Molecular Characterization of Carbapenem-Resistant Acinetobacter baumannii Isolated from Intensive Care Unit Patients in Jordanian Hospitals

Acinetobacter baumannii is a common cause of healthcare-associated infections (HAI) worldwide, mostly occurring in intensive care units (ICUs). Extended-spectrum beta lactamases (ESBL)-positive A. baumannii strains have emerged as highly resistant to most currently used antimicrobial agents, including carbapenems. The most common mechanism for carbapenem resistance in this species is β-lactamase-mediated resistance. Carbapenem-hydrolyzing class D oxacillinases are widespread among multidrug-resistant (MDR) A. baumannii strains. The present study was conducted to determine the presence and distribution of blaOXA genes among multidrug-resistant A. baumannii isolated from ICU patients and genes encoding insertion sequence (IS-1) in these isolates. Additionally, the plasmid DNA profiles of these isolates were determined. A total of 120 clinical isolates of A. baumannii from various ICU clinical specimens of four main Jordanian hospitals were collected. Bacterial isolate identification was confirmed by biochemical testing and antibiotic sensitivity was then assessed. PCR amplification and automated sequencing were carried out to detect the presence of blaOXA-51, blaOXA-23, blaOXA-24, and blaOXA-58 genes, and ISAba1 insertion sequence. Out of the 120 A. baumannii isolates, 95% of the isolates were resistant to three or more classes of the antibiotics tested and were identified as MDR. The most frequent resistance of the isolates was against piperacillin (96.7%), cephalosporins (97.5%), and β-lactam/β-lactamase inhibitor combinations antibiotics (95.8%). There were 24 (20%) ESBL-producing isolates. A co-existence of blaOXA-51 gene and ISAba1 in all the 24 ESBL-producing isolates was determined. In addition, in the 24 ESBL-producing isolates, 21 (87.5%) carried blaOXA-51 and blaOXA-23 genes, 1 (4.2%) carried blaOXA-51 and blaOXA-24, but all were negative for the blaOXA-58 gene. Plasmid DNA profile A and profile B were the most common (29%) in ESBL-positive MDR A. baumannii isolates while plasmid DNA profile A was the most common in the ESBL-negative isolates. In conclusion, there was an increase in prevalence of MDR-A. baumannii in ICU wards in Jordanian hospitals, especially those having an ESBL phenotype. Thus, identification of ESBL genes is necessary for the surveillance of their transmission in hospitals.


Introduction
Acinetobacter baumannii is a Gram-negative coccobacillus that is an aerobic, nonmotile, catalase-positive, and oxidase-negative pathogen. A. baumannii is an opportunistic pathogen in humans, affecting mainly people with compromised immune systems, and is becoming increasingly significant as a nosocomial infection [1]. A. baumannii normally inhabits mucous membranes, skin, and soil [1]. The organism is not fastidious in its growth requirement

