The First Saudi Study Investigating the Plasmid-borne Aminoglycoside and Sulfonamide Resistance among Acinetobacter baumannii Clinical Isolates Genotyped by RAPD-PCR: the Declaration of a Novel Allelic Variant Called aac(6ʹ)-SL and Three Novel Mutations in the sul1 Gene in the Acinetobacter Plasmid (s)

Introduction

Acinetobacter baumannii (A. baumannii) is one the most clinically important non-enteric Gram-negative pathogens that causes a wide range of nosocomial infections (NIs) especially among debilitated patients who are admitted to intensive care units (ICUs).¹

In clinical settings, A. baumannii is the most pathogenic and commonly encountered species in the genus Acinetobacter followed by Acinetobacter calcoaceticus (A. calcoaceticus) and Acinetobacter lwoffii (A. lwoffii).^2,3

A. baumannii was the only species in the genus Acinetobacter that intrinsically harbors bla_OXA-51 which was then used for identification and differentiation of A. baumannii from the other Acinetobacter species.^4,5

Multidrug-resistance (MDR), extensive drug resistance (XDR), and pan drug-resistance (PDR) patterns are common among A. baumannii due to the harboring of intrinsic resistance genes,² genetic mutations, and/or horizontal gene transfer (HGT) by mobile genetic elements (MGEs) like insertion sequence elements (ISE), plasmids, and transposons.⁶

High morbidity and mortality by A. baumannii are commonly occured in hospital settings¹ due to limited therapeutic options for treatment of infections caused by either MDR, XDR, and PDR strains⁷ so, in 2013, MDR A. baumannii has been termed by the center of disease control and prevention (CDC)⁸ as a “serious threat” and in 2016, carbapenem-resistant A. baumannii (CRAB) has been considered as a “level I: critical priority pathogen” in the World Health Organization (WHO) priority pathogen’s list.⁹

Aminoglycosides are broad-spectrum antibiotics that mainly achieve their action by selective inhibition of bacterial protein synthesis via selective binding to the 30S subunit of the bacterial ribosome.¹⁰ As aminoglycosides are positively charged molecules due to their amino groups, it has also been known that aminoglycosides disrupt bacterial cell membranes resulting in pore formation in the bacterial’s cell membrane resulting in bacterial cell death.^11,12

Aminoglycosides have been widely used, in clinical settings after the emergence of β-lactam and quinolone-resistant strains, for the eradication of MDR Gram-negative isolates despite their nephrotoxic and ototoxic complications.¹³ According to the previously mentioned, the aminoglycoside resistant strains have emerged worldwide.¹⁴

Bacterial resistance to aminoglycosides can be mediated by different mechanisms like overexpression of efflux pumps, downregulation of outer membrane proteins (OMPs), ribosomal target modification, and enzymatic inactivation by aminoglycoside modifying enzymes (AMEs).⁵

Production of AMEs is the most important mechanism for aminoglycoside resistance in which the dissemination of AMEs occurs via MGEs that commonly harbor other resistance determinants⁵ to other antibiotic classes leading to the failure of the other antibiotics to cure infections, leading to the selection and the ease of dissemination of aminoglycoside resistant strains in hospital settings.^15,16

The AMEs abolish the activity of aminoglycosides by their ability to acetylate, phosphorylate and adenylate the -OH or -NH₂ of the 2-deoxystreptamine nucleus or the sugar moieties¹⁷ via aminoglycoside acetyltransferases (ACC), aminoglycoside nucleotidyltransferases (ANT), and aminoglycoside phosphotransferases (APH), respectively resulting in poor binding of the aminoglycosides to the bacterial ribosome with subsequent failure to achieve their action.¹⁸

Resistance to aminoglycosides can also be achieved through ribosomal target modification via methylation of 16S rRNA by methyltransferases (MT) such as aminoglycosides resistance methyltransferase A (armA)¹⁴ and 16S rRNA methyltransferases B (rmtB).^13,19

Sulfonamides are the oldest antimicrobial agents that were effectively used in the treatment of bacterial infection since 1932²⁰ in which they exert their antimicrobial activity by inhibition of bacterial's folic acid synthesis²¹ via their binding with dihydropteroate synthase (DHPS) due to their structural similarity with P-amino-benzoic acid (PABA) which is utilized by bacteria for the biosynthesis of folic acid.²²

Resistance to sulfonamides can be either chromosomally or plasmid-mediated.²¹ Chromosomal -mediated sulfonamide resistance involves either a genetic mutation in the DHPS gene known as folP gene or the acquisition of an altered DHPS encoding gene known as sul gene²³ via MGEs.²⁴

