MOLECULAR CHARACTERISATION OF METALLO-BETA LACTAMASE AND OTHER RESISTANCE GENES IN PSEUDOMONAS AERUGINOSA FROM SEVEN TERTIARY HOSPITALS IN SOUTHWESTERN NIGERIA BY RY A OLUWATOYIN BOLA, OLANIRAN R MATRIC No.: 160617 IB B.Sc. Microbiology (Ilorin), M.Sc. PharmaceutiNcal M Licrobiology (Ibadan) DA B A A thesis in the DepartmFen t Iof Pharmaceutical Microbiology SubmitteOd to the Faculty of Pharmacy in partial fulfilYlm ent of the requirements for the degree of T SI VE R I DOCTOR OF PHILOSOPHY UN of the UNIVERSITY OF IBADAN JUNE 2019 ABSTRACT The emergence of resistance to carbapenems, a last resort antibiotic, among Pseudomonas aeruginosa is of great health concern. Detailed studies on the molecular basis of carbapenem resistance in clinical P. aeruginosa isolates are scanty in Nigeria. Therefore, this study was aimed at determining the incidence of Metallo-Beta Lactamase (MBL) and other mechanisms mediating carbapenem resistance, and evaluating clonal spread among carbapenem-resistant P. aeruginosa isolates. Four hundred and forty-seven presumptive P. aeruginosa isolates collected fromY seven tertiary hospitals laboratories in southwestern Nigeria were idenAtifiRed using biochemical tests and amplification of oprI and oprL genes. AnRtibiogram of the isolates and Minimum Inhibitory Concentrations (MIC) were dBetermined by Kirby- Bauer disk diffusion and broth microdilution, respectively.L PhIenotypic detection of carbapenemases was carried out using Modified-HodNge and combined disc tests. Carbapenem-resistant P. aeruginosa isolates were screened for class A, B and D carbapenemases, integrons and type III secreDtionA effectors by Polymerase Chain Reaction (PCR) followed by sequencingA of amplified carbapenemase genes. Transferability of MBL genes was IdBetermined by transformation experiments. Quantitative reverse transcription PFCR (RT-qPCR) was used to quantify expression levels of eight efflux pump ge nOes, ampC cephalosporinase and outer membrane porin oprD. The isolates were Yfurther genotyped using three PCR-based fingerprinting techniques. Fisher‟s eIxTact test was used to determine the association between MBL and integrons at p S≤ 0.05. Four hundrEed aRnd thirty isolates were identified as P. aeruginosa of which 185(43.0%) were ImVultidrug resistant and 50(11.6%) were extensively drug resistant. All the isolates were resistant to ampicillin, cephalothin and cefuroxime, while sensitivity to UpoNlymyxin B was most common (96.3%). The MICs ranged from 0.125 to >64 µg/mL and 0.0625 to >64 µg/mL against imipenem and meropenem, respectively. All the isolates were negative for Modified-Hodge test, while combined disc test revealed the presence of MBL. Two class B carbapenemases were detected in 86.3% of the carbapenem resistant isolates: blaVIM and blaNDM in 35.6% and 38.4% isolates, respectively, co-existing in 12.3% isolates. Fifty-one (57.5%) carbapenem-resistant P. aeruginosa strains carried class 1 integrons while class 1 and 2 integrons were present concomitantly in 12.3%. Type III effector genes, exoY and exoT were found in all ii isolates, while exoU and exoS were present in 49.3% and 53.4%, respectively. Two isolates possessed both exoU and exoS. Sequence analysis of blaVIM and blaNDM revealed maximum identity with blaVIM-5 and blaNDM-1, respectively. MBL genes were successfully transferred into Escherichia coli DH5α. MexXY-OprM was the most overexpressed pump (5.0 - 996.3 fold increase) occurring in 