Multidrug-Resistant Acinetobacter: Detection of ESBL, MBL, blaNDM-1 Genotype, and Biofilm Formation at a Tertiary Care Hospital in Eastern Nepal

Background The Acinetobacter species is an important hospital-acquired pathogen. The rapid development of resistance to multiple drugs and the ability to form biofilm make these bacteria more adaptable to survive in healthcare facilities, thus posing a challenge to their effective management. Objective This study aimed to characterize clinical isolates of Acinetobacter spp and to study their antimicrobial susceptibility patterns and ability to form biofilm. Resistant Acinetobacter was further analyzed for the detection of extended-spectrum β-lactamases (ESBLs), metallo β-lactamases (MBLs), carbapenemase production, and presence of blaNDM-1 gene. Materials and Methods A total of 324 Acinetobacter species were isolated from various clinical specimens which were submitted to the Department of Microbiology, B.P. Koirala Institute of Health Sciences, Dharan, Nepal, and were studied for antibiotic susceptibility testing, detection of ESBL and MBL production, and formerly biofilm formation was performed by standard microbiological methods. PCR was carried out to determine the presence of the blaNDM-1 gene. Results The predominant Acinetobacter species isolated was A calcoaceticus-baumannii Complex (Acb complex) 167 (51.5%). Among those, all A. species 128 (40%) were multidrug resistant (MDR). In which 13 (4.0%) were ESBL producers, 70 (61.9%) were MBL, and 12 (10.6%) were carbapenemases producers. The blaNDM1 gene was present in 33 isolates. Thirty-seven percent (121/324) of isolates formed biofilm. The majority of A. species were resistant to cefotaxime 73.8% (239) and cefepime 74.4% (241). A significant proportion of biofilm producers were MDR (p < 0.001). Conclusion Drug-resistant Acinetobacter formed a substantial proportion of this hospital's samples with a large presence of the blaNDM-1 gene. A matter of great concern is the association of multidrug-resistant phenotype with biofilm formation. This situation warranted stringent surveillance and adherence to infection prevention and control practices.


Introduction
Acinetobacter, a widely distributed, saprophytic bacterium in nature, has established itself as one of the most common nosocomial pathogen [1,2]. Although diferent species of Acinetobacter are the potential to cause infection, 80% of infections are caused by Acinetobacter baumannii. Ease of survival even in adverse environments, ability to form bioflms on surfaces, and possession of many genes for antimicrobial resistance have made this bacterium an important pathogen. Te potential ability of the bacterium to form bioflms in certain instances, indeed, provides a potential explanation for outstanding antibiotic resistance and survival properties in the harsh environment of hospitals, particularly in the intensive care setting [3][4][5].
Over the past few decades, its clinical importance had increased due to its ability to receive antimicrobial resistance factors [6,7] through the transfer of plasmid or transposons that contained antimicrobial resistant genes, particularly in a hospital setting where usage of antibiotics are huge leading to selective pressure [8,9]. Multidrug resistant (MDR) Acinetobacter species are defned as isolates resistant to the major three classes of antimicrobial agents -all penicillins and cephalosporins (including inhibitor combinations), fuoroquinolones, and aminoglycosides [7][8][9][10][11]. Tese strains are implicated in a variety of life-threatening infections such as ventilator-associated pneumonia (VAP), urinary tract infections, bloodstream infections, surgical site infections, and infections associated with medical devices, occurring especially in patients of intensive care units. Moreover, a signifcant correlation between bioflm formation and multidrug resistance has been attributed to the threat imposed by Acinetobacter to the current antibiotic era [8,9,12].
Diagnosis of multidrug-resistant Acinetobacter infection is a great challenge owing to the distribution of various species in relation to the type of infection, their antimicrobial profle, and bioflm-forming phenotype. Hence, from efective management and infection control perspectives, it is crucial to minimize the risk associated with Acinetobacter infection in a healthcare setting. Tis study was conducted to characterize the clinical Acinetobacter isolates with special reference to the detection of antimicrobial resistance, bioflm formation, and the presence of the bla NDM-1 gene.

Materials and Methods
A total of 324 Acinetobacter species isolates were obtained from various clinical specimens, and submitted to the Department of Microbiology, B. P. Koirala Institute of Health Sciences (BPKIHS) Dharan, Nepal. Tis study was conducted from September 2017 to August 2018. Ethical approval was obtained from the Institutional Review Committee of BPKIHS, Dharan, Nepal.

