Molecular analysis and antimicrobial resistance pattern of distinct strains of Pseudomonas aeruginosa isolated from cystic fibrosis patients in Iran.

Background and Objectives: Colonization of Pseudomonas aeruginosa in Cystic Fibrosis (CF) patients may lead to severe pulmonary disease and death. Different characteristics of P. aeruginosa from these patients were determined in the present study. Materials and Methods: Antimicrobial susceptibility and AmpC-overproduction were determined. The β-lactamase genes were detected by PCR and the oprD gene was sequenced in some of the carbapenem resistance isolates. Distribution of exo genes was determined by PCR. Cytotoxicity of Exo effector proteins was measured using A549 cells. Biofilm production was determined by microtiter plate assay. Random amplified polymorphic DNA (RAPD) –PCR was performed for molecular analysis. Results: Polymyxin B, piperacillin/tazobactam and meropenem were the most active antibiotics and 9.6% of isolates were ampC overproducers. The prevalence of blaVEB, blaOXA, blaVIM, and blaPER genes were as follow: 22.7%, 3.75%, 6.25% and 3.75%, respectively. A high proportion (83.5%) of isolates was able to produce biofilm. The exoT gene was present in all isolates while exoU was present in about 35% of them. RAPD-PCR revealed 49 patterns among 78 tested isolates in which 34 patterns were detected once. Conclusion: Biofilm formation ability and relatively high frequency of exoS may contribute to the persistence of bacteria within lungs of CF patients. Some characteristics of isolates recovered from a single patient after several sampling procedures were similar, while others lacked resemblance.


INTRODUCTION
Pseudomonas aeruginosa, a major opportunistic pathogen causing a diverse range of human infections, is the most frequently isolated pathogen from sputum of patients with cystic fibrosis (CF) lung http (1). Among multiple pathogenic elements involved in infection process (2), type III secretion system (T3SS) effectors have been well studied and recognized as key for establishing infection (3). Currently, four type III-secreted effector proteins, ExoS, ExoT, ExoU, ExoY, have been recognized in P. aeruginosa (4,5,6). Both ExoT and ExoS show ADP-ribosyltransferase and GTPase activating activities. They result in inhibition of actin polymerization, prevent phagocytosis and cell migration, and promote apoptosis (7). ExoU and ExoY have phospholipase and adenylate cyclase activities, respectively (8,9). Together, these secreted toxins result in disruption of key eukaryotic cellular signaling pathways (10). Furthermore, in clinical studies, the presence of these exoenzymes is associated with a dissatisfactory clinical outcome among patients with P. aeruginosa infection (9). A great deal of evidence has demonstrated that the biofilm formation is responsible, in part, for therapeutic failure and biofilms have been shown to be up to 1000 times more resistant to antibiotics than planktonic, or free-swimming, cells of the same isolate (11). Presence of quorum-sensing molecules in CF sputum, used by P. aeruginosa and other bacteria for cell to cell communication in vitro (12), has been cited as evidence that P. aeruginosa in the CF lung form biofilms (13). On the other hand, due to increased multi-drug resistant isolates and biofilm formation capacity, treatment of infections caused by this bacterium is difficult, so therapeutic options are limited and a careful selection of antibiotic is necessary (14). β-lactam antibiotics are used for the treatment of acute pulmonary exacerbations in CF patients (15). Resistance mechanisms of P. aeruginosa to β-lactam antibiotics are multifactorial, including production of β-lactamase, AmpC overproduction, loss of OprD porin and overexpression of efflux-pumps (16). The distribution of T3SS effector genes amongst clinical isolates of P. aeruginosa remains to be elucidated. Furthermore, the prevalence of effector genes in populations of isolates from different disease sites has not been systematically and thoroughly explored. RAPD analysis were also determined.