58.3% of the isolates. The ampC was overexpressed in 27.1% isolates, while oprD porin down-regulation was observed in 77.1% of the isolates. Nine disseminated clones were identified across southwestern states. There was positive association between integrons and MBL (p = 0.0064). RY There is a high incidence of transmissible metallo-beta lactamaAse genes in Pseudomonas aeruginosa from tertiary hospitals in southwestern NBigeRria with different mechanisms mediating carbapenem resistance. blaVIM-5 and LblaINDM-1 were found co- occurring for the first time. There is a need for surv eillance of resistance to carbapenems and associated resistance genes. AN Keywords: Carbapenem resistance, PseudomAonaDs aeruginosa, Efflux pumps, Integrons, Metallo-Beta Lactamases I B Word count: 494 OF ITY ER S IV N U iii ACKNOWLEDGEMENTS My genuine appreciation goes to my supervisor Professor O. E. Adeleke for his commitment, encouragement and immense contribution to the success of this research work. I also want to thank my host supervisor, Professor Habib Bukhari, for his kindness and for providing me with favourable atmosphere for conducting research. I also appreciate my Head of Department, Professor Iruka N. Okeke, for her valuable input. I also appreciate the support of Professor Bolanle A. Adeniyi, Dr P. A. Idow u, Dr Funmilola A. Ayeni, Dr Morenike E. Coker, Dr Temitope O. Lawal, Dr YO. S. Alabi, Mrs. Bolaji B. Oluremi and Mrs. Mojirayo T. Durowaye, may God reRward you accordingly. I am grateful to Professor J. O. Moody and Professor M. RA. AOdeniyi for their generosity. I also want to appreciate the staff of PharmaIceButical Microbiology laboratory, Mr. F. B. Odewale, Mr. O. S. Makinde, Mr. O. J. Olatunde, Mr. J. J. Olajubutu and Mrs. A. O. Ekundayo for their assistance.N L I am as well indebted to the management of the PoAstgraduate College, University of Ibadan, Former Dean, Professor A. A. AderintoD and the present Provost, Professor J. O. Babalola for the award of Teaching Aand Research Assistantship which has contributed greatly to the success of thi s IreBsearch work. I am grateful to The World AcademFy of Science (TWAS) and COMSATS University Islamabad, Pakistan for tYhe a w Oard of Sandwich Fellowship for providing financial assistance and travel IeTxpenses which enable me carry-out part my research work in Pakistan. S I gratefullyE recRognise the commitment of Mr. Matthew Ogundele, Microbiology unit, UniversVity College Hospital, Ibadan, Mrs. Ayeni, Federal Medical Centre, Owo; Mrs. AdeyaInkinnu and Mrs. Aboderin, Federal Medical Centre, Abeokuta; Mr. Ajewole, UOlNabisi Onabanjo University Teaching Hospital, Sagamu; Bisi and Taiwo, Ladoke Akintola University Teaching Hospital, Osogbo; Mr. Ayo, Federal Medical Centre, Ido-Ekiti and Mr. Arasanmi, Obafemi Awolowo University Teaching Hospital, Ile-Ife during isolate collection. iv I also wish to thank Reverend David Kayode-Ige, Reverend Dr. Israel Dada, Pastor Adeagbo and the Pastoral council for their prayers, encouragement and financial support. I also want to thank the Head of Biosciences Department, COMSATS University Islamabad, Pakistan, Dr. Tayyaba Yasmin, for her understanding. The input of Dr. Ramla Shahid, Nanotechnology Department, COMSATS University Islamabad, Pakistan, is greatly appreciated. I am also grateful to my dear mother, Mrs. K. F. Olaniran, who made this possiYble by supporting me prayerfully and financially. I also appreciate my sister andA heRr