Identifcation of Acinetobacter Species.
Direct microscopic examination of Gram-stained smear of all samples except blood were performed. Inoculation of samples onto appropriate culture media, incubation, and detection of growth after the recommended duration was carried out by standard microbiological techniques [13]. On blood agar suspected smooth, opaque colonies corresponding to nonlactose fermenting colonies on MacConkey and on CLED agar plates were presumed as Acinetobacter and processed further. Species identifcation of the genus Acinetobacter was carried out by several biochemical tests which included triple sugar iron (TSI) fermentation test, oxidase, indole, motility, urease, and arginine hydrolysis [14,15].

Detection of ESBL Phenotype.
According to the CLSI guidelines, probable ESBL-producing isolate had a zone of inhibition for ceftazidime (30 μg) ≤ 22 mm and cefotaxime (30 μg) ≤ 27 mm [13]. In order to confrm ESBL production, ceftazidime (30 μg) and ceftazidime + clavulanate (30/10 μg) discs were placed in Acinetobacter culture. Zones of inhibition were compared with the ceftazidime and cefotaxime discs alone and compared with the combined ceftazidime + clavulanate disc. An enhanced zone of the diameter of ≥5 mm in combination with clavulanate was confrmed isolates as ESBL [13].

Combined Disc Difusion Test.
A combined disc diffusion test was employed to determine the MBL-producing phenotype in Acinetobacter. On the MHA plate lawn culture of Acinetobacter, imipenem disc (10 μg) and imipenem disc with 10 μl of 0.5 M EDTA were applied 20 mm apart from center to center. Te zone of inhibition of >7 mm around the imipenem-EDTA disc compared to the imipenem disc alone classifed the isolate as an MBL producer [16].

Carbapenemase Production Test.
Phenotypic detection of carbapenemase-producing MDR Acinetobacter was determined by a modifed Hodge test [13]. First of all, an overnight broth culture of Escherichia coli ATCC 25922 was adjusted to 0.5 McFarland standards and spread on the dried surface of Mueller Hinton agar (MHA) plate by sterile cotton swab. After transitory drying, a 10 μg imipenem (IMP) disc was placed at the center of the plate, and tested strains were streaked from the center to the periphery of the plate in four diferent directions. Following overnight incubation at 37°C, carbapenemase-positive isolates showed the distorted zone of inhibition, and a "clove leaf pattern" was observed due to carbapenemase production by isolates [13].  Gene. New Delhi metallo beta lactamase-1 (bla NDM-1 ) is a novel MBL that confers resistance to all β-lactam antibiotics with the exception of aztreonam. In this study, the multidrug resistant organisms were determined to have the bla NDM -1 gene by conventional PCR [17].

Molecular Detection of bla
2.6. DNA Extraction of Bacterial Isolates. Te overnight broth cultures were centrifuged at 3,500 rpm for 10 minutes at 4°C. Ten, the supernatant was discarded and the pellet was washed twice with 5 ml phosphate bufered saline (PBS) followed by centrifugation. Te pellet was resuspended in 1 ml PBS and centrifuged at 10,000 RPM for 10 minutes at 4°C. Finally, the supernatant was discarded and the remaining pellet was stored at −20°C till assayed [18].

International Journal of Microbiology
Te pellet was dissolved in 200 μL of TE bufer. Ten, the mixture was heated at 100°C for 10 minutes with shaky and rapidly transferred to an ice bath for 10 minutes. Centrifugation was performed at 13,000 rpm for 30 seconds at 4°C. Finally, 100 μL of supernatant DNA was transferred to a new tube. Te concentration and purity of the DNA were measured by Nanodrop 2000 spectrometer (Termofsher) [17].

