Introduction

Hospital-acquired infection due to Acinetobacter baumannii is a rising problem globally, including the Gulf region and Saudi Arabia [1–3]. Acinetobacter baumannii is a glucose non-fermentative Gram-negative bacilli which is widespread in hospitals and healthcare environments [4]. Acinetobacter baumannii plays an important role in hospital-acquired infections due to their extended survival time periods on surfaces and their high tendency to acquire resistance to multiple drugs. The resistance mechanism that A. baumannii uses is becoming a challenge for both physicians and microbiologists. One possible explanation of such a mechanism involves the selective permeability action in the outer membrane protein (porin) which controls the efflux system and the presence of chromosomal beta-lactamase [5, 6].

PER is a term that refers to Pseudomonas aeruginosa extensive resistance, which is carried onto both chromosomes and plasmids [7–9]. Although the gene was first discovered in a P. aeruginosa strain, different hypotheses argue about its origin [8, 10]. The PER-1 gene has been detected in A. baumannii strains isolated from intensive care unit (ICU) patients in Riyadh, Saudi Arabia [3]. This might have triggered an outbreak case that was first found in the country. It is an extended-spectrum beta-lactamase enzyme and encoded by the bla -PER-1 gene. This enzyme is known for its ability to hydrolyze the beta-lactam ring of antibiotics, such as penicillin, oxyimino-cephalosporins, and aztreonam [11]. PER resistance genes have been identified worldwide, including in Europe, North and South America, and the Far East, due to their chromosomal location [12]. PER-1 gene expression was first detected in 1993 in a Turkish patient in France who was infected with P. aeruginosa [13]. Different P. aeruginosa isolates harboring the enzyme were then reported in Turkey, and a widespread detection of the PER-1 enzymes has been reported in other European countries [14–16]. On the other hand, the prevalence and genetic diversity of PER-1 in the Middle East in general and Saudi Arabia in particular, are yet to be identified [2].

Many molecular typing methods have been used to study the molecular epidemiology and genetic diversity among A. baumannii isolates. Multilocus sequence typing (MLST) is based on sequencing of the seven impartially selected genes, and the genes of each are allocated a number for each different allelic form [15, 17]. Based on the sequence types (STs) of isolates (A. baumannii isolates for example), the relatedness and phylogenetic analysis of these isolates will be determined by comparing their STs with the databases created by different laboratories worldwide (http://pasteur.fr/mlst). The purpose of the current study was to investigate the prevalence of PER-1 resistance gene among carbapenem-resistant A. baumannii isolates from Riyadh, Saudi Arabia.

Materials and methods

Bacterial identification

The clinical isolates used in this study were collected at the microbiology laboratory of the 800-bed tertiary care hospital, King Abdulaziz Medical City (KAMC), Riyadh, Saudi Arabia during 2006–2014. Isolates characterized by aerobic growth and were catalase-positive, oxidase-negative were identified using commercial the biochemical system MicroScan® WalkAway (Siemens, Germany). The isolates were initially identified at the species level using MicroScan and were then confirmed by polymerase chain reaction (PCR) of the RNA polymerase β-subunit gene (rpoB), followed by sequencing.

MIC and susceptibility testing

We collected 503 isolates for antimicrobial susceptibility testing. We used a randomly selected subset (n = 130) for the MLST to further study their minimum inhibitory concentrations (MICs). The susceptibility data were interpreted according to the guidelines of the Clinical and Laboratory Standards Institute (CLSI). It was performed using the automated MicroScan® Walkaway system. All isolates that exhibited resistance were confirmed by the Etest® (bioMérieux, France) to detect the exact MIC value. Isolates were tested for their MIC using the MicroScan Neg MIC panel type 32 including the following antibiotics: amikacin, amoxicillin/clavulanate, aztreonam, cefepime, cefotaxime, ceftazidime, ceftriaxone, ciprofloxacin, colistin, gentamicin, imipenem, meropenem, tobramycin, and trimethoprim/sulfamethoxazole. Standardized suspensions of tested isolates were inoculated and incubated at 35 °C for a minimum of 16 h and the MIC was expressed as susceptible, intermediate, and resistant according to the 2011 European Committee on Antimicrobial Susceptibility Testing (EUCAST) report. Only one isolate per patient per year was included in the current analysis. Quality control was performed by testing these same antimicrobials against Escherichia coli ATCC 25922, E. coli ATCC 35218, and P. aeruginosa ATCC 27853. The isolates were defined as multidrug-resistant (MDR) [4, 18].

DNA extraction

DNA was extracted from 100 μL of fresh subcultured samples using the DNA extraction kit MagNA Pure Compact (Roche), according to the manufacturer’s instructions. In brief, we used a combination of proteinase K enzyme and heat lysis, followed by magnetic bead capture and DNA wash. Finally, DNA was eluted with 200 μL of 10 mM Tris and 1 mM EDTA (pH 8.0) of elution buffer. The aliquots of extracted nucleic acids were stored at −20 °C until further analysis. The quantity and purity of the extracted nucleic acid samples were confirmed using a NanoDrop spectrophotometer (ND-1000, Fisher).