husband, Mr. and Mrs. S. B. Bashorun, my brothers Babatunde and Ayodeji ORlaniran for their support and care. I am also grateful to my beloved husband, MIr.B Tayo Famojuro and my lovely son, Ayomide Inioluwa Famojuro for their suppor t Land understanding. My colleagues Mrs. Oketola, Mr. Mebude, Mr. OlaArindNe, Mr. Olaleye, Salma, Anas, Zobia, Aisha, Taskeen and other Research students of Microbiology and Nanotechnology Departments, COMSATSA UDniversity Islamabad, Pakistan were appreciated for their unalloyed support . IB All glory the Most-High God: the soFurce of my strength. O ITY S VE R NIU v CERTIFICATION I certify that this project was carried out under my supervision in the Department of Pharmaceutical Microbiology, University of Ibadan, Ibadan, Nigeria. AR Y R IB ……………………………………………N…… L………… Supervisor A O. E. Adeleke (PDh.D) Professor of PharmaBceuAtical Microbiology, University of IbIadan, Ibadan, Nigeria. F Y O SI T VE R NI U vi DEDICATION This research work is dedicated to Most-High God for His endless faithfulness towards me. RY BR A N LI A D IB A OF SI TY R VE UN I vii TABLE OF CONTENTS Title Page i Abstract ii Acknowledgments iv Certification vi Dedication vii Table of contents viii List of Tables x iv List of Figures RY xvi List of Plates A xvii List of main abbreviations R xix CHAPTER ONE IB 1.0 INTRODUCTION L 1 1.1 Pseudomonas aeruginosa as a nosocomial pathogNen 1 1.2 Healthcare-associated infections (HAIs) A 1 1.3 Pseudomonas aeruginosa type III secrAetioDn system (T3SS) 2 1.4 Carbapenem resistance B 3 1.5 Justification for this study I 4 1.6 Hypotheses F 5 1.7 Objectives O 6 CHAPTER TWO TY 2.0 LITERATURIE REVIEW 7 2.1 The test orSganism: Pseudomonas aeruginosa 7 2.2 TheE PsReudomonas aeruginosa type III secretion system (T3SS) 7 2.3 VPseudomonas aeruginosa as a nosocomial pathogen 9 2.4N ICommon hospital-acquired infections (HAIs) caused by P. aeruginosa 10 U2.4.1 Ventilator-associated pneumonia (VAP) 10 2.4.2 Catheter-associated urinary tract infections (CAUTIs) 10 2.4.3 Central line- associated bloodstream infections (CLABSIs) 10 2.4.4 Surgical site /wound infections 11 2.5 Other infections caused by Peudomonas aeruginosa 11 2.51 Gastrointestinal infections 11 2.5.2 Central nervous system infections 11 2.5.3 Ocular infections 12 viii 2.5.4 Endocarditis 12 2.5.5 Ear infections 12 2.5.6 Skin infections 12 2.5.7 Bone and joint infections 13 2.6 Antibiotics 13 2.7 Beta lactam antibiotics 15 2.7.1 Carbapenems 15 2.7.1.1 Imipenem Y 15 2.7.1.2 Meropenem R 17 2.7.1.3 Doripenem A 17 2.7.1.4 Ertapenem R 17 2.8 Carbapenem resistance IB 17 2.8.1 Impermeability of antibiotics mediated by overexpre ssLion of efflux pump systems N 19 2.8.2 Down-regulation of outer membrane porin D A(oprD) 20 2.8.3 Overexpression of ampC cephalosporiAnasDe 20 2.8.4 Mechanism of action and classificaBtion of Beta lactamases 21 2.8.4.1.1 Class A carbapenemases I 22 2.8.4.2 Class D carbapenemases F 23 2.8.4.3 Molecular Class B Car bOapenemases (Metallo beta-lactamases) 24 2.8.5 TYPES OF MBL Y 24 2.8.5.1 Imipenem (IMIPT) type MBL 24 2.8.5.2 Veronese ISmipenemase (VIM) type MBL 25 2.8.5.3 SaoE PaRulo Imipenemase (SPM-1) type MBL 25 2.8.5.4 VGermany Imipenemase (GIM-1) type MBL 25 2.8N.5.5I Seoul Imipenemase (SIM-1) type MBL 25 U2.8.5.6 Adelaide Imipenemase (AIM-1) type 26 2.8.5.7 Kyorin Hospital Imipenemase (KHM) 26 2.8.5.8 New Delhi Imipenemase (NDM) 26 2.8.5.9 Dutch Imipenemase (DIM) 26 2.8.5.10 Florence Imipenemase (FIM) 