Polymerase Chain Reaction (PCR) for Detection of bla NDM-1 .
Te bla NDM-1 gene-specifc PCR was performed to detect bla NDM-1 . PCR master mix was prepared in 25 μL fnal volume which constituted 1X Qiagen PCR bufer, 2 mM MgCl 2 , 0.1 mg/ml BSA, 0.2 mM of dNTP mix, 0.8 μM of each primer NDM1-F (5′-CAGCACACTTCCTATCTC-3′) and NDM1-R (5′-CCGCAACCATCCCCTCTT-3′), 0.5 Unit of Taq polymerase, and 2 μL of DNA template. PCR amplifcation was carried out in Eppendorf Mastercycle ProS (Eppendorf, Germany) with (i) initial denaturation at 95°C for 5 minutes, (ii) 35 c ycles of denaturation at 94°C for 30 seconds, 55°C for 30 seconds, then at 72°C for 30 seconds, and (iii) a fnal extension at 72°C for 10 minutes. PCR water was used as a negative control, and the DNA from a bacterial culture with bla NDM-1 PCR positive result was considered as a positive control. After electrophoresis of amplifed DNA in 2% agarose gel at 5 V/cm and ethidium bromide staining, the DNA band was visualized with UV exposure [19]. Te sample was identifed as bla NDM-1 PCR positive result if DNA band of 300 bp, as seen in the gel.

Detection of Bioflm Formation.
Detection of bioflm formation was performed by the standard laboratory methods described elsewhere [20][21][22][23]. Te bacterial isolates were grown overnight at 37°C in 5 ml of tryptic soy broth (TSB). Methicillin-sensitiveStaphylococcus aureus (MSSA) ATCC-25923 and P. aeruginosa ATCC-27853 were used as negative and positive controls, respectively. Each well of a 96-wellfat-bottomed plates were flled with 200 μL with overnight culture broth (0.5 McFarland standard diluted with 1% glucose + TSB). Te plates were covered with lids and incubated aerobically for 24 hours at 37°C. After incubation, the content of each well was removed and washed three times with 300 μL sterile phosphate-bufered saline (PBS; pH 7.2) in order to remove freely foating bacteria.
Succeeding every washing, the plates were drained in an inverted position. Adherent bacteria was fxed with 150 μL of methanol for 20 min, those 96-well plate was emptied by simple ficking, and was left to air dry overnight at room temperature. Te adherent bioflm layer formed in each 96well plate's well were stained with 150 μL of 2% Hucker crystal violet for 15 minutes. Excess stains were rinsed of by placing those 96-well plate under the running tap water until the release of stain got stopped. Te 96-well plate were airdried at room temperature. Ten, 150 μL of 95% ethanol was gently added in each well of the microtiter-plate in order to permit cell re-suspension. Again, the plate was incubated at room temperature for 30 minutes without shaking. Te absorbance (A) of each solution, well stained with crystal violet, was measured at 570 nm using a microtiter-plate reader [23].

Statistical
Analysis. Data were entered in MS Excel 2013 worksheet and statistical analysis were carried out by using R package version 0.55 [24]. Te principle component analysis among the several factors such as MDR, MBL, blaNDM , and bioflms were carried out by using the "prcomp" function of the R stat package, correlation, and visualization of the plot were demonstrated by the ggbiplot package [25].
Amongst those diferent specimens analyzed, Acb complex was the predominant species (Table 1).
In this study, 26% of the samples were obtained from the medical ward, 20% from ICU, 12% from OPD,11% from surgery and pediatrics, 6% from gynecology, 4% of emergency, NICU, and orthopedic department. Acb complex was predominant in ICU (76.7%).

Antimicrobial Susceptibility Testing.
Te resistance percentages of Acinetobacter in the descending order of frequency were cefepime 74.4% cefotaxime 73.8%, ceftriaxone 65.7%, ceftazidime 72.5%, ceftazidime + clavulanic acid 72.2%, piperacillin 65.7% ampicillin + sulbactam 36.7%, amikacin 44.7%, ciprofoxacin 50%, and imipenem 35.2%. Acb complex was found to have the highest drug-resistant phenotypes to analyze antibiotics with 57.4% being resistant to imipenem. For the Acb complex, cefotaxime was the antibiotic with the highest resistance frequency (94%), as for A. hemolyticus, it was 31 isolates out of 38 (82%). More than 50% of A. lwofi and A. junii isolates were sensible to the investigated antimicrobials (Table 2). Acinetobacter isolates from ICU were more resistant to the antibiotics than those from other wards.
Among 324 isolates, 128 (39.5%) were MDR. Most of MDR were from patients in ICU 60.3% followed by OPD 43.5%, Ward 32.1%, and Emergency 20.0%. Acb complex had the highest rate of MDR phenotype as shown in Table 3.