As mentioned above, plasmid-mediated sulfonamide resistance is achieved by the sul gene which can be translocated between plasmids and chromosomes by MGEs.²⁵ Plasmids harboring the sul gene can be disseminated between same/different bacterial species or different bacterial genera by either conjugation of transformation.²⁴

There are three variants of the sul gene namely; sul1, sul2, and sul3.²⁴ It was reported that the sul1 gene is carried on large conjugative plasmids and class 1 integrons, while the sul2 gene was found to be carried on small non-conjugative plasmids and large conjugative plasmids as well.²⁵ The sul3 gene was reported in 2003 for the first time to be carried on a 54-kb conjugative plasmid in Escherichia coli (E. coli).²⁶

Random amplified polymorphic DNA-polymerase chain reaction (RAPD-PCR) is a rapid molecular technique which is commonly used for molecular typing and epidemiological tracing to discriminate between bacterial isolates based on the use of short oligonucleotide primers ranging from 8 to 15 nucleotides²⁷ that randomly bind to different regions on the whole bacterial genome with a subsequent amplification of many DNA fragments that migrate to different distances on agarose gel generating a complex DNA band patterns.²⁷

Due to the lack of the published data about the genetic background of aminoglycoside and sulfonamide resistance among A. baumannii clinical isolates in the Kingdom of Saudi Arabia (KSA) so, the current study adopted such issue to investigate the the existence of AMEs, 16S rRNA methyltransferases, and altered DHPS   plasmid encoding genes in A. baumanniiclinical isolates recovered from a large tertiary care hospital in the Western area, Taif, KSA. Also, the epidemiological typing using RAPD-PCR was performed to track the dissemination of the investigated isolates between the different hospital locations.

Materials and Methods

Ethical Statement

The current study was performed according to ethical approval No. 43-001 following the regulations of the ethics committee at Taif University that accredited by the national committee for bioethics with No. (HAO-02-T-105).

Bacterial Strains

The current study was conducted on 40 clinical isolates of A. baumannii that were recovered from the patients who were admitted to or attended various medical departments at a large tertiary care hospital in Taif, KSA during the period from October 2016 to May 2017. All isolates were recovered from the routinely investigated clinical specimens sent to the microbiology laboratory as a part of routine hospital laboratory procedures. The clinical isolates were recovered from sputum (n = 20), blood (n = 4), tracheal aspirate (n = 2), urine (n = 4), peritoneal fluid (n = 1), wound swab (n = 6), and catheter tip (n = 3). All strains were cryopreserved at −80 °C in tryptic soy broth (Scharlau, Spain) containing 15% glycerol to be used when needed.

Isolation and Identification

All strains were primarily isolated on blood agar then purified on MacConkey’s agar (Oxoid, UK). The isolates were provisionally identified to the genus and species level by the Vitek 2^® system (BioMérieux, France) and API 20NE^® (BioMérieux, France). Molecular confirmation to the species level was achieved via the amplification of the chromosomally- encoded gene that namely; bla_OXA-51 using the specific primers (Table 1) that previously described.²

Table 1 Primer Sequence and Cycling Conditions Used in PCR

Antimicrobial Susceptibility Testing (AST)

All clinical isolates were tested for their susceptibility to ten different antibiotics namely; gentamicin, amikacin, streptomycin, sulfamethoxazole/trimethoprim, ciprofloxacin, levofloxacin, meropenem, tigecycline, polymyxin, and colistin representing 6 different antibiotic classes. Susceptibility testing was performed by the broth microdilution method as previously described²⁸ where Klebsiella pneumoniae (K. pneumoniae) ATCC 700603 and Pseudomonas aeruginosa (P. aeruginosa) 27853 were employed as standard control strains. The susceptibility results were interpreted according to the breakpoints of the Clinical Laboratory Standards Institute (CLSI)²⁹ for all tested antibiotics, while the breakpoint for streptomycin was interpreted according to Yang et al.³⁰

Phenotypic Characterization of Antibiotic Resistance Pattern

The investigated clinical isolates were considered to have MDR phenotype when they resisted three different classes of antibiotics except for carbapenems, while XDR phenotype was considered when MDR isolate was resistant to meropenem. PDR phenotype was considered if XDR isolate was colistin and tigecycline resistant.²

Phenotypic Characterization of Aminoglycoside Resistance Level

Based on the definition proposed by Nie et al,³¹ Upadhyay et al³² and Doi et al³³ all isolates that exhibited MIC value to gentamicin and amikacin ≥512 µg/mL were considered to have high level of aminoglycoside resistance (HLAR) phenotype.

Extraction of Total DNA

Total DNA (chromosomal and plasmid DNA) was extracted as previously described³⁴ to be used in PCR reactions for the amplification of bla_OXA-51 and the genotyping by RAPD-PCR technique. The extracted DNA was kept in a sterile DNase-free 0.5 mL tube at −20 °C until use.