MATERIALS AND METHODS
Bacterial isolates. A total of 85 clinical isolates of P. aeruginosa were included in the present study. These isolates were collected from 27 CF patients attending the Children's Medical Center Hospital in Tehran, among which, six patients took part in collection process more than once. The collection process was carried out through April 2010 to June 2011. Primary identification to the species level were done based on colony morphology, pigment production, and Gram staining. All phenotypically different isolates were collected to be observed, though some of them were obtained from one patient. Routine biochemical tests used for confirmation were as follow: the ability of oxidase and catalase production, oxidative-fermentative (OF) test, and growth at 42ºC on Mueller-Hinton agar (17).
Detection of bla genes among β-lactam resistant isolates. Genomic DNAs used as template were extracted using DNA Extraction Kit (Accu-Prep® Genomic DNA Extraction Kit, USA Bioneer, Co.) according to the manufacturer's instructions. Hence, in this report, we investigate the distribution The bla PER , bla , bla , bla PSE , bla CTX-M , bla OXA of the type III effector proteins-encoding genes and bla OXA-48 , bla VIM, bla , bla GIM , bla , bla SPM , bla NDM association between cytotoxicity and exo genes in bla SIM , bla KPC and bla GES genes were detected by P. aeruginosa strains obtained from cystic fibrosis patients. Other characteristics of isolates such as antibiotic resistant pattern as well as mechanisms of resistance to β-lactam antibiotics, biofilm capacity, and PCR. Amplification of β-lactamases genes was performed using specific primers (16,20). The amplified PCR products were subjected to electrophoresis in 1.2% agarose gels, stained with ethidium bromide, IRAN Detection of AmpC overproducer among carbapenem resistant isolates. AmpC overproduction among carbapenem resistant isolates was assayed according to Rodríguez-Martínez et al. method (21). Briefly, the isolates were considered as AmpC overproducers when there was at least a two-fold dilution difference between the MICs (minimal inhibitory concentrations) of IMI (imipenem), MEM meropenem) or CAZ (ceftazidime) and MICs of these antibiotics plus 250 µg COL (cloxacillin).
Sequencing of oprD gene among carbapenem resistant isolates. PCR amplification of oprD among carbapenem resistant isolates was performed using specific primers (21). Sequences were compared with that of reference P. aeruginosa PAO1.
Detection of genes encoding type-III secretion toxins. Sequences of specific primers used in amplifying genes encoding type-III secretion toxins, exoT, exoU, exoS, exoY, based on PCR technique as previously described (22,23).

Cytotoxicity assay.
To determine cytotoxic effect of P. aeruginosa isolates on A549 human lung epithelial cells, isolated bacteria were grown in Luria-Bertani broth. Cultures were centrifuged and resultant pellets were washed three times with sterile PBS. The final bacterial cell pellets were resuspended in PBS to yield the final concentration of 1.5 × 10 8 cfu/ml. Human cells suspension were seeded on 96-well flat-bottom tissue culture plates at a density of 1.5 × 10 5 cells / well with DMEM (Dulbecco's Modified Eagle's medium) and 10% Fetal Bovine Serum (FBS). After an overnight incubation of cell culture at 37°C, wells were rinsed with sterile PBS, followed by the addition of DMEM without phenol red and 1% FBS. Consequently, the diluted bacterial suspension of each P. aeruginoasa isolates were added to the cells and incubated at 37°C for 4 h. PBS without inoculated bacteria was used for negative controls. Eventually, lactate dehydrogenase activity was assayed for quantification of cytotoxicity using the Cytotoxicity Detection Kit plus (Roche, Germany) according to the manufacturer's instructions. The assay was performed in triplicate.
Biofilm formation. Biofilm formation ability of isolates was determined based on a colorimetric microtiter plates assay as described by Lee et al. (24). In brief, an overnight bacterial culture in Tripticase Soy Broth (TSB) at 37°C was diluted (1:100) by fresh TSB and then inoculated into the sterile flat-bottomed 96-well polystyrene microtiter plates. After an incubation period of 24h at 37°C without shaking, wells were gently washed three times with PBS. 99% methanol was used for biofilm fixation for 15 min, and then the plate was air-dried. Afterwards, Crystal violet 1% (CV) was added for biofilm staining for 20 min. In the next step, unbound CV was removed and the remaining was released by adding 33% acetic acid. Finally, the absorbance of solubilized CV was measured at 590nm using a microtiter plate reader. All the assays were tested in triplicate. Uninoculated medium was considered as control negative sample. The cut-off OD (ODc) was defined as three standard deviations above the mean OD of the negative control. In accordance with the results of microtiter plate tests, Biofilm producers were characterized as follows based on the optical density: non-biofilm producers (OD test <ODc), weak biofilm producers (ODc< OD < 2× ODc), moderate biofilm producers (2× ODc< OD < 4× ODc), and strong biofilm producers (4× ODc< OD).

