FIRST ISOLATION OF CARBAPENEM-RESISTANT ACINETOBACTER BEIJERINCKII FROM AN ENVIRONMENTAL SAMPLE

The emergence of opportunistic Acinetobacter spp. in healthcare settings poses a signi ﬁ cant threat to public health. The major reasons for nosocomial spread of these species are their abilities to develop and transfer drug resistance against various classes of antibiotics. Considering that Acinetobacter spp. are ubiquitous in nature, can utilize several carbon sources, and reach humans via various pathways, our aim was to obtain information about the environmental strains of this genus. Our ﬁ rst step was to develop and test a multistep isolation procedure based on traditional scienti ﬁ c methods. Antibiotic resistance patterns of the isolated strains were determined, as susceptibility to 12 antibiotics of 7 classes was tested by MIC Test Strip method. Altogether 366 samples (groundwater, surface water, and soil) of 24 sites were investigated and a collection of 37 Acinetobacter isolates was obtained. Among others, clinically important human pathogen Acinetobacter spp., such as A. baumannii, A. johnsonii , and A. gyllenbergii were identi ﬁ ed. Three environmental strains were determined as multidrug-resistant including a carbapenem-resistant, hemolytic Acinetobacter beijer-inckii strain isolated from a hydrocarbon-contaminated groundwater sample. In summary, it has been found that the applied multistep isolation procedure is applicable to isolate various species of Acinetobacter genus. Based on the antibiotic resistance assay, we can conclude that environmental representatives of Acinetobacter spp. are able to develop multidrug resistance, but at a lower rate than their clinical counterparts.


Introduction
The necessity of simultaneous investigations of environmental and clinical isolates of opportunistic species is a rising issue in our days, since pathways between clinical and non-clinical environments are both globally and locally opening.In the case of some clinically important species such as Pseudomonas aeruginosa, comparative studies are available regarding virulence factors, genomic diversity [1], or antibiotic resistance of strains isolated from various environmental sources.However, Acinetobacter strains of clinical and environmental habitats are still considered and examined in various ways.
Acinetobacter species are ubiquitous in natural habitats (soil and water).Their presence was verified in the normal skin, throat, and rectal flora as well as in food and body lice [2].The representatives of this genus can cause serious respiratory, urinary, and biliary tract infections; empyema; bacteremia; meningitis; endocarditis; burn infections; and wound sepsis [3,4].Acinetobacter strains related to nosocomial infections are mainly found in intensive care units and in burns or high dependency units treating severely ill or debilitated patients [5].In addition, infections caused by Acinetobacter baumannii have been reported in the case of military personnel of combat regions such as Afghanistan or Iraq [6].
Among all Acinetobacter species, the Acinetobacter calcoaceticusbaumannii complex (ACB) has the greatest influence on clinical healthcare; therefore, ACB is the critical subject of the current scientific investigations.The members of the ACB complex (A.baumannii, A. calcoaceticus, Acinetobacter pittii, and Acinetobacter nosocomialis) are in close phylogenetic relation; therefore, their differentiation solely on phenotypic characterization is difficult [7,8].
Acinetobacter-related infections can pose serious clinical concerns with high mortality of patients.Due to the high rate of multidrug-resistant (MDR) Acinetobacter strains in clinical settings [9], some Acinetobacter species such as A. baumannii are mentioned as "emerging nosocomial superbugs" [10].
Regarding antibiotic resistance features, the emergence of MDR Acinetobacter strains has been reported from hospital settings all over the world, and the number of infections caused by these strains is continuously increasing [2,5].The treatment of MDR Acinetobacter strains is complicated because (due to the production of OXA-type carbapenemases) the majority of clinical Acinetobacter spp. is resistant to carbapenems, a commonly used group of first-line antimicrobial agents [11].Moreover, some A. baumannii strains were reported to be resistant to all available antibiotics [12], due to their great set of various antibiotic resistance mechanisms, which lead to the emergence of untreatable infections.
Although clinical Acinetobacter spp.are well characterized, the available information regarding antibiotic resistance patterns of Acinetobacter strains from environmental samples is much limited and controversial, despite the recognition of environment as a potential reservoir for Acinetobacter spp.-resistant isolates [13].The major reason for this is possibly the difficult isolation and identification of ACB from environmental samples.Upon the information of scientific literature, the existing methods for the isolation of Acinetobacter spp.are moderately effective and are not standardized.
The available procedures for identification are also limited and occasionally are not verified by molecular genetic methods [13].Recent publications about Acinetobacter species used only classical microbiological methods or mass spectrometry [10,14] for identification; however, none of these research works used the reliable 16S rRNA-based molecular method for identification.
Despite the clinical concerns of acinetobacters, some Acinetobacter strains are considered as industrially important [15] due to their ability to degrade several pollutants or industrial products [16][17][18] such as diesel oil [19].However, the clinically important features of these strains such as antibiotic resistance or virulence are barely described.
Only a few scientific articles have discussed the antibiotic resistance of environmental Acinetobacter strains.Dhakephalkar and Chopade [20] examined different environmental sources (soil, river water, domestic sewage, treated industrial effluent, and laboratory effluent) to isolate Acinetobacter strains and to determine their antibiotic and metal resistance.Berlau et al. [21] isolated Acinetobacter strains from vegetables and fruits to determine minimum inhibitory concentrations (MICs) of six antimicrobial agents.
Based on the overview of the available scientific information on environmental acinetobacters, our aim was to widen our knowledge on antibiotic/ multidrug resistance of naturally occurring Acinetobacter isolates, which can be a possible source of non-nosocomial infections [22].To reach our main objective, we aimed to establish a culture collection of environmental Acinetobacter strains and to determine their antibiotic resistance profiles.