Extraction of Plasmid DNA

Plasmid extraction was performed to investigate the existence of AMEs, 16S rRNA methyltransferases encoding genes and, the sul1 gene as well in which each isolate was cultured in 5 mL of lysogeny broth (LB) (Himedia^®, India) to increase the plasmid yield. The inoculated LB tubes were then placed in the shaking incubator at 37 °C for 16 h. After that, the culture tubes were centrifuged at 13,200 rpm for 3 min. The bacterial cell pellet was then washed twice with phosphate buffer and any remaining buffer over the pellet was pipetted out. The plasmid DNA was extracted from the washed pellet by GeneJET^® Plasmid Miniprep Kit (Thermo Fischer Scientific, USA) according to the manufacturer’s instructions.

Genotypic Detection of The Plasmid Mediated Aminoglycosides and Sulfonamide Resistance

The plasmid DNA from each isolate was used as a template to investigate seven different plasmid-borne AMEs encoding genes namely; ant(3″)-I, aph(3ʹ)-VI, aac(3)-II, aac(6ʹ)-II, aac(6ʹ)-Ib, aad(2)-Ia, aad(4)-Ia and two 16S rRNA methyltransferases encoding genes namely; armA and rmtB using the specific primers and cycling conditions listed in Table 1. Also, all isolates were investigated for the sul1 gene.

PCR

The PCR was performed in a 0.2 DNase-free PCR tube containing 30 μL of the total reaction mixture that was composed of 6 µL of 5x master mix (Solis BioDyne, Estonia), 4 µL of DNA template equivalent to about 12–15 ng, 0.9 µL from each of the forward (10 pmol/μL) and the reverse (10 pmol/μL) primer (Table 1).

The volume of the reaction mixture was completed to 30 µL by adding 18.2 µL of sterile DNase-free water. Five strains from the laboratory culture collection provided by the microbiology laboratory at Taif University were used as positive controls for the 11 investigated genes. E. coli ATCC 25922 was used as a negative control strain. For the RAPD-PCR technique, 1.8 μL of the HLWL74 primer³⁵ was used.

Gene Amplification and Electrophoresis

Target genes were amplified using Mastercycler gradient^® (Eppendorf, Germany), oligonucleotide primers (Korea, Seoul), and cycling conditions listed in Table 1. All the amplified PCR products were run on 1% agarose (Scharlau, Spain) gel containing 500 ng/mL ethidium bromide. The random amplified DNA fragments generated by the RAPD-PCR were run on 2.5% agarose gels to achieve well band separation.

Gel Extraction

All PCR products subjected to capillary electrophoresis sequencing (CES) were initially extracted from the agarose gel by a glass fiber membrane using gel extraction SV kit (MG^®, Korea) according to the manufacturer's instructions with a maximum yield of 90%.

DNA Sequencing

The extracted amplified PCR products of the aac(6ʹ)-Ib and the sul1 genes were subjected to CES in one direction utilizing the forward amplification primers^36,39 (Table 1) using 96 capillary type ABI PRISM^® 3730XL DNA Analyzer (Applied Biosystems, USA).

DNA Sequence Correction

All the released DNA sequencing ab1 files were manually corrected using the FinchTv software⁴⁰ version 1.5.0 (Geospiza Inc, USA).⁴⁰ The corrected sequences were uploaded to the National Center for Biotechnology Information (NCBI)⁴¹ using the blastn service in which the query DNA sequence was identified and corrected.

Translation of The Sequenced aac(6ʹ)-Ib and sul1 Genes

The corrected DNA sequences of aac(6ʹ)-Ib and the sul1 genes were uploaded to open reading frame (ORF) finder website⁴² to obtain the corresponding amino acid sequences using the following search parameters; (i) minimal ORF length of 300 and 600 for aac(6ʹ)-Ib and the sul1 genes, respectively; (ii) genetic code was bacterial and archaeal; (iii) ORF start codon to use was ATG only.

NCBI Database Search Criteria and Reference Sequence (RefSeq) Selection

The corresponding amino acid sequences of the sequenced aac(6ʹ)-Ib and the sul1 genes were searched in the NCBI database against the corresponding amino acid sequences reported only in A. baumannii using the blastp service with selection criteria of sequence identity and query coverage of ≥98% in which AAG33663.1 and WP_063855115 were used as the amino acids RefSeq for aminoglycoside 6ʹ-N-acetyltransferase-Ib and sulfonamide-resistant DHPS Sul1, respectively.

Identification of the Mutation Sites in the aac(6ʹ)-Ib and the sul1 Genes

The translated query amino acid sequences were subject to multiple sequence alignment (MSA) against the selected reference sequence using the Jalview software⁴³ utilizing the fast Fourier transform (MAFFT) web service.