PCR-RAPD typing of isolates.
The RAPD-PCR technique was used as described by Bulkanov et al. (25). This fingerprinting was carried out by using arbitrary primers, RAPD-208 (5´-AGCGG-GCCAA-3´) and 272 (5´-ACGGCCGACC-3´) (16,26,27). This amplification was performed in T-100 Gradient (BioRAD) using PCR Master kit (AMPLI-QON, Denmark) according to manufacture guideline, with the following cycling parameters: initial denaturation at 95°C for 5 min followed by 45 cycles of denaturation at 95°C for 1 min, annealing for 1 min at 38.5°C, extension at 72°C for 5 min. The final extension step was continued for another 2 min at 72°C. After visualization of amplified products, genotypic profile of each isolate was determined. BioNumerics software was used in pattern analysis and isolates with ≥85% similarities were considered as indistinguishable (one pattern) types and isolates with <85% similarities were taken as different types. RAPD types were designated numerically. Type-1 was the most frequent pattern among 49 detected patterns. IRAN Table 1 shows the antibiotic resistance patterns of P. aeruginosa isolates. As shown in Table 1, polymyxin B followed by piperacillin/tazobactam and meropenem were the most active antibiotic against the panel of isolates. Ceftazidime was effective on about 50% of the isolates. Less than half of the isolates were determined as MDR organisms. and showed decrease in MICs for ceftazidime, imipenem and meropenem in presence of cloxacillin (Table 2).
Different characteristics of P. aeruginosa isolates including type III secretion-toxin encoding gene tations which leads to internal stop codon and frameshift mutations and in two isolates oprD was not detected by PCR (  Table 3. A high proportion of isolates were able to produce biofilm (83.5%) in which, 20%, 37.64% and 25.88% were determined as weak, moderate, and strong biofilm producers, respectively.
As shown in Fig 1a, exoT was present in all isolates. Most isolates harbored exoS and exoY, while exoU was present in about 35% of isolates. In vitro cytotoxicity against A549 human lung epithelial cells was observed in strains carrying type III secretory genes. The correlation between the presence of exo genes and cytotoxicity is demonstrated in Fig.  1b. It revealed that levels of cytotoxicity were found to be significantly affected by presence of exoU gene.
Results based on RAPD-PCR, as a quick method to characterize the heterogeneity in the group by using a cut off value of 85% as the threshold, revealed 49 patterns among 78 tested isolates in which 34 patterns were detected once, while the remaining 15 patterns were repeatedly observed. RAPD-PCR generated dendrogram of 78 P. aeruginosa isolates from 27 different patients is illustrated in Fig. 2.  Properties of the isolates obtained from some of the patients who have been taking part in sampling for several times, during a year, are demonstrated in Table 4.

