Impact of target site mutations and plasmid associated resistance genes acquisition on resistance of Acinetobacter baumannii to fluoroquinolones

Among bacterial species implicated in hospital-acquired infections are the emerging Pan-Drug Resistant (PDR) and Extensively Drug-Resistant (XDR) Acinetobacter (A.) baumannii strains as they are difficult to eradicate. From 1600 clinical specimens, only 100 A. baumannii isolates could be recovered. A high prevalence of ≥ 78% resistant isolates was recorded for the recovered isolates against a total of 19 tested antimicrobial agents. These isolates could be divided into 12 profiles according to the number of antimicrobial agents to which they were resistant. The isolates were assorted as XDR (68; 68%), Multi-Drug Resistant (MDR: 30; 30%), and PDR (2; 2%). Genotypically, the isolates showed three major clusters with similarities ranging from 10.5 to 97.8% as revealed by ERIC-PCR technique. As a resistance mechanism to fluoroquinolones (FQs), target site mutation analyses in gyrA and parC genes amplified from twelve selected A. baumannii isolates and subjected to sequencing showed 12 profiles. The selected isolates included two CIP-susceptible ones, these showed the wild-type profile of being have no mutations. For the ten selected CIP-resistant isolates, 9 of them (9/10; 90%) had 1 gyrA/1 parC mutations (Ser 81 → Leu mutation for gyrA gene and Ser 84 → Leu mutation for parC gene). The remaining CIP-resistant isolate (1/10; 10%) had 0 gyrA/1 parC mutation (Ser 84 → Leu mutation for parC gene). Detection of plasmid-associated resistance genes revealed that the 86 ciprofloxacin-resistant isolates carry qnrA (66.27%; 57/86), qnrS (70.93%; 61/86), aac (6')-Ib-cr (52.32%; 45/86), oqxA (73.25%; 63/86) and oqxB (39.53%; 34/86), while qepA and qnrB were undetected in these isolates. Different isolates were selected from profiles 1, 2, and 3 and qnrS, acc(6,)-ib-cr, oqxA, and oqxB genes harbored by these isolates were amplified and sequenced. The BLAST results revealed that the oqxA and oqxB sequences were not identified previously in A. baumannii but they were identified in Klebsiella aerogenes strain NCTC9793 and Klebsiella pneumoniae, respectively. On the other hand, the sequence of qnrS, and acc(6,)-ib-cr showed homology to those of A. baumannii. MDR, XDR, and PDR A. baumannii isolates are becoming prevalent in certain hospitals. Chromosomal mutations in the sequences of GyrA and ParC encoding genes and acquisition of PAFQR encoding genes (up to five genes per isolate) are demonstrated to be resistance mechanisms exhibited by fluoroquinolones resistant A. baumannii isolates. It is advisable to monitor the antimicrobial resistance profiles of pathogens causing nosocomial infections and properly apply and update antibiotic stewardship in hospitals and outpatients to control infectious diseases and prevent development of the microbial resistance to antimicrobial agents.


Materials and methods
Specimens collection. A total of 1600 specimens were collected from blood, respiratory tract, urinary tract, catheters, wounds, and skin between January 2014 and March 2019 from two university hospitals; Al Azhar university hospital and Assuit university hospital, Assuit governorate, upper Egypt. Different clinical isolates were recovered. The study was approved by the ethics committee (Ethics Committee ENREC-ASU-63) of faculty of pharmacy-Ain Shams university. Both informed and written consents were obtained from the patients after explaining the study purpose. Also, all methods were performed following the relevant guidelines and regulations. The isolates were identified by standard microbiological methods 16 and as described below.
Phenotypic and genotypic identification of A. baumannii isolates. All specimens were cultured on nutrient agar (Oxoid Limited, England) as a general medium to recover all bacterial pathogens. The recovered bacterial isolates were characterized using two selective media, MacConky agar (Oxoid Limited, England) to identify lactose fermenting from non-lactose fermenting species and Herellea agar (Himedia, India) for selective scoring of Acinetobacter spp. Incubation was done at 37 °C for 24 h. All collected clinical isolates were propagated and maintained by standard microbiologic techniques 16 . For phenotypic identification, single separate colonies were handled for qualitative conventional diagnostic tests for A. baumannii; including Gram staining. The typical isolates showed Gram-negative reaction, catalase, and citrate utilization positive, while oxidase and indole tests were negative 17 . For genotypic identification, genomic DNA was extracted using the GeneJET Genomic DNA Purification Kit (Thermo, USA, catalog No. K0721) and used for amplification of the intrinsic bla OXA-51 -like gene. This gene is unique for A. baumannii as previously described 18 . A. baumannii ATCC 19,606 was used as a positive control.

