Molecular characterization of strong biofilm producing MDR, XDR and PDR Pseudomonas aeruginosa isolated from a tertiary care hospital in South India

Background : Pseudomonas aeruginosa is an opportunistic, gram negative bacterium that causes serious infections, especially among immunocompromised patients. An unusually high incidence of nosocomial P. aeruginosa infections was observed among patients between 2014 November and 2015 February. Some of the patients being treated for a variety of cardiac and urological disorders in the Cardiothoracic and Vascular Surgery Intensive Care Unit (CTVS-ICU) and Urology ward were found infected with P. aeruginosa . Active surveillance and environmental sampling revealed the presence of two additional MDR P. aeruginosa in the tap waters of CTVS-ICU. Based on the Antibiotic sensitivity pattern, fourteen P. aeruginosa MDR (n=7), XDR (n=6) and PDR (n=1) isolates (inclusive of the two tap water isolates) were shortlisted for additional investigations. These isolates were characterized for antibiotic sensitivity, biofilm production, minimum biofilm inhibitory concentration (MBIC), presence of antibiotic resistance genes, efflux pumps, and integrons. Results : Mutations in gyr A, gyr B, par C, mex R, nfx B, mex B and mex F genes were correlated to enhanced antibiotic resistance. Notably, the isolates were also found to harbor integrons and blaNDM-1. Pulsed Field Gel Electrophoresis (PFGE) and Random Amplified Polymorphic DNA (RAPD) based phylogenetic analyses grouped the clinical isolates into three distinct ward specific clusters, while the tap water isolates were grouped into a separate cluster. Detection of nosocomial P. aeruginosa in the CTVS-ICU and urology ward triggered the activation of enhanced surveillance and infection control measures to contain and eliminate P. aeruginosa infections. Conclusions : Multiple clones of P. aeruginosa are prevalent in the study center. Hence, continuous screening and identification of potential reservoirs is absolutely essential to control the spread of drug resistant P. aeruginosa infections.


Background
Nosocomial infections by Pseudomonas aeruginosa have become a critical clinical challenge globally, and are associated with significant morbidity and mortality particularly in developing countries (1,2). The organism is a ubiquitous, opportunistic human pathogen that causes bacteremia, ventilator-associated pneumonia (VAP), abdominal, genito-urinary, skin, post-operative infections and sepsis (3,4). P. aeruginosa is the most frequently encountered pathogen among Intensive Care Units (ICUs) (5) and is included among the list of "ESKAPE" pathogens (6) and "Meet WHO's dirty dozen" by Infectious Diseases Society of America (IDSA) and World Health Organization (WHO) respectively. P. aeruginosa infections account for more than 19.8% of all ICU infections globally. In India, the situation is even more stark (13.2 to 38%) (7). The bacterium possesses several intrinsic resistance mechanisms which include active efflux pumps, low outer membrane permeability, production of antibiotic degrading enzymes and biofilm forming capacity (8).
Recalcitrance due to robust biofilm formation by P. aeruginosa is believed to be a major contributing factor to its colonization and persistence on many biotic and abiotic surfaces including water bodies and pipes. These biofilms are highly tolerant against traditional antibiotics (9). Mutations in DNA gyrase (gyrA and gyrB) and topoisomerase IV (parC and parE) genes confer resistance to fluoroquinolones (10,11). Further, presence of multiple efflux pump-associated regulatory genes (mexR, mexZ and nfxB) have been reported to confer resistance to tetracycline, chloramphenicol and fluoroquinolones albeit at a lower level (12)(13)(14). With the capacity to acquire additional resistance determinants through Horizontal Gene Transfer (HGT) (15)(16)(17), emergence of nosocomial extensively drug resistant (XDR) and pan drug resistant (PDR) P. aeruginosa has now become a new normal.
Integrons, transposons, and conjugative plasmids are the most frequently associated mobile genetic elements (MGEs) among antimicrobial resistant (AMR) Pseudomonas infections. Class I integrons, prevalent among resistant isolates are generally found to possess gene cargo conferring resistance to b-lactams, aminoglycosides, macrolides and sulfonamides antibiotics (18,19). Scant information is available on the molecular analyses of nosocomial XDR and PDR P. aeruginosa infections in the Indian context.
In this study, we characterized drug-resistant P. aeruginosa (14 nos)