26 2.9 Clinical significance of carbapenem resistance Gram-negative bacteria 27 2.10 Mobile genetic elements 27 2.10.1 Integrons 27 ix 2.10.2 Plasmids 29 2.10.3 Miniature Inverted-repeat Transposable Elements (MITEs) 29 2.10.4. Genomic islands (GIs) 29 2.10.5 Insertion sequences (ISs) 29 2.10.6 Transposons 30 2.11 Molecular typing 30 2.11.1 Pulsed-field gel electrophoresis (PFGE) 30 2.11.2 Multi Locus Sequence Typing (MLST) 31 2.11.3 Repetitive element sequence-based polymerase chain reaction (rep-PCRR) Y 32 2.11.4 Restriction Fragment Length Polymorphism (RFLP) A 33 2.11.5 Randomly amplified polymorphic DNA (RAPD) R 33 CHAPTER THREE IB 3.0 MATERIALS AND METHODS L 35 3.1 Materials N 35 3.1.1 Equipment and glassware A 35 3.1.2 Media, Buffers, Chemicals D 35 3.1.3 DNA isolation and purification kits, enAzymes, molecular weight markers and Master mix IB 35 3.1.4 List of primers F 35 3.1.5 Study design and Place oOf study 36 3.1.6 Bacterial isolates Y 36 3.2 Methods I T 42 3.2.1 PhenotypicS identification of the study isolates 42 3.2.1.1 CulEturaRl identification of isolates 42 3.2.1.2 V Gram‟s staining 42 3.2N.1.3I Oxidase test 42 U3.2.1.4 Gelatin liquefaction test 43 3.2.1.5 Catalase test 43 3.2.1.6 Urease test 43 3.2.1.7 Hydrogen sulphide production 43 3.2.1.8 Citrate test 44 3.2.1.9 Oxidative utilisation of sugars 44 3.2.2 Antimicrobial susceptibility testing 44 3.2.2.1 Disc diffusion 44 x 3.2.2.2 Determination of minimum inhibitory concentrations (MICs) 45 3.2.3.1 Detection of beta-lactamase production 45 3.2.3.2 Screening of pathogen for carbapenemase production 46 3.2.3.3 Combined disc test for phenotypic detection of metallo-beta-lactamases (MBLs) 46 3.2.3.4 Curing of antibiotic resistance 47 3.2.4. Molecular methods 47 3.2.4.1 Isolation of plasmid DNA 47 3.2.4.2 Extraction of genomic DNA RY 48 3.2.4.3 Extraction of total RNA A 49 3.2.4.4 Quantification of DNA /RNA R 50 3.2.4.5 Gel preparation and electrophoresis IB 50 3.2.4.6 Preparation of sample for agarose gel electrophoresis L 50 3.2.4.7 Visualisation of DNA N 50 3.2.4.8 Molecular identification of Pseudomonas aerAuginosa 50 3.2.5 Detection of carbapenemase encoding genDes by PCR 51 3.2.5.1 Class A carbapenemases B A 51 3.2.5.2 Class B carbapenemases (MBL s)I 51 3.2.5.3 Class D carbapenemases F 52 3.2.6.1 Characterisation of cla ssO 1, 2 and 3 integron 53 3.2.6.2 Restriction FragmeYnt Length Polymorphism (RFLP) for differentiation of integrons I T 53 3.2.6.3 CharacterisSation of cassette arrays 53 3.2.7 PurEificaRtion and Sequencing of PCR products 54 3.2.7.1 VPurification of PCR products 54 3.2N.7.2I Sequencing of PCR products 54 U3.2.8 Transformation experiment 54 3.2.9 Statistical analysis 55 3.2.10 Quantification of gene expression 55 3.2.10.1 Removing the co-purified contaminating DNA and synthesis of First strand complementary DNA (cDNA) 55 3.2.10.2 Expressions level quantification of efflux pumps, ampC and oprD transcripts 56 3.2.10 Detection of type III secretion system 57 xi 3.2.12 PCR-based genotyping 58 3.2.12.1 Repetitive extragenic palindromic PCR (REP PCR) 58 3.2.12.2 Enterobacterial Repetitive Intragenic Consensus-PCR (ERIC PCR) 58 3.2.12.3 BOX PCR 59 CHAPTER FOUR 4.0 RESULTS 60 4.1 Identification of clinical isolates of Pseudomonas aeruginosa 60 4.2 Distribution of clinical isolates of P. aeruginosa according to site of isolation in relation to hospital RY 60 4.3 Antibiotic susceptibility profile of clinical isolates of P. aeruginosAa 