Molecular typing of CIP-resistant A. baumannii isolates. Investigation of clonal relationship and
diversity of the recovered A. baumannii isolates was determined by molecular typing of the CIP-resistant isolates using ERIC-PCR 21 . Genomic DNA was extracted using the Genomic DNA Purification Kit (Thermo Fisher Scientific) according to the manufacturer's instructions. ERIC-PCR was carried out using the ERIC-1 (5'-ATG TAA www.nature.com/scientificreports/ GCT CCT GGG GAT TCAC-3') and ERIC-2 (5'-AAG TAA GTG ACT GGG GTG AGCG-3') primers as previously described 21 . The PCR products were analyzed using agarose gel electrophoresis at 1.5% (w/v) agarose containing 0.5 mg/ml ethidium bromide that was subsequently visualized by UV transilluminator. ERIC-PCR dendrogram was constructed by the use of UPGMA clustering method, Bionumeric program version 7.6 (Applied Maths A. PCR screening of FQ-resistance target site mutation PCR condition denaturation at 95 °C for 2 min, then 35 cycles of amplification as follows: denaturation at 95 C for 30 s, annealing for 30 s at primer set-specific temperatures, and extension at 72 °C for 1 min), a final extension at 72 °C for 5 min. The oligonucleotides used to amplify and for sequencing analysis are shown in Table 1 23 .

B. Purification of PCR products
Purification of PCR products before sequencing was performed using the PCR purification kit (Thermo, USA, Catalog number: K0701). Briefly, 1 volume of DNA binding buffer was mixed with each volume of PCR product. The sample mixture was then transferred into thermo-spin and centrifuged. The column was washed with a DNA wash buffer. Finally, the column was eluted with 50 µl DNA elution buffer. The purified DNA was stored at -20 0 C for safe storage of amplicons till sequencing using Sanger sequencing technique (Applied Biosystems genetic analyzers (ThermoFisher, Uk)).

C. Target site sequence analysis
A 613 bp and A 919 bp fragments of the corresponding gyrA and parC genes were amplified from 12 selected isolates (one isolate from one out of 12 profiles, isolates of the profiles 1 to 12 showed resistance to 19,18,17,16,15,14,13,12,11,9,8, 6 out of 19 tested antimicrobial agents, respectively. The chosen 12 profiles included PDR XDR and MDR isolates). Nucleotide sequences were visualized by SnapeGene Viewer (version 5.1.4.1 software 2020). ORFs were identified using the ORF finder tool. The sequences of gyrA and parC genes from the selected 12 isolates were compared to their corresponding ones of the gyrA and parC genes of A. baumannii ATCC 19,606 as previously described 24 . Pairwise codon-based nucleotide alignments (CDS-alignments) of the gyrA and parC genes from the selected isolates against their corresponding sequences of the standard strain were carried out.

A. Plasmid DNA extraction
Plasmid DNA was extracted and purified using the GeneJET Plasmid Miniprep Kit (Thermo, USA, catalog No. K0502). Briefly, a single colony from each test isolates was picked up from a freshly streaked selective plate to inoculate 1-5 mL of Mueller Hinton broth mixed with ciprofloxacin (0.125 mg/ml) in a test tube. The tubes were incubated at 37 °C for 12-16 h under shaking at 200-250 rpm in a shaker incubator. The volume of the container either a test tube or a flask was at least 4 times the culture volume. Cell pellets of the bacterial culture were collected by centrifugation at 6800 × g (8000 rpm) for 2 min using a micro-centrifuge. The supernatant was decanted and bacterial pellets were resuspended and subjected to alkaline lysis to liberate the plasmid DNA. The resulting lysate was neutralized to create appropriate conditions for the binding plasmid DNA on the silica membrane in the spin column. By centrifugation, cell debris was pelleted, and the supernatant containing the plasmid DNA was loaded onto the spin column membrane. The adsorbed DNA was washed to remove contaminants and then eluted with 50 µL of the elution buffer (10 mM Tris-HCl, pH 8.5). The extracted plasmid DNA was stored at − 20 °C for subsequent use.

