the detection of the bla OXA-51-like gene can be used as a simple and reliable way of identifying A. baumannii [36,37]. Although it is clear that bla OXA-51-like genes are present in at least the vast majority of isolates of A. baumannii, there has been some debate as to whether they are present in all isolates of this species [36]. Zander et al. [38] reported that three A. baumannii isolates did not harbour the predicted bla OXA-51-like gene (353 bp) and instead amplified PCR products of 1.2 kb and 1.6 kb. The authors then used external primers and the amplicons revealed that the one isolate (Ab-511) possessed the bla OXA-51 variant bla OXA-66 , which was interrupted by the insertion sequence ISAba15 at nucleotide number 435, while the two other isolates (Ab-508 and Ab-653) possessed the bla OXA-51 variant bla OXA-78 , which was disrupted at nucleotide number 379 by a novel insertion sequence ISAba19 [38]. Further investigation of the genomic environment of the isolate identified as A. baumannii by MALDI-TOF will reveal the changes in the OXA-51 genetic structure. Additionally the 16S-23S ribosomal spacer region can be sequenced for further species confirmation [37].
The occurrence of the OXA-23 gene has been reported since 1985 in A. baumannii isolates and has now disseminated worldwide [5,19,39,40]. The prevalence of OXA-23 in this study was 77 %, compared to Kock et al. [5], who reported a 59 % prevalence in the same clinical setting in 2008. It is clear that the prevalence of OXA-23 in A. baumannii isolates in this clinical setting is increasing. This increased occurrence can be due to poor infection control practices, spread through contact with healthcare workers or from patient to patient and prolonged antibiotic use [11]. Stricter adherence to infection prevention and control policies by healthcare workers are needed to prevent further dissemination [4,41,42]. Correct use of disinfectants and solutions containing ethanol is particularly important because if applied at very low concentrations A. baumannii bacterium can become more pathogenic [41].
The high prevalence of OXA-23 can be due to the acquisition of genetic elements, such as plasmids and transposons [1,42]. Liakopoulos et al. [43] conducted a study in 2010 to 2011 and reported a 95 % prevalence of  OXA-23 in Greece. The OXA-23 cluster can be located on a plasmid or chromosome [5]. Mugnier et al. [44] reported in 2010 that eleven isolates had the bla OXA-23 gene on the chromosome, nine isolates had the bla OXA-23 gene on a plasmid and one isolate had two copies of the bla OXA-23 gene, one on the plasmid and one on the chromosome. The OXA-23 enzyme has a higher affinity to hydrolyse imipenem than meropenem, ertapenem or doripenem [19]. The MDR A. baumannii isolates in this study that contained the OXA-23 gene showed a 69 % resistance to imipenem (1 % had intermediate susceptibility and 6 % were susceptible to imipenem). Imipenem susceptibility may be explained by the absence of the ISAbaI element upstream of the bla OXA-23 gene [45]. Twenty-two percent of the isolates were positive for the OXA-51 gene only. However, 17 % of these  isolates were resistant to imipenem while 5 % were susceptible.
The prevalence rates of isolates that contained both the OXA-23 and OXA-51 genes were 77 %. It is evident that the carbapenem-hydrolysing class D β-lactamases (CHDL's), such as OXA-23 and OXA-51 were prevalent in clinical A. baumannii isolates in this study. The high prevalence of the OXA-51 and OXA-23 group (77 %) is due to the spread and establishment of several successful clones in our clinical setting.
None of the clinical A. baumannii isolates contained the genes that were screened for in M-PCR II to IV as well as the OXA-24 and OXA-58 genes that were part of M-PCR I. The GIM, SIM and SPM metallo-β-lactamase (MBL) genes are rarely found in A. baumannii isolates [46]. Kock et al. [5] reported one A. baumannii isolate that was positive for the VIM-like gene. Safari et al. [47] reported that 99 % of the isolates in ICU wards of three educational hospitals of Hamadan City in Iran produced metallo-β-lactamases. In 2013, the first outbreak of NDM-1 producing A. baumannii was reported in France, which occurred in a surgical ICU [48]. Fortunately NDM-1 in A. baumannii has not yet been detected in our clinical setting. The NDM-1 gene encodes an MBL carbapenemase that has a high hydrolytic activity for carbapenem antibiotics, and much more [49].
PFGE is suited for studying isolates from a defined temporal and spatial epidemiologic setting, whereas MLST has an advantage in defining clonal lineages of isolates from larger geographic areas over time [50]. Three clonal lineages of A. baumannii, commonly referred to as the pan-European clonal lineages (EUI, EUII, and EUIII), have predominated in many European countries since the 1990s [50]. Isolates belonging to the pan-European clonal lineages I and II (EUI and EUII) were identified in our clinical setting. EUI (ST258 and ST758) was predominant together with the OXA-23 gene. Other diverse clones, such as ST106 and ST339 are also circulating in our clinical setting. The origin of the A. baumannii isolates in our setting cannot be attributed to the dissemination of one or two successful clones but rather to the dissemination of a diverse group of successful clones.

Conclusions
The increasing resistance to carbapenems are of great concern due to the limited number of effective drugs and the Isolate ST blaOXA-23 Fig. 3 PFGE patterns of A. baumannii isolates. The bandings patterns were analysed using the GelCompar II software programme (GelCompar II, Applied Maths, Belgium). A distance matrix was constructed using the Dice coefficient and a dendogramme was constructed from the distance matrix using UPGMA. Major pulsotype designation was based on five or more isolates showing ≥80 % relatedness. Isolate name, ST and the carriage of bla OXA-23 (Y, positive; N, negative) are indicated association with increased morbidity and mortality. There was a noticeable change in the resistance profiles of the A. baumannii isolates from 2008 to 2013 in the same setting. The high prevalence of MDR A. baumannii isolates has a severe impact on treatment outcomes in the studied healthcare facilities. The presence of the OXA-23 gene was not related to a specific ST and can be due to the acquisition of genetic elements, such as plasmids and transposons. OXA-23 and ST758 were dominant among the isolates. A limitation of this study is the small sample size and only tertiary academic hospitals in the Pretoria region were included and thus the findings in this study are not a true reflection of the situation in the whole of SA. Another limitation of this study is that there were not any positive controls included for the screened genes in M-PCR assay III, which could have caused that these genes were missed due to not using the optimal PCR conditions. It is vital to identify and characterise healthcareassociated pathogens and their circulating antibiotic resistance genes as it is of great concern to public health systems. Understanding the resistance genes can lead to the understanding of their origin and can assist in better infection control measures and ultimately limit the spread of these pathogens.