Identification of ESBL-Positive A. baumannii Isolates
Initial susceptibility screening of the A. baumannii isolates to both aztreonam (30 µg), and ceftazidime (30 µg) was evaluated by disk diffusion method, according to CLSI-2015 recommendation [17]. The double-disk diffusion method was used, and a 0.5 McFarland bacterial suspension spread over Mueller-Hinton agar (Oxoid, Hampshire, UK). A disc of ceftazidime (30 µg) and a disc of ceftazidime in combination with clavulanic acid (30/10 µg) were placed at a distance of 20 mm, incubated overnight at 37 • C and results were compared by the disk diffusion method. An increase in a zone diameter in the presence of clavulanate significantly (≥5 mm) compared to the inhibition zone around ceftazidime disc; the result is interpreted as confirmatory of ESBL production. E. coli ATCC 25922 and K. pneumoniae ATCC 700603 were used as negative and positive quality controls, respectively, according to CLSI-2015 recommendation [17].  , bla OXA-23 , bla OXA-24 , bla  Genes and the ISAba1 Insertion Sequence Genomic DNA was extracted from 24 ESBL-positive isolated colonies using Wizard ® Genomic DNA Purification Kit (Promega, Madison, WI, USA). Plasmid DNA was extracted using PureYield™Plasmid Miniprep System Kit (Promega, Madison, WI, USA). DNA quantification and purity was performed using NanoDrop (ND-1000 V3.7.1) micro volume UV-Vis spectrophotometer (Thermo Scientific, Wilmington, DE, USA). The 24 ESBL-positive isolates were subjected to PCR to detect the presence of bla OXA-51 , bla OXA-23 , bla OXA-24 , bla OXA-58 genes as described below. All primers used and PCR amplification size are listed in (Table 1) [10,11]. Each reaction mixture contained: 50-100 ng of genomic DNA, 12.5 µL PCR Master Mix, which contains Taq DNA polymerase, Qiagen PCR Buffer, and dNTPs (Qiagen, Valencia, CA, USA), and 1 µL of 20 pM of each primer, in a final volume of 25 µL. The 24 ESBL-positive isolates were also assayed for the presence of the ISAba1 insertion sequence by PCR using the primers ISAba1F and ISAba1R [18]. Each reaction mixture contained: 50-100 ng of genomic DNA; 12.5 µL PCR Master Mix, which contains Taq DNA polymerase; Qiagen PCR Buffer; and dNTPs (Qiagen, Valencia, CA, USA), and 1 µL of 20 pM of each primer, in a final volume of 25 µL. The PCR conditions consisted of 95 • C for 5 min, followed by 35 cycles at 95 • C for 45 s, 56 • C for 45 s, and 72 • C for 3 min, and a final extension step at 72 • C for 5 min [18].

Detection of bla
DNA sequencing was conducted at Princess Haya Biotechnology Center (PHBC/Jordan) for the purified PCR products of selected samples that were positive for the genes bla OXA-23 , bla OXA-24 genes, and the insertion sequence ISAba1 using the primers in (Table 1). Sequencing was conducted to confirm the presence of these genes since reference QC strains were not available for all genes tested. The nested PCR is intended to reduce non-specific binding in products due to the amplification of unexpected primer binding sites. It involves two sets of primers, used in two successive PCR runs; the second set purpose was to amplify a secondary target within the first run product. The primers used to amplify a fragment in bla OXA-58 gene ( Table 2) gave non-specific products. Therefore, new primers were designed to detect the bla OXA-58 gene by the same conditions of the nested PCR [17] (Table 2). Each PCR reaction mixture contained: 50-100 ng of genomic DNA, 12.5 µL Taq PCR Master Mix (Qiagen, USA), 1 µL of 20 pM of each primer, in a final volume of 25 µL. The PCR products were diluted 1:20 and 1:10 and were used as DNA template in the nested PCR. The newly designed primers used in the nested PCR are shown in (Table 2 note). PCR condition consisted of 94 • C for 5 min, followed by 32 cycles at 94 • C for 40 s, 55 • C for 25 s, and 72 • C for 1 min, and a final extension step at 72 • C for 5 min. The reaction mixture contained: 50-100 ng of the diluted PCR reaction, 12.5 µL Taq PCR Master Mix (Qiagen, Valencia, CA, USA), 1 µL of 20 pM of each nested primer in a final volume of 25 µL. The second PCR run used the same amplification conditions but with the outer primers as in (Table 1 note).

Plasmid Profiling of MDR-A. baumannii Isolates
Plasmid DNA was extracted from 44 MDR-A. baumannii clinical isolates including 24 ESBL-positive isolates and 20 ESBL-negative randomly picked isolates. Extraction was performed by the centrifugation method using PureYield™ plasmid MiniPrep System (Promega, USA) and the extracted plasmid DNA was stored at −20 • C. Electrophoresis was performed on a gel containing 0.8% agarose in Tris-Boret-EDTA (TBE) buffer. Five microliters of the sample and 1µL of running dye were loaded into wells in the gel and were run for 3 h at 80 V. Gels were stained with ethidium bromide and were examined under UV illumination to visualize plasmid bands of particular size. Molecular weight of bands was estimated by comparison with molecular size markers lambda DNA hind III Digest (BioLabs, New England, UK).