Detection of resistance genes

To identify the resistance genes, genotyping was performed by PCR amplification reactions with a list of primers representing the four different resistance mechanisms. Twenty pairs of primers were used for A. baumannii resistance genotyping (Eurofins, Germany) including TEM, VIM, and PER-1, as previously well established by Bonomo and others and explained elsewhere [3]. Final PCR reactions were carried out in 20-μL volumes containing 20 ng DNA, 17 μL of MegaMix-Blue (Microzone, UK) containing 1.1× PCR reaction buffer, 220 μM of dNTP, 2.75 mM MgCl2, 0.25 units of recombinant Taq DNA polymerase, and 0.40 μM each of the forward and reverse primers. PCRs were carried out using the PCR thermocycling program as recommended by the PCR mix supplier. In general, PCRs were done as follows: after the initial 10 min of incubation at 95 °C, a 30 cycle program comprising 94 °C for 30 s, 55 °C annealing temperature for 60 s, 72 °C for 60 s, and a final extension for 10 min at 72 °C was used. Some modifications in annealing temperature, number of cycles, etc. were made to optimize the amplification for certain genes. The PCR products were analyzed by electrophoresis on a 1.5 % (wt/vol) agarose gel containing 0.125 μg/mL ethidium bromide.

Sequence-based typing (SBT) and sequencing

The PCR amplicons were purified using the QIAquick PCR Extraction Kit (Qiagen, USA). Both strands of the amplicons were sequenced by using an ABI Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit v1.1 (Applera, France) and an ABI 3100 automated sequencer (Biosystems, USA). PCR sequences were compared with all nucleotides sequences in the GenBank database using the PubMed National Center for Biotechnology Information BLAST program.

Multilocus sequence typing

MLST of 100 randomly selected isolates representing all clusters form the SBT-based phylogenies were carried out using the following seven housekeeping genes: cpn60 (60-KDa chaperonin), fusA (elongation factor EF-G), gltA (citrate synthase), pyrG (CTP synthase), recA (homologous recombination factor), rplB (50S ribosomal protein L2), and, finally, rpoB (RNA polymerase subunit B), as described by Diancourt et al. [19]. The primers used for MLST amplification and sequencing are listed in Table 1 and the PCR conditions used were as previously described by others [19, 20]. We used the Pasteur MLST database for allelic profiles identification (http://pubmlst.org/perl/bigsdb/bigsdb.pl?db=pubmlst_abaumannii_pasteur_seqdef).

Table 1 Scheme of Acinetobacter multilocus sequence typing (MLST) amplification and sequencing primers developed by Diancourt et al. [19]
Full size table

Data analysis

Data were collected and analyzed by Excel and SPSS version 20. A phylogenetic tree was generated from the MLST allelic profile using the START 2 software using default criteria and the unweighted pair group method with arithmetic mean [21]. The MLST genotyping alleles were used to construct the tree. Moreover, the identified sequence type, patients’ demographic data, and their corresponding TEM and PER-1 genes were incorporated with the lineage tree. The isolates with new sequence types are shown with red arrows.

Results

MDR and carbapenem-resistant phenotypes

A total of 503 isolates were collected from KAMC during the period 2006–2014. The sources of the clinical isolates were as follows: respiratory 60 (46.2 %), wound and burn 31 (23.8 %), blood 18 (13.8 %), urine 5 (3.8 %), and others 16 (12.3 %). The demographic data showed that the isolates were collected from patients with the following age distribution: 2 to 24 years with 22.2 %, followed by aged 25 to 49 years (28.6 %), and, finally, 49.2 % of the patents were elderly, between 50 to 90 years. furthermore, we did not observe any significant differences in antibiotic resistance among A. baumannii isolated from males (49.2 %) versus females (50.8 %). Susceptibility testing showed that all isolates were resistant to four or more antimicrobial groups. Antibiotic susceptibility isolates against imipenem and meropenem are represented in Tables 2.

Table 2 Antibiotic susceptibility of multidrug-resistant (MDR) Acinetobacter isolates represented as number of isolates (percentage)
Full size table

PER-1 and other carbapenems resistance genes

Carbapenems resistance phenotypes among the isolates were confirmed by MIC ≥8 μg/ml for both meropenem (96.4 %) and imipenem (88.3 %). PCR typing showed that 76.3 % of the isolates (n = 384/503) contained the PER-1 gene. On the other hand, the genotyping results showed that n = 132/503 (26.2 %) isolates were positive to bla -TEM and 457/503 (90.8 %) were positive for the OXA-51-like gene. Nevertheless, all isolates were negative to the OXA-24-like gene.