B. PCR screening of PAFQR determinant genes
Plasmid extracts of all isolates were screened for the following genes: qnrA, qnrB, qnrS, aac(6′)-Ib-cr, qepA, oqxA , and oqxB using gene-specific primers listed in Table 2. PCR conditions and primer sequences were as previously described 25 . Ethical approval. The study was approved by the Ethics Committee (Ethics committee ENREC-ASU-63) at Faculty of Pharmacy-Ain Shams University where both informed and written consent were obtained from the patient after explaining the study purpose.

Results
Phenotypic and genotypic identification of A. baumannii isolates. The isolates were identified according to standard microbiological methods 16 . The isolates showing characters suspected to be Acinetobacter species were subjected to some biochemical and growth conditions tests. The suspected isolates of Acinetobacter spp gave the following test results: negative reaction with oxidase, and indole, positive reaction with catalase test, growth at 44 °C, and positive for citrate utilization test. Examination of cell morphology of Gram-stained films revealed that: all Acinetobacter isolates were short, Gram-negative diplo-coccobacillus (sometimes de-staining of primary stain was difficult due to tendency of some isolates to retain crystal violet) with pairing or clustering arrangements. From a total of 623 non-lactose fermenters isolates, only 151 had these mentioned characters.
All Acinetobacter species candidates (151 isolates) were further subjected to conventional PCR to confirm their identity. The characteristic band at 353 bp for the bla OXA-51 -like gene of A. baumannii was used for this identification and confirmation. From the total 151 isolates subjected to PCR, only 100 isolates were positive for bla OXA-51 and confirmed to be A. baumannii as shown in Supplementary Fig. S1.
The specimens that showed the highest percentage 61% (61/100) of A. baumannii contamination were obtained from the respiratory tract (ETT 29%, nasal 17%, sputum 13%, throat 2%), followed by urinary tract infection 17% (urine 9%; urinary tract catheter 8%) and blood 12%, while the lowest (2%; 2/100) was from skin and CVC. Table 3 all A. baumannii isolates exhibited high resistance to most of the tested antimicrobial agents. However, the higher resistance was recorded to piperacillin (99%) and cephalosporins (98%). On the other hand, the most effective antimicrobial agents were recorded to be colistin (only 5% of isolates showed resistance) followed by doxycycline (only 57% of isolates showed resistance).
Analyses of the resulting A. baumannii susceptibility to the 19 tested antimicrobial agents showed the diversity of isolates' resistance to the tested agents. They were divided into 12 profiles according to the number of antimicrobial agents to which they were resistant. The resistance scope ranged from 6 to 19 antimicrobial agents. The first profile is PDR (two isolates), which represents isolates resistant to 19 tested antimicrobial agents. The Table 2. Oligonucleotide primers, their sequences, and annealing temperatures used for detection and sequencing of PAFQR genes, and the expected sizes of resulting amplicons.  www.nature.com/scientificreports/ and the results were subjected to phylogenetic analysis (Fig. 2). As a result, the isolates could be divided into three major clusters (I, II, and III) with similarities ranging from 10.5 to 97.8%. The high diversity indicates multiple contamination sources with A. baumannii. Cluster I represents 63.95% (55/86). Cluster II represents 2.32%, and included only two isolates namely; AS-44 and AS-46, while cluster III constitutes 37% (29/86) of the total isolated A. baumannii (Fig. 2). All A. baumannii in clusters II and III were recovered from Assiut University while most isolates included in cluster I was recovered from Al-Azhar University.

Molecular characterization of A. baumannii resistance mechanisms to FQs. Detection of target
site mutation. Out of 12 phenotypic profiles (Fig. 1), one isolate from each profile was selected, and its purified genome was subjected to PCR amplification of gyrA and parC genes followed by sequencing of the resulting amplicons. Out of the 12 isolates, 10 isolates were fluoroquinolone-resistant, and two isolates were fluoroquinolone susceptible (AS-01 and AS-05). gyrA and parC gene sequences of the selected A. baumannii isolates were analyzed and they showed similarities ranging from 95 to 100% to their corresponding sequences deposited in GenBank nucleotide database under accession numbers shown in Supplementary Table 1. Target site mutation analyses in gyrA and parC gene sequences of the selected twelve A. baumannii isolates represent 12 profiles. Two isolates (susceptible ones to CIP) had a wild-type profile. For the ten isolates (CIP-resistant ones), 9 of them (9/10; 90%) had 1 gyrA and 1 parC mutations Ser 81 → Leu mutation for gyrA gene and Ser 84 → Leu mutation for parC gene. The remaining CIP-resistant isolate (1/10; 10%) had (0 gyrA /1 parC) mutation (Ser 84 → Leu mutation) for parC gene. All tested isolates had a silent mutation in one or more positions of either gyrA or parC or both gyrA and parC (Supplementary Table 2, and Figs. 3 and 4).