Isolation of Acinetobacter species
Successful isolation of Acinetobacter strains was carried out based on classical microbiological methods described in scientific literature.The recommended media were examined with two Acinetobacter strains (Acinetobacter beijerinckii and Acinetobacter johnsonii) formerly isolated in our department.The colony morphology of Acinetobacter spp. was examined in cetrimide agar, tryptone glucose yeast (TGY) agar, nutrient agar, McConkey agar [18] and Leeds Acinetobacter Medium [25,26] with different incubation parameters (28, 30, and 42 °C).Furthermore, Acinetobacter sp.enrichment liquid medium [27] was tested in combination with TGY and Leeds Acinetobacter Medium agar.
Based on these preliminary investigations, the following protocol was implemented in the case of environmental samples: each incoming sample was diluted in tenfold series with physiological saline solution (10 1 -10 4 ) and 1-1 ml of diluted samples were inoculated into tubes containing Acinetobacter sp.enrichment liquid medium (9 ml) [27].After 48 h of incubation (28-30 °C and 160 rpm) agarized Leeds Acinetobacter Medium [25] plates were poured containing 1 ml of the shaken enriched cultures.After another incubation period of 72 h, colonies resemble Acinetobacter spp.(mauve, round, intact edge, and shiny) on the surface of agar plates were purified using streak plate method and subsequently inoculated onto nutrient agar slants (MERCK 105450) for conservation and further investigations.

Identification of strains
Before 16S rRNA gene base identification, pure cultures of Acinetobacter candidates were examined with oxidase and catalase tests [28].Strains with oxidase-negative and catalase-positive profiles were used for further examinations (16S rRNA gene sequencing and antibiotic resistance investigations).

16S rRNA gene-based identification of strains
Acinetobacter candidates were identified based on 16S rRNA gene sequence analysis.The 16S rRNA gene was amplified from chromosomal DNA.DNA of strains from liquid cultures (24 h) was extracted and purified using the MOBIO Ultra Clean Microbial DNA Isolation Kit (MOBIO Laboratories, USA) following the manufacturer's instructions, detected by 1% agarose gel electrophoresis and visualized under ultraviolet light.
For amplification of 16S rRNA genes, universal 27f (5′ GAG TTG ATC CTG GTC AG 3′) forward and 1492r (5′ TAC GGG TAC CTT GTT ACG ACT T 3′) reverse primers were used [29].The reaction parameters were as follows: 98 °C for 5 s, 32 cycles of 94 °C for 30 s, 52 °C for 30 s, 72 °C for 45 s, and then a final extension at 72 °C for 10 min.PCR reactions were performed in 50 μl reaction volumes containing 5 μl Taq-Buffer (Fermentas, Lithuania), 10 μl (1 mM) dNTP, 0.25 μl Taq DNA polymerase, 0.5 μl (1 μM) forward primer (27f), 0.5 μl (1 μM) reverse primer (1492r), 1 μl DNA template, and 32.75 μl nuclease-free water.Amplicons were detected by agarose gel electrophoresis, purified by NucleoSpin Extract II DNA Clean-up set (Macherey-Nagel, Germany) and used as a template for sequencing reactions using 27f as sequencing primer.The nucleotide sequence determination was performed using the Big Dye Terminator version 3.1 (Applied Biosystems, USA).Cycle Sequencing Kit (Applied Biosystems) and sequences were analyzed using ABI 3130 Genetic Analyzer (Applied Biosystems).Prior capillary gel electrophoresis products of the sequencing reactions had been purified using ethanol precipitation method.The obtained sequences were edited and assembled using MEGA5 software [30], and then homology BLAST searches [31] were performed in the GenBank database.The determination of closest type strain was carried out by EzTaxon-e server [32,33].Species-level identification was accepted above 98% sequence homology.Sequencing data obtained in this study were submitted to the European Nucleotide Archive database and can be found under accession numbers HG810368-HG810404.