Statistical Analysis
Statistical analysis was performed using IBM SPSS Version 21 (Armonk, NY, USA) as previously described [19,20]. A p value of ≤0.05 was considered statistically significant.

Antibiotic Susceptibility Testing
All 120 A. baumannii isolates were tested against a panel of 17 antibiotic discs as recommended by the CLSI-2015 [17]. Most of the A. baumannii isolates were multi-resistant to the majority of antibiotics tested ( Figure 1). The highest resistance rate was observed against cefepime with 99.2% of the isolates resistant, followed by 98.3% resistance against each cefoperazone and ceftriaxone, and 96.7% for piperacillin. The resistance rate against the other antibiotics was 95.8% against each of ciprofloxacin, meropenem, and piperacillin/tazobactam; 94.2% each against aztreonam and sulfamethoxazole /trimethoprim; 93.35% against ceftazidime; 92.5% resistance against each imipenem and against tetracycline; 86.7% against gentamicin; 84.2% against levofloxacin; 83.3% against amikacin; 84.2% against tobramycin; and 0% against colistin.

Antibiotic Susceptibility Testing
All 120 A. baumannii isolates were tested against a panel of 17 antibiotic discs as recommended by the CLSI-2015 [17]. Most of the A. baumannii isolates were multi-resistant to the majority of antibiotics tested (Figure 1). The highest resistance rate was observed against cefepime with 99.2% of the isolates resistant, followed by 98.3% resistance against each cefoperazone and ceftriaxone, and 96.7% for piperacillin. The resistance rate against the other antibiotics was 95.8% against each of ciprofloxacin, meropenem, and piperacillin/tazobactam; 94.2% each against aztreonam and sulfamethoxazole /trimethoprim; 93.35% against ceftazidime; 92.5% resistance against each imipenem and against tetracycline; 86.7% against gentamicin; 84.2% against levofloxacin; 83.3% against amikacin; 84.2% against tobramycin; and 0% against colistin.

Detection of ESBL-Positive A. baumannii Isolates
Out of the 120 tested isolates, 24 (20%) were ESBL-positive. The positive isolates were tested further by an ESBL confirmatory test. There was a statistically significant association between the ESBL production in these isolates and the resistance to 10 antibiotics (amikacin, aztreonam, ceftazidime, ciprofloxacin, gentamycin, levofloxacin, meropenem, piperacillin/tazobactam, tobramycin, and sulfamethoxazole/trimethoprim; Table 2). This association may be due to the fact that genes coding for ESBLs are carried on plasmids and these plasmids also carry resistant genes for other antibiotics. Out of the 20% (24) ESBL-positive isolates, 9.2% (11) were from sputum, 4.2% (5) from urine, 2.5% (3) from

Detection of ESBL-Positive A. baumannii Isolates
Out of the 120 tested isolates, 24 (20%) were ESBL-positive. The positive isolates were tested further by an ESBL confirmatory test. There was a statistically significant association between the ESBL production in these isolates and the resistance to 10 antibiotics (amikacin, aztreonam, ceftazidime, ciprofloxacin, gentamycin, levofloxacin, meropenem, piperacillin/tazobactam, tobramycin, and sulfamethoxazole/trimethoprim; Table 2). This association may be due to the fact that genes coding for ESBLs are carried on plasmids and these plasmids also carry resistant genes for other antibiotics. Out of the 20% (24) ESBL-positive isolates, 9.2% (11) were from sputum, 4.2% (5) from urine, 2.5% (3) from blood, and 1.7% (2) from bronchial wash, while only 0.8% (1) each was isolated from triple-lumen central line, nasal culture, and wound specimen. The CHDLs OXA-type genes bla (OXA-58, OXA-51, OXA-24, OXA-23) and the insertion sequence ISAba1 were detected among the 24 ESBL-positive isolates at the following rates: bla OXA-51 (100%), bla OXA-23 (87.5%), bla OXA-24 (4.2%), bla OXA-58 (0%), and ISAba1 (100%). Our study showed significant associations (p < 0.05) between the bla OXA-23 gene and resistance to Antibiotics 2022, 11, 835 7 of 14 the following antibiotics: cefoperazone, ceftriaxone, ciprofloxacin, imipenem, levofloxacin, meropenem, piperacillin/tazobactam, piperacillin, and sulfamethoxazole/trimethoprim (Table 2). Additionally, statistically significant associations were identified between the OXA genes as some genes were more likely to coexist with other genes on the same chromosome or plasmid such as bla OXA-23 and bla OXA-24 (p < 0.05). Furthermore, the ESBL phenotype was not statistically significant with either of the OXA genes. Direct DNA sequencing was conducted for the purified PCR ISAba1 insertion sequence. The results were compared to the gene sequences at Genome BLAST, NCBI and it was a 100% match. No DNA band was observed for any of the 24 isolates in the agarose gel electrophoresis and they were considered negative for the OXA-58 gene.