GenBank Sequence Accession Numbers

The partial sequences of the mutated aac(6ʹ)-Ib and sul1 genes were submitted to the GenBank⁴⁴ via the submission portal⁴⁵ on the NCBI website in which each mutated gene has assigned a specific GenBank accession number and specific protein accession number.

Cluster Analysis of RAPD Fragments

The amplified DNA fragments generated by RAPD-PCR were analyzed as previously described³⁶ using BioNumerics^® 7.5 software⁴⁶ in which the clonal relatedness between the 40 A. baumannii clinical isolates was determined based on the generated dendrogram using the unweighted pair-group method with arithmetic average (UPGMA) utilizing the Dice coefficient .

Results

Isolation and Identification

The present study revealed that all the recovered isolates belonged to the genus Acinetobacter based on the phenotypic profiles obtained from the Vitek2 and API 20NE systems, also all isolates were genetically confirmed to be A. baumannii in which 100% of isolates tested positive for the chromosomal -borne bla_OXA-51 gene (Figure 1).

Figure 1 PCR products of amplified blaOXA-51used in molecular confirmation of A. baumannii. L1; 100 bp DNA ladder supplied from Solis BioDyne, Estonia; lanes 60–65 and 66–72 represent the isolate number .

AST and Phenotypic Characterization of Resistance Pattern

As regards the tested aminoglycosides, the current study revealed that 55% (22/40), 57.5% (23/40), and 95% (38/40) of A. baumannii clinical were resistant to gentamicin, amikacin, and streptomycin, respectively.

Apart from tigecycline to which all isolates were sensitive, the susceptibility testing revealed that polymyxin B and colistin were the most effective agents following tigecycline in which 85% (34/40) and 67.5% (27/40) of isolates were sensitive, respectively.

Regarding the other tested antibiotics, it was found that meropenem and ciprofloxacin were the lowest effective agents in which 97.5% (39/40) of isolates were resistant for each. Also, the current study revealed that 70% (28/40) and 77.5% (31/40) of isolates were resistant to levofloxacin and sulfamethoxazole/trimethoprim, respectively.

Fortunately, no isolate exhibited a PDR pattern in which all isolates were sensitive to at least either tigecycline or colistin. On the other hand, it was found that 90% (36/40) of isolates exhibited an XDR pattern in which 67.5% (27/40) and 22.5% (9/40) of isolates were resistant to four and five different antibiotic classes, respectively. Only 5% of isolates exhibited MDR pattern in which only two isolates were resistant to only three antibiotic classes as shown in Table 2 and 3.

Table 2 Susceptibility Pattern of A. baumannii Clinical Isolates

The values of MIC₅₀ for the tested aminoglycosides ranged from 16 to 1024 μg/mL in which the MIC₅₀ for gentamicin, amikacin, and streptomycin were 16, 64, and 1024 μg/mL, respectively. On the other hand, the MIC₉₀ for all of the tested aminoglycosides was 1024 μg/mL.

It was also found that the values of MIC₅₀ and MIC₉₀ for the tested polymyxins ranged from ≤0.5 to 2 μg/mL and 8 to 32 μg/mL, respectively.

Regarding sulfamethoxazole/trimethoprim, it was found that the values of MIC₅₀ and MIC₉₀ were 13.4/256 μg/mL and 53.8/1025 μg/mL, respectively. In addition, obvious elevated values of MIC₅₀ and MIC₉₀ for ciprofloxacin and levofloxacin were observed (Table 2).

Phenotypic Characterization of Aminoglycoside Resistance Level

The current study exhibited that 42.5% of isolates showed HLAR in which the MIC values of 17 isolates were ≥512 µg/mL for gentamicin and amikacin, while 52.5% of isolates exhibited LLAR in which 21 isolates were streptomycin-resistant and exhibited MIC value <512 µg/mL to gentamicin and amikacin. On the other hand, 5% (2/40) of isolates were sensitive to all of the tested aminoglycosides.

Genotypic Detection of Plasmid Mediated Aminoglycosides and Sulfonamide Resistance

The current study revealed that aph(3ʹ)-VI and aac(6ʹ)-Ib gene variants [aac(6ʹ)-Ib and aac(6ʹ)-SL] (Figure 2 and Supplementary Figures 1–3) were the most prevalent plasmid encoding AMEs in which 90% (36/40) and 87.5% (35/40) of isolates tested positive, respectively (Table 3). On the other hand, ant(3″)-I and armA (Figure 2 and Supplementary Figures 4–6) ranked the 3rd and the 4th most prevalent plasmid-borne aminoglycoside resistance encoding genes in which 85% (34/40) and 45% (18/40) of isolates tested positive, respectively.