Sequence types of A. baumannii isolates and identification of regional clones

The sequence types for 100 randomly selected isolates typed by MLST were gathered in four clusters. The MLSTs of some isolates (n = 34) were not available because they were untypable or missing alleles. However, the most common sequence type (ST) was ST2 (n = 26), which is part of clonal complex 2 (CC2) and represents the European clone II (EUII) or Worldwide clone 2 (WW2). The second most common type was ST20 (CC1) (n = 19), followed by other STs. Moreover, MLST enabled the identification of four new isolates (486–489) with novel allelic profiles (Table 3). The MLST allelic profiles of the four novel STs isolates have been entered into the Institut Pasteur MLST databases. Full data of the new isolates ST-486, ST-487, ST-488, and ST-489 can be accessed at (http://pubmlst.org/perl/bigsdb/bigsdb.pl?page=info&db=pubmlst_abaumannii_isolates&id=1602).

Table 3 For selected isolates, demographic data and sequence type alleles and beta-lactamases and metallo-beta-lactamases resistance genotyping among MDR Acinetobacter isolates (full dataset in Supplementary Table 1)
Full size table

Diversity of PER among Acinetobacter isolates

The PCR primers used for sequencing bla- PER-1 genes in A. baumannii identified two genotypic forms, PER-1-1 and PER-1-7. In addition, based on hit definition, the PER-1-positive clones were sub-divided into five groups (Table 4). The majority of PER-1 isolates are similar to isolate AB11 (bla-PER-1–1) (Table 4) and the remaining four groups represents one isolate of PER-1-1 in each, as shown in Table 4. On the other hand, the sequences of 29 isolates were submitted to the BLAST server and found to be similar to PER-7 (Table 4). This group was sub-divided into two groups: 17 isolates belong to isolate AB16 and the remaining 12 belong to pAB154.

Table 4 Results of BLAST of the isolated clones
Full size table

Association of PER-1 diversity and MLST

A lineage tree was generated from MLST allelic profiles and SBT-PER using the unweighted pair group method with arithmetic mean (Fig. 1). MDR A. baumannii isolates were grouped into three clusters, where the first cluster consisted of isolates that harbored the PER-1 gene (n = 28) and belonged to ST2. Distinctively, many isolates typed as ST20 and possessing the PER-1 gene (n = 19) were the main contributors to the second cluster. However, the third cluster was a mix of STs that confer positive TEM and PER resistance genes together (Fig. 1).

Fig. 1
figure 1

Lineage tree generated from the multilocus sequence typing (MLST) allelic profile and showing the MLST a sequence type, patients’ demographic data, and their corresponding TEM and PER-1 gene occurrence in Acinetobacter clinical isolates from Saudi Arabia, with new sequence type isolates shown with bold font

Full size image

Discussion

Genetically, PER-1 assumed to be part of the class A beta-lactamase enzyme but different studies contradict the idea by separating it into independent groups within classes [7–9]. This is due to the fact that the PER-1 gene shares only 26 % sequence homology with other class A beta enzymes such as TEM and SHV. Also, there are two essential conserved domains in the class A beta-lactamase structure: the omega loop fold and the cis confirmation peptide bond between 166 and 167 residues were not present in the PER-1 protein [22]. On the other hand, PER-1 was found to form a new fold in the omega loop which possessed an aspartic acid residue 138, conferring trans confirmation between 166 and 167 residues [23]. Recently, A. baumannii isolates with antibiotic resistance have been on the increase in the Gulf countries [22–26]. The characterization of its prevalence will help contain this problem. Studies dissecting the genetic structure and mode of acquisition of certain specific elements revealed that two transposons elements, ISPa12 in the upstream and ISPa13 in the downstream, surround the PER-1 gene to form the TN1213 transposon in both A. baumannii and P. aeruginosa. Also, there is a promoter in the ISPa12 element that activates the expression of the PER-1 gene [23]. In this study, 73 isolates were characterized by PCR and sequencing of the bla PER gene. The sequenced products were then analyzed for homology with equivalent sequences available at NCBI BLASTN (http://blast.ncbi.nlm.nih.gov/Blast.cgi). All of the isolates were found to be positive for this gene. Among the 73 isolates, 55 % of samples were related to the Chinese origin isolate AB11 (bla PER-1), 23 % related to the Chinese origin isolate AB16 (bla PER-1), 16 % related to the Egyptian origin plasmid pAB154 (blaPER-7), and 6 % of the isolates were related to Chinese, French, and British origin of A. baumannii. Wang et al. [27] reported a similar high prevalence of PER-1 A. baumannii in China as well. In their study, about 78 % of the Chinese imipenem-resistant Acinetobacter isolates were found to harbor the PER-1 gene. These results indicated that the prevalence of PER-1 genes is similar to that China (77.8 %) and higher than in Italy (34.61 %) and Turkey (46 %) [13, 27, 28]. Thus, the isolated strains in Riyadh, Saudi Arabia are either transferred by nationals of these countries or by Saudis who visited these countries. It is noteworthy that PER-7 represents a slight challenge, as it can hydrolyze cephalosporins more rapidly than PER-1 [29].

Conclusion

Here, we report for the first time the high prevalence of PER-1 resistance gene among Acinetobacter baumannii in Saudi Arabia. The PER-1 resistance gene was previously reported in some European countries (Turkey, France, and Italy) and North America. Further studies are needed to explore the origin and molecular mechanisms that confer the resistance phenotype