64 4.4. Minimum inhibitory concentrations (MICs) of selected antibioRtics against clinical isolates of P. aeruginosa IB 65 4.5 Phenotypic detection of beta lactamase and metallo-b eLta lactamase (MBL) 71 4.6 Curing of antibiotic resistance in carbapenem-resNistant P. aeruginosa 71 4.7 Molecular identification of P. aeruginosa A 82 4.8 Molecular detection of class A and D AcarbDapenemases 86 4.9 Amplification of MBL-resistance gBenes in carbapenem resistant P. aeruginosa I 86 4.10 Analysis of sequenced PCR Fproducts 87 4.10.1 Alignments of blaVIM a nOd blaNDM sequenced amplicons using Basic Local Alignment Search TYools (BLASTn) 87 4.10.2 GenBank acceIssTion numbers of sequenced blaNDM-1 and blaVIM-5 genes 88 4.11 TransformaStion experiments 89 4.12 AmEplifRication of integron and integrase gene cassette 105 4.13 VPrevalence of type III effector toxins in carbapenem resistant P. NIaeruginosa 106 U4.14 Statistical analysis 106 4.15 Quantification of efflux pumps expression in carbapenem-resistant P. L 2 3 4 5 6 7 8 9 10 11 12 13 14 15 N aeruginosa 128 4.16 Quantification of ampC overexpression and diminished expression of oprD Porin 129 4.17 MIC of antibiotics against carbapenem-resistant P. aeruginosa isolates in relation to MBL and efflux pump genes 129 4.18 Molecular typing of carbapenem-resistant P. aeruginosa 141 xii CHAPTER FIVE 5.0 DISCUSSION 154 5.1 Distribution of isolates in clinical samples 154 5.2 Antibiotic susceptibility profile of clinical isolates of P. aeruginosa 154 5.3 Evaluation of antibiotic susceptibility profile of phenotypically detected MBL-producing and non-MBL-producing strains 157 5.4 Molecular detection of carbapenemases 157 5.5 PCR-RFLP analysis of integrons in carbapenem-resistant P. aeruginosa RY 161 5.6 MDR efflux pump overexpression, ampC overexpression and oprDA underexpression in carbapenem-resistant P. aeruginosa R 162 5.7 Association of carbapenem resistance with increased expIreBssion of ampC, efflux pump and oprD underexpression L 164 5.8 Combination of resistance mechanisms in CRPAN 165 5.9 Prevalence of type III effector toxins in CRPAA 165 5.10 Molecular typing of CRPA with three repDetitive sequence-based PCR methods A 167 5.11 Relationship between Type Thr eeI BSecretion System (T3SS) and Repetitive Element Sequence-Based PCFR (rep-PCR) 169 5.12 Limitations of this stud yO 169 5.13 Recommendations Y 169 5.14 Contributions ItoT knowledge 170 5.15 ConclusionS 170 ReferencesE R 172 LIST OVF AWARDS 197 AppenIdices UApNpendix I. Media and antibiotics 198 Appendix II. Chemicals, Enzymes, Master mix and DNA isolation kits 204 Appendix III. Statistical analysis 206 Appendix IV. Selected genbank flatfiles and BLASTn of blaVIM and blaNDM sequences 216 xiii LIST OF TABLES Table 3.1. Primers used for molecular screening in this study 37 Table 3.2. Primers used for typing in this study 41 Table 4.1. Occurrence of P. aeruginosa strains according to hospital with respect to their clinical sources 63 Table 4.2. Antibiotic susceptibility profile of clinical isolates of P. aeruginosa 67 Table 4.3. Classification of clinical isolates of P. aeruginosa based on antibiotic resistance profiles Y 69 Table 4.4. Antibiotic susceptibility rates of 430 clinical isolates of P. aeruginRosa at indicated MIC in µg/mL A 70 Table 4.5. Prevalence of beta lactamase and MBL in clinical isolates oRf CRPA using phenotypic method IB 73 Table 4.6a. Effect of 100 μg/mL ethidium bromide on the su scLeptibility