Table 3 Resistance and Genetic Profile of A. baumannii Clinical Isolates

Figure 2 PCR products of the genes coding for aminoglycoside and sulfonamide resistance. L1; 100 bp DNA ladder supplied from Solis BioDyne, Estonia; lanes 1–5 represent the amplified resistance genes as stated above each PCR product.

The lowest detected gene responsible for aminoglycoside modification by adenylation was aad(4ʹ)-Ia in which 5% (2/40) of isolates tested positive. The other investigated aminoglycoside resistance encoding genes were not detected in which all isolates tested negative for aad (2ʹʹ)-Ia, aac(6ʹ)-II, aac(3)-II, and rmtB.

As regards the investigation of plasmid-mediated sulfonamide resistance, the current study revealed that 15% (6/40) of the isolates were found to harbor the altered DHPS encoding gene namely; the sul1 gene (Figure 2 and Supplementary Figures 7 and 8).

Correlated Phenotypic and Genotypic Resistance Profiles

The current study revealed 19 different resistance genotypes (Table 4) among the 38 aminoglycoside resistant isolates in which the resistance genotype A (bla_OXA-51, ant(3″)-I, aph(3ʹ)-VI, aac(6ʹ)-Ib) was the most prevalent to which 4–6 antibiotics were resistant. On the other hand, the resistance genotype B (bla_OXA-51, ant(3″)-I, aph(3ʹ)-VI, aac(6ʹ)-SL) and the resistance genotype C (bla_OXA-51, ant(3″)-I, aph(3ʹ)-VI, armA, aac(6ʹ)-Ib) were the second prevalent resistance genotypes to which 4–6 and 6–9 antibiotics were resistant, respectively.

Table 4 Combined Phenotypic and Genotypic Resistance Profile Among Aminoglycosides Resistant A. baumannii Clinical Isolates

Detection of the Novel aac(6ʹ)-SL Allelic Variant in The Acinetobacter Plasmid

Only 34.29% (12/35) of the isolates which tested positive for acc(6ʹ)-Ib were found to harbor a common missense mutation in position 102 in which leucine (hydrophobic) was substituted for serine (neutral-polar)amino acid in position 102 indicating a novel allelic variant that was named as aac(6ʹ)-SL due to the presence of serine (S) amino acid instead of leucine in position 102 (L102S) and the normal presence of leucine (L) in position 117 (Figure 3).

Figure 3 Multiple sequence alignment of the amino acids sequences of the 12 mutated aminoglycoside 6'-N-acetyltransferase-Ib [AAC(6')-Ib] encoded by the novel aac(6')-SL allelic variant. The code AAG33663.1 represents the accession Id of the reference AAC(6')-Ib protein. The codes QZB49340-QZB4934051 represent the newly  accession Ids assigned  by the GenBank for the mutated AAC (6')-Ib protein detected in the current study.

Detection of the Mutation Sites in sul1 Gene in The Acinetobacter Plasmid

The sequencing of the sulfonamide-resistant DHPS encoding gene; namely sul1revealed a common novel misense mutation in 66.6% (4/6) of isolates tested positive in which cysteine (neutral-polar) was substituted for glycine (hydrophobic) in position 98 (C98G). Other two novel missense mutations were detected in only one isolate (Acb_96) in which aspartate (acidic-charged) was substituted for asparagine (neutral-polar) in position 40 (D40N), secondly; alanine (hydrophobic) was substituted for threonine (neutral-polar) in position 233 (A232T) as shown in Figure 4.

Figure 4 Multiple sequence alignment of the amino acid sequences of the mutated sulfonamide-resistant dihydropteroate synthase (DHPS). The code WP_063855115.1 represents the accession Id of the reference protein sequence. The codes QZH81639- QZH8163942 represent the newly accession Ids assigned by the GenBank for the mutated sulfonamide-resistant DHPS detected in the current study.

GenBank Accession Numbers of the the Novel aac(6ʹ)-SL Allelic Variant

The mutated aac(6ʹ)-Ib genes that detected in the current study and named acc(6')SL allelic variants were assigned the following gene accession numbers by the GenBank: MZ820065 MZ820066, MZ820067, MZ820071, MZ820072, MZ820064, MZ820068, MZ820063, MZ820069, MZ820070, MZ820073 and, MZ820074 for the isolates No. Acb73, Acb75, Acb81, Acb75, Acb80, Acb82, Acb83, Acb90, Acb91, Acb97, Acb108, and Acb109, repectively.