Antibiotic resistance examination of environmental strains
The antibiotic sensitivity tests of environmental Acinetobacter strains were performed according to the guidelines of the Clinical and Laboratory Standards Institute (CLSI) [34].
For the determination of MIC values, MIC Test Strip (Liofilchem, Italy) was applied with 12 antibacterial agents belonging to 7 antibiotic groups (Table II).In the first step, bacterial suspensions of 0.5 McFarland value [35] were spread onto the surface of Mueller-Hinton agar (MERCK 1.05435.0500) in accordance with the recommendation of the manufacturer.MIC Test Strips were positioned on the surface of the inoculated medium.Escherichia coli ATCC 25922 and P. aeruginosa ATCC 27853 were used for quality control of test strips.After an incubation period of 24 h at 35 °C, MIC values were determined and evaluated according to CLSI's recommendation.Strains were classified into resistant (R), intermediate resistant (I), and sensitive (S) categories.

Results and Discussion
As it was described in "Methods" section, a multistep isolation protocol was implemented upon the experienced effectivity of the examined media.The main steps of this selection procedure for acinetobacters are summarized in Figure 1.
The multiple, parallel checks of examined media showed that our protocol is suitable for selective isolation of environmental Acinetobacter strains with a rate similar to other available procedures [13], but without the addition of antibiotics, as selective agents.The main advantage of avoidance of antibiotics during isolation is their possible effect on the antibiotic resistance profile of a given strain; therefore, our procedure appears to be more reliable regarding antibiotic resistance profiling than other methods.On the whole, this protocol can be recommended for other researchers for further testing and adaptation to reach a good detection rate of environmental acinetobacters with a decreasing rate of false identifications and/or with a negative effect on antibiotic resistance profiling.
During our examination period, altogether 37 environmental strains belonging to the genus Acinetobacter were isolated with the aforementioned multistep protocol from groundwater (26 isolates), soil (3 isolates), and surface water -fishpond (8 isolates) samples (Table I).Among environmental isolates, the overwhelming occurrence of A. calcoaceticus-related strains (n = 20) was observed followed by representatives of species A. baumannii (n = 4), A. johnsonii (n = 4), A. pittii (n = 4), A. beijerinckii (n = 2), Acinetobacter gyllenbergii (n = 1), Acinetobacter radioresistens (n = 1), and Acinetobacter soli (n = 1).Fifteen strains were isolated from TPH-contaminated sites, 3 strains from PAH-contaminated soil, 1 strain from MTBE-contaminated groundwater, and 18 strains from non-contaminated samples (10 strains from groundwater and 8 strains from fishponds).As we know, Acinetobacter species can degrade different hydrocarbons [15-17, 36, 37], but based on our results, their adaptation and degradation abilities are not determining factors regarding their environmental frequency.About 75.6% of the identified Acinetobacter spp.were the members of the ACB complex with verified human health concerns and three of the remaining species (A.beijerinckii, A. johnsonii, and A. gyllenbergii) are considered as emerging human pathogens that may become increasingly important in the future [4].Surprisingly, the numbers of A. baumannii strains in the examined environmental samples were quite low, i.e., only 4 of 37 strains (Table I), while previous studies reported environmental A. baumannii in greater numbers [14,20,21].However, the 16S rRNA-based identification was missed from those papers and identification was based on phenotypic methods or matrix-assisted laser desorption ionization-time of flight mass spectrometry accepted with relatively low (2.005-2.418)score values.Considering the different identification procedures, it is possible that the real environmental frequency of A. baumannii is lower than it was previously expected, and our results reflect the proportion closer to the reality.
A. baumannii was detected mainly in fishponds (three of four strains), but the low detection rate does not let us draw general conclusions.Definite correlation between the occurence of A. baumannii and the level or type of the contamination of environmental samples could not be found.Further investigations may help us to inform about the effects of the available nutrients, oxygen, or other biogeochemical factors of the environmental samples on the frequency of A. baumannii.
The results of antibiotic susceptibility tests and the minimal inhibitory concentrations of antibiotics in the case of the examined environmental strains are summarized in Table III.
The results of the MIC test indicated that the majority of the investigated environmental acinetobacters were sensitive to the most widely used antibiotics.Imipenem, meropenem, gentamicin, ciprofloxacin, ofloxacin, and ceftazidime proved to be highly effective antibiotics, inhibiting the growth of nearly all tested acinetobacters (from 89% to 97%).
According to reports of National Center for Epidemiology (NCE) in Hungary, since 2010 there has been a steady increase in the prevalence of clinical acinetobacters resistant to imipenem, meropenem, gentamicin, ciprofloxacin, and ceftazidime [38].According to NCE, 49.4% of clinical ACB complex strains exhibited resistance to imipenem and 53% against meropenem in 2012.Approximately 20.1% of other clinical acinetobacters (other than the members of ACB complex) showed resistance against both tested carbapenems [39].On the contrary, 96.5% of the environmental strains with a closer phylogenetic relationship to the ACB complex were susceptible to the tested carbapenems.Solely A. pittii "II HT-1/1" strain (isolated from a fishpond sample) exhibited intermediate resistance against meropenem.Based on the NBS Annual Reports [39], ciprofloxacin resistance among clinical ACB complex strains and other clinical acinetobacters was 80.4% and 41.3% in Hungary, respectively.Conversely, 100% of our environmental ACB complex strains showed susceptibility to ciprofloxacin.Although there are no available data from 2012, in 2011, 84.5% of clinical ACB complex strains and 32.3% of other clinical acinetobacters exhibited resistance to ceftazidime [40].