of carbapenem-resistant strains of P. aeruginosaA to Nimipenem and meropenem D 78 Table 4.6b. Effect of 50 μg/mL ethidium bro mAide on the susceptibility of carbapenem- resistant strains of P. aeruginosa toB imipenem and meropenem 80 Table 4.7. Molecular detection of CFlass BI Carbapenemases (MBLs) in carbapenem- resistant P. aeruginosa 90 Table 4.8. Distribution of meta lOlo beta-lactamase (MBL) genes in carbapenem- resistant P. aeruginYosa according to hospital 93 Table 4.9. DistribuStionI Tof metallo beta-lactamase (MBL) genes in carbapenem- resistanRt P. aeruginosa according to clinical source 94 Table 4.10.E MIC of selected antibiotics against carbapenem-resistant clinical isolates Vof P. aeruginosa in relation to type of MBL gene possessed 95 TaNble I4.11. Distribution of integrase genes in carbapenem-resistant P. aeruginosa U according to hospital 107 Table 4.12. Distribution of integrase genes in carbapenem-resistant P. aeruginosa according to clinical source 108 Table 4.13. PCR-Restriction fragment length polymorphism investigation of integrons and characterisation of cassette arrays in carbapenem-resistant P. aeruginosa 109 xiv Table 4.14. Presence of MBL gene(s) and integrons in carbapenem non-susceptible P. aeruginosa 112 Table 4.15. Distribution of T3SS in carbapenem-resistant P. aeruginosa according to hospital 120 Table 4.16. Prevalence of Type III effector toxins in carbapenem-resistant P. aeruginosa according to clinical source 121 Table 4.17. Relative fold expression of efflux pump genes in carbapenem- resistant P. aeruginosa clinical isolates compared to P. aeruginosa ATCC 27853 RY 131 Table 4.18. Relative fold expression of ampC and outermembrane porin (AoprD) in carbapenem-resistant P. aeruginosa clinical isolates comparedR to P. aeruginosa ATCC 27853 IB 133 Table 4.19. MDR efflux gene overexpression, ampC overexp rLession and oprD in carbapenem-resistant P. aeruginosa N 135 Table 4.20. MIC of antibiotics against carbapenem-reAsistant P. aeruginosa isolates in relation to MDR efflux gene overexpreDssed and MBL 137 Table 4.21. Singles and combinations of varioAus resistance mechanisms in CRPA 140 Table 4.22. Comparison between REP- , IEBRIC- and BOX- PCR 150 Table 4.23. Indistinguishable isolateFs with REP-PCR 151 Table 4.24. Indistinguishable i sOolates with BOX-PCR 152 Table 4.25. Genotypic chaYracteristic of four clones confirmed by two or more rep- PCR methods I T 153 ER S IV U N xv LIST OF FIGURES Figure 2.1. A typical image of P. aeruginosa with its virulence factors 8 Figure 2.2. Mechanism of action of antibacterial drugs 14 Figure 2.3. Structures of beta-lactam families 16 Figure 2.4. Mechanisms of antibiotic resistance in bacteria 18 Figure 4.1. Distribution of the isolates according to clinical source 62 Figure 4.2. Percentage antibiotic resistance of 430 isolates of P. aeruginosa from seven hospitals 68 Figure 4.3. Comparison of antibiotic sensitivity among MBL-positive and MBRL-Y negative P. aeruginosa isolates *(p<0.05) A 77 Figure 4.4. Distribution of MBL(s), integron and exoU in carbapenemR-resistant P. aeruginosa (CRPA) according to hospital IB 126 Figure 4.5. Distribution of MBL(s), integron and exoU in ca rbLapenem-resistant P. aeruginosa (CRPA) according to clinical sourNce 127 Figure 4.6. Comparison of proportion of two efflux gAenes overexpressed in each of the four efflux pump system in carbapAeneDm-resistant P. aeruginosa 136 Figure 4.7. Occurrence of various