GenBank Accession Numbers of sul1 Genes

The mutated sul1 genes that detected in the present study were assigned the following gene accession numbers by the Genbank: MZ751055, MZ751056, MZ751057, and MZ751058 for the isolates No. Acb62, Acb81, Acb82, and Acb96, repectively.

Cluster Analysis of RAPD Fragments

The amplified fragments generated by RAPD-PCR (Figure 5 and Supplementary Figure 9) were able to classify the 40 clinical isolates of A. baumannii into three main clusters (Figure 6) in which 92.5% (37/40) of the isolates belonged to cluster II, while 5% (2/40) of the isolates were related to cluster I exhibiting identical band profile. Only one isolate was related to cluster III.

Figure 5 DNA fingerprint profile generated by RAPD-PCR for A. baumannii clinical isolates. L1 and L2, represent 100 bp and 1 kb DNA ladders, respectively that were supplied from Solis BioDyne, Estonia. Lanes 94 to 99 and 100 to 110 represent the isolate number.

Figure 6 Clonal relatedness between A. baumannii clinical isolates based on the constructed unweighted pair-group method with arithmetic average (UPGMA) dendrogram for the generated RAPD-PCR banding profile. The aac(6')-Ib gene detected in the isolates Acb_73, Acb_75, Acb_80, Acb_81, Acb_83, Acb_84, Acb_90, Acb_91, Acb_97, Acb_108, Acb_109  was  termed more specifically as aac(6')-SL allelic variant in this study.

As cluster II was the main cluster to which 37 isolates were related, this cluster was found to include 24 different banding profiles in which 14 isolates exhibited 14 single different unique profiles with different RAPD-PCR banding patterns, whereas 23 isolates showed 10 different combined profiles (include more than one isolate within the same profile) exhibiting specific identical RAPD-PCR banding pattern, for instance, the isolates Acb_66, Acb_98, Acb_68, Acb_108 were included in one combined profile (Figure 6) exhibiting an identical RAPD-PCR banding pattern, also the isolates Acb_110 and Acb_67, Acb_107 and Acb_99, Acb_71 and Acb_73, Acb_90 and Acb_69, Acb_94 and Acb_65, Acb_62 and Acb_63, Acb_70, Acb_74 and Acb_76, Acb_75 and Acb_82, Acb_80 and Acb_81 exhibited nine different combined profiles with a specific identical RAPD-PCR banding pattern for each.

Discussion

The problem of antibiotic resistance represents one of the biggest obstacles facing the health system in most countries of the world, if not all, as this problem increases the financial burden of the health system due to the prolonged stay of the infected patients in the hospitals,³⁴ especially in ICUs, as the patients infected with MDR bacterial strains usually suffer from life-threatening conditions⁴⁷ that may require admission to to ICUs and placing them on ventilators leading to an increase in the cost of treatment . A. baumannii is one of the most important members of those MDR bacterial strains and therefore the research on A. baumannii occupies a great importance worldwide, especially in KSA.²

Based on reviewing the available cited literature and according to the Saudi review conducted by Yezli et al⁴⁸ in 2014 and Ibrahim et al⁴⁹ in 2019 there is no data about the resistance pattern and the genetic background of the plasmid-mediated aminoglycoside and sulfonamide resistance among A. baumannii clinical isolates in the Taif area so, we were prompted to investigate such a topic in an attempt to contribute in building the data concerning such an issue.

By reviewing the cited literature we found that gentamicin, tobramycin, and amikacin were officially introduced to KSA in 1975, 1983, and 1984, respectively.⁵⁰ Based on the data retrieved from previous literature, it was found that the emergence of gentamicin resistant strains began to appear in KSA slowly from low-level to high-level of resistance in which Moaz et al⁵⁰ and Memish et al⁵¹ reported that the resistance level to gentamicin was 26.6% (254/959) during 1983 and 1984 and 31.2% (1930/6189) in 2009, respectively among Saudi P. aeruginosa clinical isolates without any data about the resistance among A. baumannii.⁵⁶

It is not surprising to report a high incidence of aminoglycoside resistance among A. baumannii clinical isolates in which the recent studies confirmed the presence of certain intrinsic aminoglycoside resistance genes on the Acinetobacter chromosome. For instance, in 2017⁵² it was reported that the gene encoding for ANT (3″)-II is located on the Acinetobacter chromosome and was found to be transferred horizontally among the Acinetobacter spp. by homologous recombination. Based on the previous data, intrinsic and acquired resistance to aminoglycosides via AMEs is easy to occur via either intrinsic and/or acquired resistance.

The current study demonstrated that 55% of A. baumannii clinical isolates were gentamicin resistant, this finding is relatively consistent to a high extent with the findings of Haseeb et al⁵³ and Abdalhamid et al⁵⁴ who reported that 46% and 54.6% of A. baumannii clinical isolates recovered from Makkah and Dammam were gentamicin resistant, respectively.