Plasmid-associated fluoroquinolone resistance (PAFQR) genes.
A. DNA plasmid extraction The variable number of bands per isolate may be plasmids with different molecular weights that were detected (no endonuclease digestion was used) in 99% of A. baumannii. Besides, no plasmid could be detected in 1% of the isolates.

B. PCR screening A. baumannii isolates for PAFQR genes
All isolates were screened for PAFQR genes (qnrA, qnrB, qnrS, acc(6)-ib, qepA, oqxA, and oqxB) using conventional PCR. The expected sizes of PCR products mentioned in Table 2 (Table 4). The prevalence of PAFQR genes was highly observed among 84 isolates. The plasmid extracts of isolates might contain up to five genes per isolate. On the other hand, the plasmid extract of one isolate (AS-29) did not show any resistance gene and another isolate (AZ-18) which did not carry any plasmid. The latter isolate was recovered from the sputum of an admitted patient who stayed for only three days in Al-Azhar university hospital. (ii) Among CIP-sensitive isolates (n = 14) For the 14 CIP-sensitive isolates, analyses of occurrence and distribution of PAFQR genes in their corresponding plasmid extracts revealed that these plasmids harbor up to five genes, which gave 8 profiles of genes association (Fig. 5). On the other hand, one isolate did not harbor any PAFQR resistance gene. (iii) PAFQR gene sequences analyses PCR products of different genes from an isolate representing each profile 1, 2, and 3 were selected for sequencing (contain more life-threatening isolates), these included qnrA, qnrS, acc(6,)-ib-cr, oqxA, and oqxB. The amplicon sizes were 347, 255, 480, 489, and 480 bp, respectively. The BLAST of the NCBI (www. ncbi. nlm. nhi. gov) was used to search databases for detecting the similarity in nucleotides and amino acid sequences of these studied genes to those deposited in the databases. The BLAST results revealed that the oqxA and oqxB sequences were not identified previously in A. baumannii but they were identified in Klebsiella aerogenes strain NCTC9793 and Klebsiella pneumoniae, with an identity of 99.78% and 99.77%, respectively. On the other hand, the sequences of qnrA, qnrS, acc(6,)-ib-cr, and oqxA, showed homology to those of A. baumannii deposited in GeneBank database with identity ranged from 97.98 to 98.28% (Table 5).