Bioflm Formation.
Among 324 isolates, bioflm production was detected in 121 (37.3%) isolates. Forty-fve (37.2%) bioflm-forming isolates were obtained from the device, 36 (29.8%) from pus, 20 (16.5%) from blood, 18 (14.9%) from urine, and 7 (5.8%) from sputum. Bioflm production was found in all species as depicted in Table 3. About 60.3% of MDR and 64.4% bioflm-forming isolates were from ICU. Moreover, MDR phenotype and bioflm formation phenotypes were signifcantly associated (p value < 0.0001) whereas no association was determined among other virulence phenotypes such as ESBL, MBL, carbapenemase, and bla NDM-1 . Te principal component analysis (PCA) among Acinetobacter isolates, origin, bioflm formation, MDR, and number of antibiotic resistance in the diagram showed the fndings that the circle is more closer to Y -axis which represents bioflm whereas the circles represents the origin of isolates and Acinetobacter species in Figure 2 and Figure 3, respectively. Similarly, bioflm formations were consistently found in isolates from ICU ( Figure 2) and Acb complex (Figure 3) as shown by principal component analysis.

Discussion
Acinetobacter is one of the notorious nosocomial pathogen and its tendency to develop resistance against antimicrobial drugs is an important rationale for infection control at Health care facility. Among fve Acinetobacter species, Acb (Acinetobacter calcoaceticus-A. baumannii) complex was one of the most predominating species (51.5%) in this study, which was comparable to the fndings of other studies [15]. It suggests Acb complex has more survival rate even in an unfavorable environment and causes hospital acquired infection. About 20% of isolates were obtained from ICU which is similar to fndings reported in the previous study from Nepal [26]. Tis indicates that ICU could be the most important location for the colonization and survival of Acinetobacter in at hospital environment [5,27]. ICU patients usually require a prolonged hospital stay, need repeated invasive procedures and utilizes various devices for life support, and frequently receives treatment with broad-spectrum antimicrobials. Most of the sample isolates were of the cases of sepsis from the ICU. Previous antimicrobial therapy, medical devices, and prolonged hospitalization are the known risk factors for bacteremia in such patients [28].
Resistance to cefepime (74.4%) and cefotaxime (73.8%) were detected in 74.4% and 73.8% of isolates respectively, followed by ceftazidime (72.5%), ceftriaxone (65%), and piperacillin 65%. It was found that the isolates resistance to amikacin was 44.7% and to ciprofoxacin 50.0% which were consistent with other reports [28,29]. Tis indicates that Acinetobacter species have intrinsic and/or easily acquired mechanisms of resistance against many of the available antimicrobial agents making this pathogen one of the most signifcant microbial challenges for the current period.
Although carbapenem was the frst-line drug against Acinetobacter infection in the late 1990s, carbapenemresistant strains are increasingly reported worldwide [10]. Among the ICU isolates, 42.5% were sensitive to ampicillin/ In this study, 128 (39.5%) isolates were determined as multidrug resistant (MDR), in which it was found that all species were MDR strains. Acinetobacter appeared to have the propensity to develop antibiotic resistance rapidly, as a consequence of prolonged antibiotic exposure. Hence, the increasing trend of Acinetobacter MDR strains were reported globally [30,31].
In this study, 235 (72.5%) of the strains were ceftazidime resistant, and 13 (4.0%) of them demonstrated ESBL production by double disc synergy test which disagree with other reports [30,32]. Since the antimicrobial susceptibility pattern could be variable depending on several factors, the surveillance studies have a crucial role in deciding the therapy against Acinetobacter infection [15]. In this study Among MDR isolates, 10.7% had demonstrated carbapenemase production by the MHT method. Te MHT and combined disc difusion tests with Imipenem and EDTA have been extensively used as a phenotypic assay for the detection of carbapenemase and MBL enzyme as it is a simple test to perform. MHT had been found to be sensitive in the detection Ambler class A (KPC) and class D (OXA-48) whereas its sensitivity is very low for NDM-1 producer which was demonstrated in this study as well [15,33]. Whereas, there is high sensitivity but low specifcity rate of combined disc test for detection of MBL production by a phenotypic method which result may increase the false positive rate of MBL [34].