About amikacin resistance, our study showed a low rate of amikacin resistance as compared with the recent Saudi study conducted by Almaghrabi et al⁵⁵ in which 74.5% of A. baumannii clinical isolates collected from Aseer region in the southwest of the Kingdom were amikacin resistant, while only 27.5% of isolates included in the present study were amikacin resistant indicating that rate of amikacin resistance in the west region of the Kingdom is much lower than that reported in the southwest region at the time of the study.

Based on the definition proposed by Nie et al,³¹ Upadhyay et al³² and Doi et al³³ a HLAR in A. baumannii is considered when the MIC value to gentamicin and amikacin ≥512 µg/mL so, the present study exhibited that 42.5% of isolates showed HLAR in which the MIC values for 17 isolates were ≥512 µg/mL to either gentamicin and/or amikacin, this finding is lower than the report of Upadhyay et al³² who showed that 79.2% of investigated A. baumannii clinical isolates exhibited HLAR.

Due to the scarcity of the studies dealing with the genetic background of AMEs among A. baumannii clinical isolates in KSA so, in this section, the findings of the current study were discussed in comparison with the findings of the other studies that were performed in the countries near to KSA like the Gulf countries and the countries from which most laborers in the KSA are hired from, like Egypt, India, and Pakistan.

The current study revealed a high prevalence of genes encoding for AMEs in which all isolates that exhibited HLAR resistance were found to harbor at least one of plasmid encoding genes for AMEs and/or 16S rRNA methylase, this finding is closely related to the finding of the recent Indian³² study which reported that 83.8% of HLAR A. baumannii clinical isolates harbored AMEs and/or 16s methyltransferase encoding genes.

The molecular investigation of AMEs in the present study showed that aph (3ʹ)-VI and aac(6ʹ)-Ib were the most prevalent AMEs encoding genes in which 90% and 87.50% of isolates tested positive, respectively, a closely related finding was reported by Polotto et al⁵⁶ in which aph (3ʹ)-VI and aac(6ʹ)-Ib were the most prevalent AMEs encoding genes that were detected in 55% and 47% of A. baumannii clinical isolates, respectively.

As regards the prevalence of nucleotidyl transferases encoding genes, the current study revealed that 85%, 5%, and no isolates were found to harbor ant(3′′)-I, aad(4ʹ)-Ia, and aad(2ʹ)-Ia, respectively, these findings are highly consistent with the findings of Nie et al³¹ who reported a high prevalence rate for ant(3′′)-I in which 95.1% of their isolates tested positive, while all isolates tested negative for aad(2ʹ)-Ia and aad(4ʹ)-Ia, these findings are similar to our findings that demonstrated a low prevalence rate for aad(4ʹ)-Ia and failure to demonstrate aad(2ʹ)-Ia. Based on the previouly mentioned data, the authors of the current study concluded that ant(3′′)-I is the most commonly detected variant of ANTs encoding genes among A. baumannii.

In contrast to the findings of the current study and the findings of Nie et al³¹ a relatively low prevalence rate of 33% for ant(3ʹ)-Ia was reported by the recent Iranian study conducted by Jouybari et al⁵⁷ among A. baumannii clinical isolates.

Concerning the genes encoding for 16S rRNA methylases, only armA gene was detected in the current study in which 45% (18/40) of the isolates tested positive, while rmtB was not detected. Closely related findings were conducted by the previous Chinese study³¹ which reported that 59.54% (103/173) of A. baumannii clinical isolates tested positive for armA, while no isolates tested positive for rmtB .

The current study demonstrated that the existence of AMEs encoding genes is not an evidence for the incidence of HLAR pattern at least in this study in which aph(3ʹ)-VI, aac(6ʹ)-Ib, ant(3')-I, and aad(4ʹ)-Ia were detected in 76.19% (16/21), 71.43% (15/21), 90.48% (19/21), and 9.52% (2/21) of isolates that did not exhibit a HLAR pattern. In addition to, 80.95% (17/21) of isolates that did not exhibit a HLAR were found to harbor 2–4 AMEs encoding genes assuring that that the co-existence of AMEs is not a condition  for the incidence of HLAR pattern.

On the contrary, the current study demonstrated that 94.11% (16/17) of isolates that showed HLAR pattern were found to harbor an armA gene indicating that there is a close relationship between the existence of armA gene and the incidence of HLAR, this conclusive finding was also reported by Doi et al.³³

In the current study, 19 different genetic aminoglycoside resistance profiles were detected. In the same context, 22 aminoglycoside resistance genetic profiles were reported by the Iranian study that was conducted by Jouybari et al⁵⁷ indicating that the Iranian A. baumannii clinical isolates are more genetically diverse than Saudi isolates.