carbapeBnem resistance mechanisms in carbapenem- resistant P. aeruginosa I 139 Figure 4.8. Dendrogram showing clFuster analysis of carbapenem-resistant P. aeruginosa by REP-PC RO using Phylotree software and Unweighted Pair Group MethodY with Arithmetic Mean (UPGMA) 145 Figure 4.9. DendrograImT showing cluster analysis of carbapenem-resistant P. aeruginosaS by ERIC-PCR using Phylotree software and Unweighted PairE GrRoup Method with Arithmetic Mean (UPGMA) 147 Figure 4V.10. Dendrogram showing cluster analysis of carbapenem-resistant P. N I aeruginosa by BOX-PCR using Phylotree software and Unweighted U Pair Group Method with Arithmetic Mean (UPGMA) 149 xvi LIST OF PLATES Plate 4.1. Zone of growth inhibition around antibiotic discs 66 Plate 4.2. Modified Hodges test for detection of carbapenemases 74 Plate 4.3a. EDTA impregnated imipenem and meropenem discs in comparison with the plain imipenem and meropenem discs 75 Plate 4.3b. Inhibition zone size augmentation with the EDTA soaked imipenem and meropenem discs in comparison with the plain imipenem and meropenem discs 76 Plate 4.4a. Row 1: Agarose gel electrophoresis of PCR products (1.5%) for thRe Y identification of Pseudomonas spp. using oprI Genus specific primeArs. Row 2: PCR products for the identification of P. aeruginosa usiRng oprL species specific primers IB 83 Plate 4.4b. Agarose gel electrophoresis of PCR products for thLe proof of identity of P. aeruginosa using oprL species specific primeNr 84 Plate 4.4c. Agarose gel electrophoresis of PCR produActs for the identification of P. aeruginosa using oprL species specAificD primer 85 Plate 4.5a. Agarose gel (1.5%) showing PCBR products for blaVIM 98 Plate 4.5b. Agarose gel (1.5%) showing IPCR products for blaVIM 99 Plate 4.5c. Agarose gel (1.5%) showFing PCR products for blaVIM 100 Plate 4.6a. Agarose gel (1.5%) sOhowing PCR products for blaNDM 101 Plate 4.6b. Agarose gel (1.Y5%) showing PCR products for blaT NDM 102 Plate 4.6c. Agarose geIl (1.5%) showing PCR products for blaNDM 103 Plate 4.7. AgaroseS gel (1.5%) showing PCR products for blaNDM after tranEsfoRrmation 104 Plate 4.V8a. Agarose gel (1.5%) showing PCR products for integrons before digestion N I with Rsa1 enzyme 115 UPlate 4.8b. Agarose gel (1.5%) showing PCR products for integrons before digestion with Rsa1 enzyme 116 Plate 4.9a. Agarose gel (1.5%) showing PCR products for integrons showing different carbapenem-resistant P. aeruginosa positive for class 1 and 2 integrons after digestion of PCR-products with RsaI enzyme 117 Plate 4.9b. Agarose gel electrophoresis of integrons showing different carbapenem resistant P. aeruginosa positive for class 1 and 2 integrons after digestion of PCR-products with RsaI enzyme 118 xvii Plate 4.10. Agarose gel electrophoresis (1.5%) of class 1 integron gene cassette 119 Plate 4.11a. Genotyping of exoU, exoS, exoT and exoY in carbapenem-resistant P. aeruginosa (CRPA) isolates with multiplex PCR 122 Plate 4.11b. Genotyping of exoU, exoS, exoT and exoY in carbapenem-resistant P. aeruginosa (CRPA) isolates by multiplex PCR 123 Plate 4.11c. Genotyping of exoU, exoS, exoT and exoY in carbapenem-resistant P. aeruginosa (CRPA) isolates by multiplex PCR 124 Plate 4.11d. Genotyping of exoU, exoS, exoT and exoY in carbapenem-resistant P. aeruginosa (CRPA) isolates by multiplex PCR R Y 125 Plate 4.12. Agarose gel (1.0%) showing total RNA extracted from carbapeAnem- resistant P. aeruginosa with PureLinkTM Micro-to-Midi