Clinical isolation
Due to the unusually high incidence of P. aeruginosa infections (50 nos) among the patients of the CTVS-ICU and urology wards at Sri Sathya Sai Institute of Higher Medical Sciences -Prasanthigram (SSSIHMS-PG), Puttaparthi, India; enhanced surveillance in and around the hospital was initiated to identify and eliminate any potential sources of infections. Bedside rails, dressing trolleys, blood pressure monitoring machines and other biomedical equipment located in the CTVS-ICU and Urology ward were sampled. Hand swabs from nurses and attending physicians who came in contact with these patients were also collected for testing. Further, samples of hand sanitizers, hand washing solutions, and tap water from these units were investigated. All P. aeruginosa isolates were analyzed for their antibiotic resistance profiles. Fourteen drug resistant isolates (MDR, XDR and PDR) which include the two MDR CTVS -ICU tap water isolates were characterized further.
The remaining isolates (36 nos) which were found to be sensitive to the tested antibiotics were excluded from the study.
Bacterial identification and antibiotic susceptibility testing (AST) Initial identification of clinical and environmental isolates was performed by Vitek-2 (BioMérieux) as per manufacturer's instructions. Further, identity of these isolates was confirmed by 16s rDNA sequencing (20). Antimicrobial susceptibility testing (AST) was carried out using Kirby-Bauer disk diffusion method (Hi-media) for all the 50 P. aeruginosa isolates. Reconfirmation of the resistance profile for the fourteen drug resistant P. aeruginosa isolates was done by Vitek-2 (AST-N281 cards) and the isolates were RAPD was performed to understand the genetic diversity among the isolates. Genomic DNA was isolated from overnight cultures using NucleoSpin® tissue kits (Macherey-Nagel).
Genomic DNA from each isolate was subjected to RAPD-PCR (Sapphire Amp Fast PCR Master Mix) using the primer Kt5: 5'-AGCAAGCCGG-3'. The PCR conditions used for the amplification were initial denaturation at 94˚C -4 min followed by 40 cycles of 94˚C -1 min, 35°C -1min, 72°C -2min and a final extension of 72°C -5min. A phylogenetic tree was constructed from the RAPD finger print data using Neighbor Joining method by NTSYS software. Inter node branches with bootstrap value more than 50% were inferred as distinct clusters in the tree.

Pulsed-Field Gel Electrophoresis (PFGE)
Genomic DNA from the 14 P. aeruginosa isolates was fingerprinted by PFGE following SpeI restriction enzyme digestion. SpeI digested genomic DNA fragments were separated in two step runs; Step1: 16h at 6V/cm with initial and final pulse times of 0.5 sec and 25 secs respectively; Step2: 2h at 6V/cm with initial and final pulse times of 30 secs and 60 secs respectively using the CHEF-DR II system (Bio-Rad, Melville, NY, USA). DNA from Salmonella serotype Braenderup (H9812) digested with XbaI was used as a DNA size marker. The gel was photographed and profiles were visualized using the GelDoc® (Bio-Rad, Hercules, CA, USA) photo documentation system. Banding patterns were analyzed using BioNumerics Version 7.6 (Applied Maths, Austin, TX, USA). The resulting dendrograms were generated at 2.0% position tolerance and 1.5% optimization using the unweighted pair-group method with arithmetic mean (UPGMA) and the Sørensen-Dice similarity coefficient.