The diverse aminoglycoside genetic profiles of the investigated isolates in the current study suggest that foreign labor recruited from the neighboring countries may contribute to the introduction of genetically diverse isolates.

In addition to what was mentioned in the previous paragraph, the ease of transmission of AMEs and 16S rRNA methylase encoding genes via plasmids or other MGEs like integrons and transposons is another possibility that explains the wide diversity of aminoglycoside genetic profiles that detected in the current study. All the previous possibilities limit the effectiveness of aminoglycosides in the treatment of infection caused by aminoglycoside resistant isolates so, effective measures should be implemented to overcome the spread of antibiotic resistant strains, especially in the hospital environment by (i) following infection control measures, (ii) following the antibiotic stewardship policy, (iii) avoiding the irrational use of broad-spectrum antibiotics except for cases which require the use of these antibiotics.⁵⁸

As the current study revealed that ciprofloxacin was the least effective agent to which 97.5% of isolates were resistant so, the authors decided to sequence the amplified aac(6ʹ)-Ib gene to investigate if the previously reported aac(6ʹ)-Ib-cr variant (responsible for ciprofloxacin acetylation)⁵⁹ is the cause for that finding, but instead it was found that only 34.29% (12/35) of isolates tested positive for aac(6ʹ)-Ib have showed a new allelic variant that was named aac(6ʹ)-SL due to the substitution of leucine (L) by serine (S) in position 102 (L102S).

The detected aac(6ʹ)-SL variant was considered to be a novel variant for aac(6ʹ)-Ib in which all the previously reported aac(6ʹ)-Ib allelic variants⁵⁹ associated with ciprofloxacin resistance was found to have leucine (L) in position 117 and either arginine (R) or tryptophan (W) in position 102 which differ from our findings that revealed the presence of serine in position 102 and leucine in position 117.

The authors of the current study thought that the newly detected aac(6ʹ)-SL allelic variant may greatly contribute to 100% ciprofloxacin resistance among the isolates that horboured such novel allelic  variant based on the fact that, the missense mutation in the position 102 in the wild type of aac(6')-Ib gene that resulted in the previously reported aac(6')-Ib-cr variant has been proved to acetylate ciprofloxacin with subsequent faliure of ciprofloxacin to achieve its action and hence ciprofloxacin resistance.

As regards to sulfonamide resistance, the present study revealed that 77.5% (31/40) of the investigated isolates were resistant to sulfamethoxazole/trimethoprim, in spite of  the existence of the plasmid borne sul1 gene in 19.53%(6/31) of the sulfamethoxazole/trimethoprim resistant isolates indicating that the resistance to sulfonamid in the isolates tested negative for the sul1gene is greatly contributed to the existence of other resistance determinants -like efflux pump system⁶⁰ and/or other sul gene variants like sul2 and/or sul3.²¹

In contrast to the current study, which revealed that 15% of isolates tested positive for the sul1 gene, a relatively high prevalence for the sul1 gene was reported in A. baumannii according to the recent Iranian study conducted by Tavakol et al⁶¹ who declared that 63.63% A. baumannii isolates were found to harbor the sul1 gene indicating that the sul1 gene is less prevalent in Saudi Arabia than Iran among A.baumanniiduring the period of the study.

The molecular epidemiological investigation by RAPD-PCR revealed that 92.5% of the investigated isolates were confirmed to be nosocomially transmitted as they were included under one cluster (cluster II) with high clonal similarity, despite that these isolates were recovered from different hospital locations at different time intervals of approximately 7 months starting from 19 October 2016 to 14 May 2017, which confirms the circulation of these isolates in the hospital environment during the period of the study .

Limitation of the Study

Gene cloning and expression experiments for the detected novel aac(6ʹ)-SL allelic variant is required to confirm the role of this novel gene variant in ciprofloxacin resistance.

Conclusion

This is the first Saudi study to shed light on the plasmid-borne aminoglycoside resistance via screening of plasmid-encoding genes for AMEs and 16S rRNA methyltransferases among A. baumannii clinical isolates. The current study demonstrated a close association between the existence of the armA gene and the occurrence of HLAR also, the coexistence of one or more AMEs encoding genes is not a prerequisite for the incidence of a HLAR pattern. aph (3ʹ)-VI and aac(6ʹ)-Ib were the most prevalent AMEs encoding genes that circulate in the hospital environment, while the armA gene was the only detected 16S rRNA methylases encoding genes among clonally related nosocomially transmitted A. baumannii clinical isolates fingerprinted by RAPD-PCR. This is the first study that addresses the detection of a novel allelic variant that we named aac(6ʹ)-SL and three novel mutations in the sul1 gene.