Plasmid DNA Profiling
Plasmid profiling was conducted for 24 ESBL-positive MDR-A. baumannii isolates, and 20 ESBL-negative MDR-A. baumannii isolates to identify the number and the size of plasmids present in each. Plasmids were found in all the 24 ESBL (100%)-positive isolates that were MDR-A. baumannii. Additionally, plasmids were found in 18 (90%) ESBL-negative MDR-A. baumannii isolates, and no plasmid was found in 2 isolates (10%). The number of plasmids ranged from 1 to 10 plasmids among the ESBL-positive isolates, and 0 to 8 among ESBL-negative isolates. The size of plasmids ranged from <2 kb to 23 kb (Table S2, Supplementary Data). The plasmid profiles found in MDR-A. baumannii ESBLpositive and ESBL-negative isolates are shown in Tables 3 and 4, respectively. The statistical analysis shows a significant correlation (p value ≤ 0.05) between plasmid number and resistance to antibiotics tested in this study (Table 5). Additionally, a significant correlation was found between the increase in plasmid number and the resistance to the following antibiotics: amikacin, cefoperazone, ceftazidime, ceftriaxone, imipenem, tetracycline, and sulfamethoxazole /trimethoprim, while no significant correlation was found between the existence of OXA genes and the number of plasmids ( Table 5). The results also revealed that and all ESBL-positive MDR-A. baumannii isolates contain plasmids and that these isolates contain higher plasmid number than the ESBL-negative MDR-A. baumannii isolates. These results suggest that the genes encoding for the ESBL enzyme are carried on plasmids.     Table 5. Cont.