Discussion
A. baumannii is a Gram-negative bacterium, can withstand a wide range of environmental conditions and also can survive on surfaces. These characters enable it to be implicated in many nosocomial infections and outbreaks. It is a strict aerobic organism 2 . The infection caused by A. baumannii is difficult to treat 2 . It has been recognized by the Infectious Disease Society of America as one of the six highly drug-resistant hospital pathogens 26 .
Antimicrobial agents use strategies are known to significantly reduce the frequency of bacterial infections in patients 27    www.nature.com/scientificreports/ specifically is becoming increasingly serious with their extensive use. FQs in the last forty years had shown good activity against A. baumannii isolates, However, resistance to these drugs has rapidly emerged 15 . FQs are widely prescribed medication in Egypt, and resistance to FQs has skipped pointedly 5 . The developing resistance of A. baumannii to antimicrobial agents has been described and this was attributed to the abundance of these antibiotics in multiple pharmaceutical markets 28 , besides their misuse 29 . A. baumannii infection is difficult to remedy, as of its everlasting fullness to acquire antimicrobial resistance due to the suppleness of its genome 30 . Many acquired resistance mechanisms have been reported for this pathogen and therefore, render it able to express MDR, XDR, or PDR phenotypes that were associated with significant morbidities and mortalities 4 . Therefore, this study aimed to determine the mechanisms behind A. baumannii resistance to FQs. For achieving this aim, the following objectives were studied: (i) the antibiotic-resistant phenotypes of A. baumannii recovered from In this study, a total of 1600 specimens were collected from different clinical hospitalized patients of two major university hospitals in Upper Egypt during the period between January 2014 and March 2019. Out of 151 Acinetobacter candidate isolates subjected to PCR, only 100 isolates were positive for bla OXA-51 and confirmed to be A. baumannii. The identification of A. baumannii phenotypically is difficult due to significant phenotypic overlapping with other species, which are genotypically closely related to each other 31 .
The study results revealed that the highest number of A. baumannii isolates was recovered from respiratory and urinary tracts as well as blood specimens indicating the involvement of this pathogen in infection of these sites. The 100 recovered A. baumannii isolates comprised 61 isolates from respiratory tract infection {Enotreacheal tubes (29), nasal (17), sputum (13), and throat (2)}, blood infection (12), and urinary tract infections (9). It has been reported 32 that the respiratory tract, blood, and urinary tract constitute the most predominant sources of A. baumannii pathogen. A. baumannii was also recovered from wounds or soft-tissue infections (6), skin (2), catheter-associated infections {central venous catheter (2), and urinary tract catheter (8)} in agreement with the results reported by 33 . Worldwide, it was reported that A. baumannii infection differs according to both the anatomical site and the clinical conditions of the patients 34 .
FQs in the last four decades had shown good activity against A. baumannii isolates, however, resistance to these drugs has quickly been rised 15 . FQs are a widely prescribed medication in Egypt, and FQs resistance has jumped up sharply 4,5,35 . Our findings emphasize that most A. baumannii isolates (71%) were recovered from patients treated for long period with FQs, while the rest (29%) of A. baumannii isolates were recovered from patients treated with other antimicrobial agents such as cephalosporins, imipenem, and the penicillin derivative (amoxicillin) or combination (amoxicillin-clavulanic acid). Un-rational use, low dose, and misuse of antimicrobial agents (empirical antibiotcs use by non hospitalized patients) result in the development of microbial resistance and so increase the risk of nosocomial Acinetobacter infections in hospitals 4 .
The increase in A. baumannii resistance constitutes a global issue 36 . Every year, the life of millions of hospitalized patients are seriously affected by incurable strains of A. baumannii 37 . As an infection control measure, continuous studies on the resistance profile of this organism are highly required and are a must for at least decreasing its devastating effect on the quality of medical treatment 4 . In the current study, the prevalence of resistance among the 100 recovered A. baumannii isolates against the tested antimicrobial agents was high. A resistance prevalence of ≥ 78% was recorded for the tested isolates against the 19 antimicrobial agents used. On the other hand, colistin proved to be the most effective anti-microbial agent against these isolates (95% of isolates were sensitive) followed by doxycycline (43% of isolates were sensitive). Our finding agrees with previous studies which stated that, A. baumannii pathogen is an opportunistic organism, often susceptible to colistin and having a low susceptibility to other antimicrobial agents. This organism is involved in radical morbidity and mortality 38 .
The emerged pan drug-resistant (PDR), extensively drug-resistant (XDR)-A. baumannii strains could be a leading cause of hospital-acquired infections by this opportunistic pathogen 4 . PDR and XDR are being recorded increasingly among A. baumannii isolates recovered from clinical 4,39 or environmental (such as soil) specimens 40 . In our finding, two isolates (2%) were detected as PDR, 68% isolates were XDR and 30% isolates were MDR. PDR and most of XDR-A. baumannii were isolated from Assiut university hospitals. The prevalence of MDR, XDR, and PDR A. baumannii could be attributed to the misuse of antimicrobial agents 41,42 , or due to the differences in rates of infections by the respective pathogens (mostly related to the degree of strict hygiene protocols applied in different hospitals) 43 , in addition to the plasticity and endless capacity of changes demonstrated in A. baumannii genomes 44 .
Genotyping method using ERIC-PCR technique was applied for determining the clonal relationship and diversity of isolated A. baumannii used in the present study. This technique can be used during nosocomial outbreaks to investigate if the involved isolates are genetically related or originated from the same strains 45 . The use of strain typing in infectious disease control decisions in hospitals is based on several assumptions, (i) whether the isolates associated with an outbreak are the progeny of a single clone, or (ii) have identical genotype, or (ii) epidemiologically unrelated so have different genotypes 46 . In our study, the phylogenetic dendrogram of ERIC-PCR showed that the isolates can be divided into three major clusters. This diversity might be due to multiple contamination sources by this organism, a finding that is different from that reported by some authors 47 Table 5. The PAFQR gene amplicons sequences of some selected A. baumannii isolates and their identity to sequences deposited in GeneBank database. The level of significance for the % of identity of the studied sequence to its homology in database was mentioned. (a) no successful sequence of qnrA could be obtained although its amplicon was sent twice to different laboratories. (b) at p-value ≤ 0.05 level of significance. (c) these genes gave negative results with profile number 1 isolate.