Biofilm formation and MBIC analysis
Biofilm production was determined using the method of O'Toole GA (22,23) with minor modifications. Briefly, 250 µl of P. aeruginosa cultures grown in Trypticase Soy Broth (TSB) (OD650 = 0.1) were seeded in a 96 well microtitre plate (Nunc TSP System). Plates were covered with peg lids (TSP lids; Nunc International) and incubated without shaking at 37ºC for 24 hrs. Biofilms that formed on the peg lids were stained using 200µl of crystal violet (0.1% w/v). Post washing, biofilms attached to the pegs were immersed in 250 µl of 70% ethanol for 10 min to solubilize the crystal violet. Absorbance was recorded at 650 nm using SpectraMaxM5 multi-mode microplate reader. The mean readings of four independent measurements of each strain were calculated (24). To identify the MBIC value (23) for each of the tested antibiotics, peg lids with biofilms were transferred into a 96well plate containing 100μL of antibiotics (2 to 1024μg/ml) and incubated at 37°C for 24hrs without shaking. Based on the AST patterns of the isolates, antibiotics for MBIC analysis were chosen. Peg lids with biofilms were then transferred to a new sterile 96-well plate, containing 150μl LB broth and centrifuged at 2,000 rpm for 30 min to recover the biofilm. Initial mean absorbance was measured at 600nm. The plates were then reincubated at 37°C for 12hrs and final mean absorbance was recorded (600nm). The difference between mean absorbance values, before and after incubation was calculated.
All experiments were performed in triplicates. P. aeruginosa ATCC 27853 (quality control strain) was used as a negative control.

Clinical characteristics and categorization of nosocomial isolates
Of the 14 drug-resistant study isolates, two were isolated from tap waters of the CTVS-ICU, while the remaining 12 were clinical isolates. Clinical characteristics of patients at the time of the isolation of the nosocomial strains include urological abnormalities (8 nos), Congenital Heart Disease (CHD) (3 nos) and Rheumatic Heart Disease (RHD) (1 no). (Table   1). Clinical and tap water isolates recovered in the study were identified as P. aeruginosa by Vitek-2 and re-confirmed by 16s rDNA sequencing. The study isolates were tested against Penicillins (carbenicillin, piperacillin+tazobactum and ticarcillin+clavulanic acid), cephalosporins (ceftazidime, cefipime and cefoperazone and sulbactum), monobactum (aztreonam), aminoglycosides (amikacin, gentamicin, netilmicin and tobramycin), carbapenems (imipenem, meropenem and doripenem), fluoroquinolones (ciprofloxacin, levofloxacin and ofloxacin), and polymyxins (colistin and polymyxin -B). Further, they were categorized as MDR, XDR and PDR based on the antibiotic sensitivity patterns ( Table   2).
Seven of the 14 isolates were found to be MDR with resistance to more than three categories of the tested antibiotics. Six isolates which exhibited resistance to more than six categories of antibiotics were grouped as XDR. XDR strains were found to be sensitive to colistin. Notably, one isolate (208) was found to be PDR with resistance against all the eight categories of antibiotics tested and an MIC value of >=16 μg/ml against colistin. Of note, the other 13 isolates had MIC values of 0.5 to 2 μg/ml against colistin (Table 2).

Phylogenetic analyses
Phylogenetic analyses using PFGE and RAPD finger print analyses unambiguously revealed the presence of four distinct clusters. Clinical isolates were grouped into three different clusters, while the two tap water isolates (T3, T4) grouped into an independent cluster.  figure 1) analyses, it may be inferred that the clonality of the study isolates is ward and location specific.

Biofilm production and MBIC
In vitro, all the 14 study isolates were found to produce strong biofilms (mean absorbance: >2.224 @OD 600nm) (Figures 2a, 2b and supplementary table 3). In this experiment, P.
aeruginosa ATCC 27853 was used as a negative control as it does not produce biofilms. As expected, the study isolates were found to exhibit higher MBIC values compared to their corresponding MICs (supplementary table 4 468Glu-Asp mutation was found in seven isolates, 504Asp-His mutation in five isolates, 503Val-Leu and 533 Gln-His mutations in two isolates and 490Leu-Val mutation was seen in one isolate (Table 3).
Int1 gene and 5̍ -3 ̍ conserved segment of class 1 integrons were observed among twelve of the analyzed isolates (Table 3). Notably, isolate 208 which was resistant to all the tested antibiotics did not possess class 1 integrons raising the possibility that other mobile genetic elements may have played a role in its emergence into a PDR phenotype.