Discussion
Evolving antibiotic resistance is a clinically challenging problem among ICU patients with limited effective agents especially in cases of MDR-A. baumannii. [21]. In this study, 120 isolates of A. baumannii were isolated from various clinical specimens from patients in the intensive care units (ICU) of four main Jordanian hospitals. Of these isolates, 69 (57.5%) were from male patients, and 51 (42.5%) were from female patients with age range (<1 month to 90 years). The most common infections caused by A. baumannii were the respiratory tract infections (RTIs, 47.5%), followed by urinary tract infections (UTIs, 14.2%), bacteremia (11.7%), wound and pus (13.3%), cerebrospinal fluid (5.8%), triple-lumen central line (3.3%), peritoneal fluid (1.7%), ear infection, tissue infection, and nasal swab each (0.8%). The distribution of these infections varied from one hospital ICU to another. This difference might be associated with the type of procedure performed on each patient and the variations in the effectiveness of infection control programs of each hospital. A study performed in Saudi Arabia reported the most frequent infections among ICU patients is the upper respiratory tract (30.48%), followed by lower respiratory tract (47.65%) and subcutaneous tissue (9.50%) [22]. On the other hand, a study conducted in Egypt on 53 A. baumannii isolates from three ICUs showed that RTIs isolates counted for (45%), followed by wound infections (42%), UTIs (11%), and bacteremia (2%) [23]. In the current study, the results of the AST demonstrated a high level of resistance against most of the antibiotics tested. About 95% of the isolates were resistant to three or more classes of the antibiotics and, therefore, were identified as MDR (please check results section). The antibiotic resistance of A. baumannii isolates in our study was comparable to other studies conducted in the Middle East and Iran [24,25].
However, our AST showed less resistance to some of the antibiotics reported in a Chinese study which revealed that all the extensive drug-resistant A. baumannii isolates were resistant to all cephalosporins, all aminoglycosides, imipenem, ciprofloxacin, ampicillin/sulbactam, piperacillin/tazobactam, sulfamethoxazole-trimethoprim, and tigecycline [26]. The differences in resistance patterns observed for A. baumannii isolated in our study from other studies could be attributed to differences in the testing methods used, the environmental factors in each hospital, and frequent use of the antimicrobial agents. The antibiotic susceptibility test results of the QC strains, P. aeruginosa ATCC 27853 and E. coli ATCC 25922, were within the antimicrobial ranges reported in the CLSI-2015 guidelines. One exception was the inhibition zone (about 14 mm) for P. aeruginosa ATCC 27853 against cefepime (FEP) (30 µg), which is less than what is reported in the CLSI guidelines (24-30 mm). The explanation for this difference could be that the strain P. aeruginosa ATCC 27853 started to develop more resistance against cefepime, which warrants further investigation.
In our study, 24 (20%) isolates were ESBL-producing as confirmed by the ESBL confirmatory test. Our results are in accordance with another study that reported the detection of ESBL production in 27.5% of their isolates, and resistance to cefepime and ceftazidime could be detected in most of the ESBLs in Acinetobacter isolates by the double disc approximation test [27]. However, our results were much lower than what was reported in other studies that reported the presence of ESBLs in Iraq (61.53%), Iran (59%) and India (77%) of the Acinetobacter isolates [28][29][30]. The difference between these studies might be due to the method used to detect the ESBL in the isolates, where the number of ESBL-producing isolates by sulbactam was found to be much higher (77%) than the number of ESBL-producing isolates by the CLSI method (4%) [30].
Our study detected the presence of the bla-OXA genes in the ICU of four hospitals, which indicated that all the 24 ESBL-positive isolates (100%) were simultaneously positive for both bla OXA-51 and ISAba1, whereas bla OXA-51 , bla OXA-23 , and ISAba1 were simultaneously positive in 21 (87.5%) of the tested isolates. Additionally, bla OXA-51 , bla OXA-24 , and ISAba1 were simultaneously positive in 1 (4.2%) isolate, and bla OXA-51 , bla OXA-23 , bla OXA-24 , and ISAba1 were simultaneously positive in 1 (4.2%) isolate. Our findings also indicated that bla OXA-23 -like genes are predominant in A. baumannii, where bla OXA-23 was detected in 21 (87.5%) of the ESBL-positive isolates; our results are in agreement with a Saudi study that showed 94% of isolates carried OXA-51-like gene, and 91% contained OXA-23-like genes [31]. In contrast the bla OXA-23 gene was not dominant in Poland, and only 3 out of 104 (2.9%) carbapenem-resistant A. baumannii isolates carried this gene [10]. OXA-24/OXA-40-like carbapenemase was only detected in 1 (4.2%) isolate. OXA-58-like carbapenemase, which includes OXA-58, OXA-96 and OXA-97, was not detected in any of the A. baumannii isolates. The same negative results for this gene were reported by another study conducted in Poland [10] and in the UK [11]. The spread of antibiotic-resistant bacteria can be attributed to factors such as horizontal gene transfer, which refers