Gene (a)
GeneBank accession no. for the deposited studied sequence Strain showing sequence homology to the studied sequence % of the identity of the studied sequence to its homology in database (b) Profile In the current study, investigation of the mechanisms responsible for the resistance of A. baumannii isolates to FQs involved detection of target-site mutation occurrence and acquisition of PAFQR genes. The analyses of target site mutations in gyrA and parC gene sequences of twelve selected A. baumannii isolates showed 12 profiles. Two isolates were susceptible to CIP and had a wild-type profile of being have no mutations. Ten isolates were CIP-resistant, 9 of them (9/10; 90%) had (1 gyrA /1 parC) mutation: Ser 81 → Leu mutation for gyrA gene and Ser 84 → Leu mutation for parC gene. The remaining CIP-resistant isolate (1/10; 10%) had (0 gyrA /1 parC) mutation: Ser 84 → Leu mutation for parC gene as reported previously in North Egypt 48 . Previous studies stated that double mutations (1gyrA/1parC) could be sufficient to confer CIP resistance in A. baumannii isolates 23,48,49 and this in agreement with our results. . It was reported that these two mutations are enough to predict resistance to CIP and LEV 48 . However, gyrA and parC mutations did not occur in all resistant mutant strains and the resistance may be attributed to changes in outer membrane protein expression and drug efflux pumps 50 . All tested isolates had a silent mutation in one or more positions of either gyrA or parC or even both gyrA and parC that do not lead to a change in amino acid composition. Twenty-three isogenic mutations in gyrA genes were as follows: (6 Gly112, 7 Ala115, 4 Ile166, 4 Ala170, 2 Asp197) and eleven isogenic mutations in parC gene as follows: (2 Ala 52, 6 Leu 35, 1 Ala 127, 1 Gly 143, 1 Ala163). The isogenic mutations were identified previously in Gram-negative 51-53 , Gram-positive 54,55 , Mycobacterium 56 , Mycoplasma 57 , and A. baumannii 58 . Our findings showed that four amino acid affected by isogenic mutations were detected in gyrA, these included Ala, Asp, Gly, and Ile while the three amino acid-affected by isogenic mutations in parC included Ala, Gly, and Ile and occurred in different codon positions. A storm of silent mutations was identified previously in A. baumannii 50 . Isogenic or silent mutations may be due to that the patient's administrated insufficient doses of FQs in different periods which may induce bacterial resistance to the drug 58 .
The dissemination of mobile elements harboring resistance genes between Acinetobacter spp. is not fully understood. Many earlier investigations have given special emphasis to plasmid-mediated horizontal transfer of antibiotic resistance genes [59][60][61] . In the present study, 99%; 99/100 of A. baumannii have harbored plasmids. However, one isolate that represents 1% has not harbored any plasmids. Despite plasmids themselves may be insufficient to confer FQs resistance, PAFQR genes execute an important role in the procuration of resistance to FQs by facilitating the selection of additional chromosomal resistance mechanisms, leading to a higher level of quinolone resistance and enabling bacteria to become fully resistant 62 . Most importantly, PAFQR genes can spread horizontally among A. baumannii 10 . Resistance to quinolones can be mediated by plasmids 63 . Three kinds of PAFQR determinants have been described: qnr genes (qnrA, qnrB, and qnrS) that encode pentapeptide protein repeats, which protect the quinolone targets from inhibition 8,63 . Inactivation of fluoroquinolones occurs by acetylation with the common aminoglycoside acetyltransferase aac (6′)-Ib-cr 11 and can be pumped out by efflux pumps QepAB and OqxAB 12 . The first PAFQR gene type is qnr genes, the 86, ciprofloxacin-resistant isolates carried qnrA (66.27%; 57/86), qnrS (70.93%; 61/86), while qnrB was undetected in these isolates. Such a high prevalence reported here and in other studies performed worldwide among FQ-resistant A. baumannii isolates reflects their fetal role in the acquisition of FQs resistance genes. In disparity, the lower prevalence of qnr was reported in CIP-resistance A. baumannii isolates from Brazil (37.5%) and north Egypt 48.3% 64 . On the other hand, qnr genes were more frequently detected among the isolates of K. pneumoniae (70.4%) and E. coli (67.5%) 61,62,65,66 . The second PAFQR gene type is the aac(6')-Ib-cr gene, a new variant of common aminoglycoside