Te gold-standard for the identifcation of carbapenemase production and detection of the bla NDM-1 gene is PCR assay [35]. Moreover, other assay such as loop-mediated isothermal amplifcation [36] had been developed for the detection of the bla NDM-1 gene. In this study, NDMproducing Acinetobacter isolates was 10.2% (33/324) which was consistent with the study, carried out in western Nepal [37]. Te evidence of Acinetobacter with the presence of the bla NDM-1 gene had already been reported worldwide [38,39]. In this study, 91% of NDM producers were resistant to second and third-generation cephalosporins. Moreover, 25.8% (33/128) of bla NDM-1 gene-containing isolates had MDR phenotype. However, Imipenem was the most efective antibiotic in the study with only 35.2% imipenem resistance.
Te bioflm-forming phenotype of Acinetobacter was determined by 96 well plate assay and were found 37.3% of isolates were bioflm formers of which was inconsistent with the previously reported 73.7% [40]. Moreover, 65.3% of bioflm-forming isolates had signifcantly associated with multidrug resistant phenotype. Isolates with the bioflm producer were Acb complex (56.2%) followed by A. lwofi (14.5%), A. haemolyticus (21.1%), A. radioresistens (20%) and A. Junii (16.7%). Bioflm-forming isolates were found mostly on samples from medical devices (37.2%) followed by pus (29.8%) which was reliable with most of the studies, that reported more than 90% bioflm-producing isolates from medical devices as well as showed multidrug resistance pattern [41,42]. In addition, the bioflm-forming isolates were predominantly in ICU (64.4%) followed by wards (31.1%) which was also consistent with other studies and it suggested that medical devices help to colonize prior to the development of bioflm formation [43]. Among the bioflm formers in this study, resistance to imipenem at 62% (75/ 121) was higher than the 46.7% reported by Yadav et al. [5]. Tis study reported that the clinical Acinetobacter isolates from ICU had both phenotype bioflm producers and multidrug resistance [3,4,42]. Moreover, the association between bioflm producers and antibiotics had been  demonstrated by multiple studies. Te rationale is that, once devices are colonized, the bioflms that develop share fundamental characteristics of all bacterial bioflms, and the cells within the bioflm were protected by the extracellular matrix, that protective material could decrease the efectiveness of both antibiotics and host defense mechanisms [44,45]. However, it was difcult to identify the antibiotic resistant pattern of bioflm-producer microorganism by manual AST, because planktonic bacterial cells were used for susceptibility tests. In addition, antibiotics resistance might be even higher than observed in the present study, as bacteria without a molecular basis of resistance (susceptible in vitro) could be resistant in a bioflm environment.
Although there were innovative antibioflm therapeutics which included combining APDT with antibiotics, plant extracts, or bioflm-disrupting enzymes that could assist in managing such cases. It helped to increase the sensitivity of the microorganism to antibiotic therapies by violating the structure of the bioflm or disturbing the communication between populations of microorganisms in the bioflm [46].
However, diferent results were seen in diferent studies due to variations as of geography, arrangement of specimens in the study groups, condition of the patient, and use of antibiotic.
Te data from this study demonstrated that Acinetobacter species were resistant to many of the available antimicrobial agents, making those nosocomial pathogens as one of the most signifcant microbial challenges to have the control in future.

Conclusion
Te clinical isolates of Acinetobacter in this setting were multidrug-resistant MBL producers with blaNDM-1 gene and bioflm formers. In addition, there is evidence that the bioflm formation is a potential marker to determine the multidrug resistant (MDR) phenotype. Tese isolates have been proven to cause nosocomial infection in healthcare settings and are challenging to treat. Terefore, a consolidated efort by all healthcare providers by strict implementation of infection prevention and control activities, early diagnosis, and antibiotic stewardship are recommended to reduce the burden of antimicrobial resistance on patients and health facilities.

Data Availability
Tis is a hospital-based study. Samples were collected during the routine diagnosis and genotyping procedure. Terefore, the data for the analysis will available upon the request from the corresponding author or head of the department of Microbiology (hod.microbiology@bpkihs.edu), BP Koirala Institute of Health Sciences, Dharan, Nepal.

Disclosure
Te preprint version of this paper was submitted to Research Square [47]. Te authors also gratefully thank the Research Square for the peel review process of our research paper.

Conflicts of Interest
Te authors declare that they have no conficts of interest regarding the publication of this paper.