to the transfer of genes between organisms in a way other than traditional reproduction and can even occur between different bacterial species. This horizontal gene transfer involves primarily temperate bacteriophages and plasmids [32]. Therefore, plasmid profiling was performed on 44 MDR-A. baumannii isolates in the current study including the 24 ESBL-positive isolates and 20 ESBL negative isolates. Results showed that all 24 ESBL-positive isolates contain plasmids with 10 plasmid profiles from A to J. Profiles A and B were the most predominant profiles present, which were found in 29% of the ESBL-positive isolates, profiles D and H were found in 8% of the ESBL-positive isolates, while profiles C, E, F, G, I, and J were found in a lower percentage (4%). Our results also showed that 18 (90%) of the 20 ESBL-negative MDR-A. baumannii isolates contain plasmids with profiles from A to J. Profile A was the dominant profile present in 40% of the ESBL-negative isolates; profile C was found in 10% of these isolates; and profiles B, D, E, F, G, H, I, and J were found in lower percentage (5%). The plasmid size varied in ESBL-positive isolates between ≥23,130 bp and ≥2322 to <4361 bp and were found in 35% and 20%, respectively, of the ESBL-positive isolates, while the dominant plasmid sizes in ESBL-negative isolates were longer and ranged between ≥9416 to <23,130 bp and were found in 40% of ESBL-negative isolates. The number of plasmids ranged 1 to 10 in ESBL-positive isolates and from 0 to 8 in ESBL-negative isolates. Consequently, based on the plasmid number, the MDR-A. baumannii isolates were divided into three groups: Group A (0-2 plasmids) was represented in 58% of the ESBL-positive isolates, and were represented in 70% of ESBL-negative isolates. Group B (3 to 4 plasmids) were represented in 12.5% ESBL-positive isolates and represented in 20% of ESBL-negative isolates. Group C (more than four plasmids) were represented in 29% of ESBL-positive isolates and represented in only 10% of the ESBL-negative isolates. All of the 44 isolates subjected to plasmid profiling were MDR-A. baumannii, and the results showed that 2 (10%) of the ESBL-negative isolates contained no plasmids, and 29% of the ESBL-positive isolates and 45% of the ESBL-negative isolates contained only one plasmid. These results may mean that antibiotic-resistant genes, but not ESBL genes, could be carried on the chromosome only, and that one plasmid may carry a number of genes coding for MDR. The results also revealed that all ESBL-positive MDR-A. baumannii isolates contain plasmids and that these isolates contain a higher plasmid number than the ESBL-negative MDR-A. baumannii isolates. These results also suggest that the genes encoding for the ESBL enzyme are carried on plasmids. In a study conducted in India, ten different plasmid profiles were detected among 55 A. baumannii clinical isolates, and most of the isolates had at least one plasmid. The number of plasmids ranged from 1 to 9, and the sizes of the plasmids ranged from 2 to >25 kb and showed that Acinetobacter may carry several copies of huge plasmids of 10 kb.
These results could be due to the presence of many transferable integron cassettes in this bacterium, which are evidently responsible for resistance [33]. A study in Iran reported seven different plasmid profiles were found in 112 Acinetobacter isolates (95.5%); the sizes of the plasmids ranged between 1 and >21 kb, and no significant association between plasmid profiles and the source of infection was found. However, they found a significant association between plasmid profiles with antibiotic resistance profiles. The study suggested that plasmid profiling is a reliable method to predict antibiotic resistance on the molecular bases [34]. MDR in A. baumannii is a result of acquired and innate resistance. The intrinsic resistance of A. baumannii is characterized by housekeeping genes encoded in the chromosomes and expressed in all strains. These genes include efflux pumps and -lactamases, such as the bla AmpC chromosomal gene, which encodes AmpC cephalosporinase, which is expressed in all strains of Acinetobacter baumannii. Horizontal gene transfer, a natural genetic transformation mechanism that involves the uptake of a short piece of bare DNA, allows Acinetobacter to acquire genes that encode for resistance elements [35].

Conclusions
Our results highlighted the prevalence of MDR among A. baumannii isolates from ICUs in four Jordanian hospitals. A significant association between the bla OXA-23 gene and resistance phenotypes was identified. Out of the 120 isolates, 24 isolates were identified as ESBL-producers. The simultaneous co-existence of insertion sequence with the naturally intrinsic bla OXA-51 gene could enhance the resistance rates to antibiotic agents. The finding that all ESBL-positive MDR-A. baumannii isolates contain plasmids, with higher plasmid numbers in the ESBL-positive isolates than ESBL-negative isolates, suggests that genes encoding for the ESBL enzyme are carried on plasmids. In addition, a significant association was found between the numbers of plasmids in each isolate with antibiotic resistance profiles. Our data may have an important impact on infection control policies in hospitals and on the availability of alternative treatment of such infections.