acetyltransferase that acetylates piperazinyl substituent of some fluoroquinolones, including ciprofloxacin 11 . The bifunctional aminoglycoside and fluoroquinolone active variant aac (6')-Ib-cr catalyzes the acetylation of both drug classes 67 . PCR screening of isolates tested in the present study showed the prevalence of aac (6')-Ibcr (52.32%; 45/86) among CIP-resistance isolates. On the other hand, CIP-susceptible isolates harbored aac (6')-Ib-cr by only 21.42% (3/14), these isolates were already resistant to aminoglycosides. Our result agrees those obtained in Iran and Brazil 68, 69 but disagrees with that of Hamed et al. 25 in north Egypt. The third PAFQR gene types are those encoding the efflux pumps QepA, and oqxAB 12 . QepA is a proton-dependent transporter belonging to the major facilitator superfamily that causes hydrophilic quinolone resistance 12 , while OqxAB is a transmissible resistance-nodulation-division multidrug efflux pump that was found to reduce susceptibility to CIP and nalidixic acid 70 . PCR screening showed the prevalence of oqxA by 73.25% (63/86) and oqxB by 39.53% (34/86) while qepA was undetected in these isolates. Our results don't agree with that reported by Hamed et al. 25 since they were unable to detect oqxAB in A. baumannii and E. coli while they could detect both in Klebsiella spp 25 . qepA was not detected in A. baumannii , E. coli, and K. pneumoniae but detected in Enterobacter spp. in the study conducted by 71 . Summing up, high variation in the prevalence of PAFQR efflux genes among different microbial species significantly limits the treatment options of infected patients and provides a potential source for the horizontal spread of resistance 71 . The differences mentioned above can result from the geographical distance, surveillance strategies, and variability in following up antibiotic stewardship among organizations.
Our findings showed that, although some of A. baumannii isolates (14%; 14/100) were CIP-sensitive, they harbored PAFQR genes: qnrA (7/14), qnrS (7/14), aac (6')-Ib-cr (3/14), oqxA (12/14), and oqxB (11/14) with collective number reached five PAFQR genes per a single isolate. This finding agrees with the results reported by other authors who showed that imipenem sensitive A. baumannii isolates harbored certain resistance genes 72,73 and the same was exhibited by other members of gram-negative bacilli worldwide 74 . The occurrence of resistance genes in CIP-sensitive A. baumannii isolates represents a major problem as this could facilitates their horizontal transfer between A. baumannii and other members of gram-negative bacilli in hospitals.
Much earlier theory has given special affirmation for transferable antibiotic resistance among A. baumannii by plasmid-associated quinolone resistance determinant genes 13 25 . This result suggests the possibility of acquisition of transferable antibiotic resistance genes that may be carried on plasmids by A. baumannii.
Our results showed the existence of several PAFQR genes and their co-occurrence in A. baumannii recovered isolates. The distribution of tested PAFQR genes gave 26 profiles for the 86 CIP-resistant-and 8 profiles for the 14 CIP-sensitive isolates (both harbor up to five genes). The storm of association due to suppleness and the nonlimit capacity of A. baumannii for genome changes 75 , which may be caused by horizontal gene transfer 76 , the introduction of mobile genetic elements like plasmids which mediate new genes, integrative conjugative elements, and transposons 60,76 . These properties give rise to PDR, XDR, and MDR gene cassettes 77 .

Conclusion
MDR, XDR, and PDR A. baumannii isolates are becoming prevalent in a number of hospitals, among the reasons behind this spread could be due to the empirical and un-rational use of antimicrobial agents which usually occur among outpatients in addition to improper infection control applied measures. It was observed that chromosomal mutations in the sequences of GyrA and ParC encoding genes as well as the acquisition of PAFQR encoding genes are molecular resistance mechanisms demonstrated among fluoroquinolones resistant A. baumannii isolates. It is advisable to monitor the antimicrobial resistance profiles of pathogens causing nosocomial infections and properly apply and update antibiotic stewardship in hospitals and for outpatients to control infectious diseases and prevent development of the microbial resistance to antimicrobial agents.