DISCUSSION
As antibiotic resistant pattern shows, less than half of the present isolates were resistant to all tested antibiotics, except cefatazidime. The difference in antibiotic resistant pattern between the obtained results in the present and other studies (28) may be due to either the different common colones present in a certain region or different antibiotic treatment pattern. The frequency of MDR isolates in the present study was considerably less than the previous investigation done by our team on isolates from wound infections of burn patients (about 92%) (23). Results obtained by Rao et al. (28) even showed less frequency of MDR isolates from CF patients.
As shown in Table 2, AmpC overproduction and oprD mutations were associated with resistance to imipenem and meropenem. Furthermore, AmpC overproduction was associated with resistance to ceftazidime. In the present investigation, results revealed that the association between OprD gene deficiency and resistance to imipenem and meropenem is similar to other literatures. Mutations in oprD gene reportedly lead to increased MIC to imipenem (16).
To date, presence of different β-lactamases genes (TEM, SHV, CTXM, PER, VEB, GES and IBC families) were reported among P. aeruginosa isolates (16,20). In the present study, VEB type of β-lactamases was the most predominant ESBLs. The results of this study demonstrated bla VIM as the most common MBLs (Metallo-β-lactamases) found in carbapenem-resistant bacteria, including carbapenem-resistant P. aeruginosa in non CF patients (16,20).
It was shown that type-III secretion system is an important virulence mechanism and cytotoxins encoded by exoS, exoT, exoU and exoY genes may be important contributors to the dissemination of the organism from the site of infection, bacterial evasion of the host immune response and inhibition of DNA synthesis leading to host cell death (29). The majority of P. aeruginosa strains carry exoT and exoY genes; however, the presence of exoS and exoU differ noticeably between the isolates and appear to be mutually exclusive (29). As the previous study done IRAN    by our team on P. aeruginosa isolated from burn patients (23), exoT was the most prevalent cytotoxin gene found in isolates of CF patient and was reported in all isolates (100%), though exoY was not as frequent as exoT (about 79%). In contrast to our previous results that showed exoS as the least prevalent gene (29%) (23), present study revealed that exoU has less frequency than other cytotoxin genes (36.58%). Data showed that the least prevalence of exoU in CF patients, as a chronic disease, is in agreement with results reported by Feltman et al. (22), which may be due to the straight correlation between exoU and acute infections. So, the high prevalence of exoU gene in burn isolates, as was reported before is predictable. Relatively high frequency of isolates harboring exoS (about 80%) may be related to hypothesis presented by Feltman et al. (22). They suggested that production of exoS may provide P. aeruginosa isolates with an advantage in colonizing or persisting in CF lung. Furthermore, exoS may be genetically linked to a second factor that is important in the pathogenesis of P. aeruginosa pulmonary colonization and infection in CF patients. Taken together, our findings suggest that at least some of the genes encoding type III secretion effector proteins are present in all P. aeruginosa isolates, though the prevalence of them was different. Variation observed in the prevalence of exo genes may be the consequence of clinical specimen used for collecting isolates.
In the present study, results related to in vitro cytotoxic effect of strains carrying type III secretion proteins concur with the previous studies, in which the type III secretion system is reported to be the primary contributor to the virulence of P. aeruginosa and may lead to the increased morbidity and mortality in patients and animals (23).
High frequency of biofilm producers in the present study (more than 90%), as in other report, suggests that P. aeruginosa is able to form biofilms within the lungs of CF patients and this mode of growth likely facilitates persistence of the bacterium within this niche (30) and biofilm development in CF airway is one of the most striking features of P. aeruginosa adaptation to airways of patients with CF (31). However, CF isolates can be highly variable with respect to biofilm density. In the current study, the prevalence rate of strong biofilm producers was about 25%, whereas Perez et al. (32) could not find such a biofilm state among CF patients. No correlation was observed between biofilm formation and cytotoxicity.
Because of the presence of problems related to phenotypic variations of bacterial isolates throughout the period of infection in CF patients (33) and the difficulty of ascertaining the reinfection of patients or difficulty in determining whether the patient is infected by different strains over a period of time, or whether the isolated strains differ according to the respiratory tract site where the sample was taken (34), molecular typing was selected as a more sensitive and specific method in this study. RAPD typing of the 78 P. aeruginosa isolates gave 49 different RAPD fingerprints, 34 of which contained only 1 strain. The dominant RAPD type amongst the remaining 15 fingerprints contained 9 strains. Presence of most isolates harboring unique RAPD patterns may be due to the acquisition of these isolates from environmental sources where horizontal gene transfer occurs and may result in acquisition of new genes.
As shown in Table 4, our results revealed that more than one P. aeruginosa strains can be isolated from a CF patient in one time attendance for sampling, though a given strain was not recovered from a given patient after a period of time. In other words, isolation of different strains with different properties from one patient while attending for several times revealed the possibility of both phenotypic and genotypic alterations in causative strains. Furthermore, the similarity between some of the genotypic aspects of different strains, such as the same RAPD pattern in isolates 1 and 2 in patient 20 (P20) was observed, though another genotypic pattern related to exo genes was different. On the other hand, despite the presence of analogous exo genes and RAPD patterns in isolates P1-5 and P1-10, which were recovered from the same patient, their cytotoxic capability were drastically different. Moreover, recovery of an isolate with a similar RAPD pattern, 7, from patient 5, was observed in each round of sampling, though isolation of another different pattern, 51, from the mentioned patient, was restricted to a defined round of sampling. Interestingly, cytotoxicity of isolates with pattern 7 from patient 5, which were recovered constantly, was low and the last isolated strain with this pattern has a strong biofilm intensity. Failure of the treatment in CF patients may be due to the strong ability of causative P. aeruginosa to form biofilm. On the other hand, isolated strain with pattern 51, recovered once, from patient 5 was significantly more cytotoxic than other isolates with pattern 7. However, treatment and management of patient 5 was not ac-ceptable and victorious. Such findings revealed that isolation of stable strains present in the whole duration of the disease along with new strains showing new different RAPD patterns in CF patients may be due to the acquisition of the latter strains from hospitals where they took part in sampling. Undoubtedly, antibiotic therapy against such new and transient strains may result in its elimination from patients, while stable strains with strong ability to form biofilm do not respond to such treatment and can be isolated for several times. However, it is well documented that in the initial stages of infection, when the bacteria are mainly planktonic and not yet persistent, their eradication may be easier and so it reveals the importance of early identification of infection (35).

CONCLUSION
To our knowledge, the compelling role of biofilm formation capacity and cytotoxicity effect of isolates in bacterial invasion process validate probable and powerful therapeutic strategies other than conventional and traditional antibiotic treatment. Furthermore, relatively high frequency of antibiotic resistance among isolated P. aeruginosa may persuade the pharmaceutical industry to introduce new antibiotics to solve the resistance problem.