In addition to the Hungarian reports of NCE, other European [41] and American [42] study groups observed a high rate of resistance among clinical Acinetobacter spp. to carbapenems (12% and 47%), to quinolones (51% and 67%), and to the third-generation cephalosporins (47% and 62%, respectively).Moreover, a recent study on A. baumannii originated from a sewage treatment plant verified that 82% of the isolates were resistant to carbapenems, fluoroquinolones, and colistin [14].Until the early 1970s, gentamicin (aminoglycoside) was successfully applied against nosocomial Acinetobacter infections [42].According to Towner [5], acinetobacters usually show a higher rate of aminoglycoside resistance than other pathogens, which can correlate with the results of NCE.In 2012, 69% of clinical ACB complex strains and 34.1% of other clinical Acinetobacter strains were resistant to gentamicin in Hungary [39], but our environmental strains showed nearly complete susceptibility to gentamicin (97.3%).
When the aforementioned antibiotics are not effective in nosocomial Acinetobacter-related infections, tigecycline is considered as the drug of choice [43].In this study, tigecycline was the sole antibiotic that inhibited the proliferation of all environmental Acinetobacter strains (100% efficacy).
Cefotaxime, ceftriaxone, cefepime, piperacillin, and rifampicin were the least effective antibiotics against the proliferation of the investigated environmental Acinetobacter strains.Acinetobacters obtained in this study widely exhibited intermediate resistance/resistance against the tested the third-generation cephalosporins (70% resistance to cefotaxime, 65% to ceftriaxone, and 32.4% to cefepime).Furthermore, 78% of strains were resistant to piperacillin and 81% to rifampicin.The rifampicin resistance of strains presented in this study is in contrast with the findings of Dhakephalkar and Chopade [20] who obtained rifampicin susceptibility higher than 90% among environmental acinetobacters.
Regarding multidrug resistance, A. baumannii that usually exhibits an elevated ability to rapidly develop antibiotic resistance [44] showed the lowest antibiotic resistance among all tested environmental strains (susceptibility against 10-12 antibiotics) followed by A. pittii and A. calcoaceticus (susceptibility from 7 to 10 antibiotics).
Environmental strains exhibiting simultaneous resistance against several classes of antibiotics were species A. beijerinckii, A. gyllenbergii, and A. johnsonii that are considered as emerging human pathogens [4].These isolates met the currently accepted definition of multidrug resistance [45]. A. beijerinckii "T3N6" isolated from TPH-contaminated groundwater was resistant to cephalosporins and to the tested carbapenem, too.Since carbapenems are often considered as antibiotics of "last resort" in treatment of MDR Acinetobacter [46], the occurrence of a multidrug-and carbapenem-resistant environmental A. beijerinckii isolate, whose clinical counterparts may cause endocarditis [4], can be worrying.To the best of our knowledge, the antibiotic resistance of A. beijerinckii against carbapenems and cephalosporins has never been reported earlier (neither among nosocomial nor among environmental strains).
A. johnsonii strain "6/1" (Table I) isolated from an overwintering fishpond showed the second highest rate of resistance to five antibiotics: ciprofloxacin, cefotaxime, ceftazidime, ceftriaxone, and piperacillin.Therefore, strain "6/1" is considered as MDR. A. gyllenbergii "6/2" (originated from the same fishpond as strain "6/1") showed resistance against cephalosporins, piperacillin, and exhibited intermediate resistance to ofloxacin; therefore, it is also considered as an MDR strain.Based on the overview of scientific sources, the resistance of A. johnsonii strains against ciprofloxacin, cephalosporins, and piperacillin has never been reported earlier.
In comparison with clinical isolates, it can be summarized that the occurrence of antibiotic resistance among terrestrial and aquatic representatives of the genus Acinetobacter does not reach the clinical level, but it needs more attention and continuous monitoring.The detected rates of antibiotic resistance among environmental strains can be explained with several reasons such as natural resistance or horizontal gene transfer with other naturally occurring bacteria.The antibiotic residues in the environment and the effect of the misuse or overuse of antibiotics may also cause problems among environmental strains.However, we did not have the chance to examine these scenarios.Further investigations of fishponds with antibiotic treatment or communal, industrial effluents with antibiotic residues may help us to clarify this possibility.Wastewaters can be important sources of MDR bacteria and antibiotic residues in the environment because antibiotics are not completely metabolized by the human body [47].Since modern wastewater treatment plants are still not designed to remove antibiotic residues found in domestic wastewaters [48], and urban sewage system can be considered as the secondary habitat of A. baumannii [14], the active compounds of antibiotics in the form of micropollutants may trigger the evolution of drug-resistant pathogens.
Regarding environmental contamination, we found that hydrocarboncontaminated environments (soil and groundwater) may function as reservoirs or incubators for antibiotic-resistant acinetobacters.Considering that catabolic and resistance genes are often located on the same mobile genetic elements (plasmids) and a hydrocarbon pollution may act like an impulsive force for horizontal gene transfer [49], the prevalence of MDR acinetobacters within contaminated environments is still a possibility.However, the low number of isolated ACB strains cannot let us draw general conclusions.