Mutations in the efflux pump and efflux pump regulatory genes
Mutations in efflux pump genes were identified among some of the study isolates. In the mexB gene, isolate 208 was found to have 576Val-Gly and 578Thr-Asn mutations; isolate 227 had 607Ser-Thr, 611Thr-Ala and 612Val-Glu mutations; while isolate 237 had 615Phe-Leu mutation and isolate 239 possessed 589Gln-Pro mutation. mexF gene analysis revealed that isolate 217 had 792Glu-Ala and 816Lys-Met mutations while T3 and T4 isolates had 843Ala-Thr mutation ( Table 3). Only the tap water isolate T4 was found to possess 93Leu-Val and 131Gln-His mutations in the nfxB gene. 126Val-Glu mutation was observed in the mexR gene among six of the isolates (227, 236, 237, 241, T3 and T4).
Further, isolates 236 and 237 were found to possess an additional mutation (143Pro-Leu) in the mexR gene.

Interventions
Repeated chlorine treatment of water for one week was performed to eliminate the presence of the pathogens in the hospital water sources. Further, disinfection of all potential reservoirs (overhead water tanks, sumps and taps) was attempted and strict infection control measures were implemented. Additionally, plumbing (pipes and faucets) in the CTVS-ICU where P. aeruginosa spp. were isolated from tap water has been totally replaced. A new overhead water tank with hyper-chlorinated (2ppm) water was installed for the purpose of scrubbing in all the ICUs and operation theaters. Enhanced surveillance, awareness programs for the staff and continuous monitoring has been initiated by the infection control team. Follow up environmental surveillance in and around the hospital premises (repeat sampling over a period of three months) did not reveal the presence of P. aeruginosa and the incidence of P. aeruginosa infections subsided. However, regular screening of water samples in the hospital premises has been made mandatory to identify any potential source of infection.

Discussion
In this manuscript, we describe the microbiological and molecular characterization of drug-resistant clinical (n=12) and environmental (n=2) P. aeruginosa spp. from an outbreak at a tertiary care hospital in south India. Antibiotic susceptibility pattern of the isolates from patients from different wards and tap water sources indicated that P.
aeruginosa isolates belonged to multiple genotypes. It was interesting to note that the two environmental P. aeruginosa isolates with similar AST profile clustered together in both PFGE and RAPD analyses and were distinct from the clinical isolates.
Biofilm forming capacity among P. aeruginosa spp. is a major antibiotic resistance factor and contributes to their ability to cause persistent infections. Even antibiotics that are effective against planktonic forms are unable to effectively penetrate biofilm, thus rendering them ineffective (20,21). Although all 14 drug-resistant isolates were found to produce robust biofilms, the MBIC values indicate differential antibiotic response across Fluoroquinolone resistance in many of these isolates can be correlated to the presence of gyrA (83Thr-Iso) and parC (87Ser-leu) mutations. gyrA -parC mutations have been reported to be responsible for higher ciprofloxacin resistance than mutations in either one of the genes among P. aeruginosa isolates (24). To the best of our knowledge, this is the first report describing multiple mutations in the gyrB gene among fluoroquinolone resistant P. aeruginosa spp. It is intriguing to note that a single gyrB (533Gln-His) mutation in isolate 241 together with mexR (126Val-Glu) contributed to fluoroquinolone resistance while mutations in gyrB (503 Val-Leu and 504Glu-His) and mexR (126Val-Glu) genes in T3 isolate did not contribute to robust fluoroquinolone resistance. The presence of class 1 integrons and bla NDM-1 in some of the tested isolates further confounds the situation. Our findings suggest that there may be a correlation in the acquisition of bla NDM-1 gene and gyrB mutations. Consistent with previous reports, all the isolates that had gyrA mutation also possessed parC mutations (25).

Conclusions
Although the environmental isolates of P. aeruginosa recovered in the study could not be

Availability of data and materials
Data sharing is not applicable to this article as no datasets were generated or analysed during the current study

Competing Interests
The authors declare that they have no competing interests.

Funding
This research project received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.