Conclusions
This study provides a new, multistep method for the effective and selective isolation of environmental Acinetobacter species, which was checked by molecular genetic identification based on 16S rRNA.Based on this isolation protocol, a collection of well-characterized environmental acinetobacters was available for further investigations.
We can conclude that environmental habitats may function as reservoirs or incubators for acinetobacters regardless their contamination (TPH, PAH, and MTBE) or matrix (groundwater, fishpond, and soil).As their presence is verified, the chance that environmental Acinetobacter strains can reach the infectious dose and may pose a public health risk cannot be excluded.
The antibiotic assay served pieces of evidence for the presence of antibiotic resistance and in some cases, multidrug resistance among the representatives of the environmental Acinetobacter strain collection.However, the antibiotic susceptibility patterns of environmental acinetobacters remarkably differ from clinical counterparts.Tigecycline (glycylcycline) proved to be the most effective antibiotic agent against environmental Acinetobacter proliferation (100% efficiency).
Based on our investigations, the detection rates and the antibiotic resistance profiles of the examined Acinetobacter spp.are not worrying but cannot be ignored and need further investigations.It is a warning sign that one MDR organism A. beijerinckii "T3N6" was found in a TPH-contaminated groundwater, which exhibited resistance to antibiotics of the "last resort" and its clinical counterparts can cause life-threatening infection (endocarditis).This result suggests the necessity of the detailed examination of acinetobacters that are used for industrial purposes.As a precaution, it can be suggested that opportunistic or unidentified microorganisms with the chance to cause infections or spread antibiotic resistance should be replaced with safe (Risk Group 1) ones as the classification of 2000/54/EC Directive suggests [50].
Based on our results, environmental acinetobacters other than the members of the clinically important ACB complex, namely A. beijerinckii, A. johnsonii, and A. gyllenbergii require detailed examinations regarding their public health and antibiotic resistance issues.
Our results may give a good basis for further examinations to reveal the significance of the human risk of environmental Acinetobacter strains.The improved multistep isolation and identification method may help to reach further results regarding the environmental role of Acinetobacter strains of various habitats.

Figure 1 .
Figure 1.The multistep method for selective isolation and identification of Acinetobacter strains

Table I .
The isolated and identified environmental Acinetobacter strains with the sampling data CARBAPENEM-RESISTANT A. BEIJERINCKII 117

Table III .
Antibiotic susceptibility patterns of the isolated Acinetobacter strains investigated with MIC Test Strips method