TotaRl RNA Extraction System (Invitrogen) IB 130 Plate 4.13. Representative image of fingerprinting patterns o f Lcarbapenem-resistant P. aeruginosa by gel electrophoresis following RNEP-PCR 144 Plate 4.14. Representative image of fingerprinting paAtterns of carbapenem-resistant P. aeruginosa by gel electrophoresis fAolloDwing ERIC-PCR 146 Plate 4.15. Representative image of fingerBprinting patterns of carbapenem-resistant P. aeruginosa by gel electropho rIesis following BOX-PCR 148 OF ITY RS IV E U N xviii LIST OF MAIN ABBREVIATIONS Abbreviations Full meaning ABC Adenosine triphosphate-binding cassette superfamily AIM Adelaide Imipenemase ampC Chromosomal cephalosporinase ATCC American Typed Culture Collection BLAST Basic local alignment search tool BIC-1 Bicêtre carbapenemase CDC Centre for disease prevention and control RY CI Chromosomal integrons A CLSI Clinical laboratory standard institute R CRPA Carbapenem-resistant Pseudomonas aerugIinBosa DHP-I Dehydropeptidase I L DIM Dutch imipenemase N DNA Deoxyribonucleic acid A EDTA Ethylene diamine tetra AaceDtic acid ERIC-PCR Enterobacterial repeBtitive intragenic consensus sequence PCR ESBL Extended spectru mI beta-lactamase FIM Florence ImipFenemase FMCA Federal MOedical Centre, Abeokuta FMCI FedeYral Medical Centre, Ido-Ekiti FMCO IFTederal Medical Centre, Owo GES S Guiana extended spectrum GIM R Germany Imipenemase GIs V E Genomic islands HANI I Healthcare associated infections UICU Intensive care unit IMI-1 Imipenem-hydrolysing beta-lactamase IMP Imipenem type metallo-beta lactamase Intl Integrase gene IS Insertion sequence ISCR19 Insertion sequence common region 19 kb kilobase kDa kilodalton xix KHM Kyorin Hospital Imipenemase KPC Klebsiella pneumoniae carbapenemase LTHO Ladoke Akintola University Teaching Hospital, Osogbo MATE Multidrug and toxic compound extrusion family MBL Metallo beta-lactamase MDR Multidrug resistance MFS Major facilitator superfamily MGE Mobile genetic element MIC Minimum inhibitory concentration RY MITEs Miniature Inverted-repeat Transposable Elements A MLEE Multilocus enzyme electrophoresis R MLST Multi Locus Sequence Typing IB NDM New Delhi metallo-beta lactamase L NmcA not metalloenzyme carbapenemasNe A OmpF Outer membrane protein F A OprD Outer membrane porin D D OTHI Obafemi Awolowo BUnAiversity Teaching Hospital Complex. Ile-Ife I OTHS Olabisi OnabFanjo University Teaching Hospital, Sagamu OXA Oxacilli nO-hydrolyzing beta-lactamases PBPs PeniYcillin-binding proteins PCR IPTolymerase chain reaction PFGE S Pulsed-field gel electrophoresis RAPD ER Randomly Amplified Polymorphic DNA REP-PCVR Repetitive extragenic palindromic PCR RFNLPI Restriction fragment length polymorphism UR A Ribonucleic acid RND Resistance nodulation division family RT-PCR Reverse transcriptase polymerase chain reaction s seconds SDS Sodium dodecyl sulphate SFC-1 Serratia fonticola carbapenemase-1 SIM Seoul imipenemase SME Serratia marcescens enzyme xx SMR Small multidrug resistance family SPM Sao Paulo Imipenemase SSI Surgical site infection T3SS Type III secretion system TAE Tris acetate EDTA Tns Transposons UCHI University College Hospital, Ibadan UTI Urinary tract infection UV Ultraviolet RY VAP Ventilator-Associated Pneumonia A VIM Veronese Imipenemase R WHO World Health Organisation IB XDR Extensively drug resistance L N AD A B OF I ITY S ER IV UN xxi Y AR R IB L N DA Detection of Bacterial PathogAens in Cerebrospinal Fluid using Restriction Fragme nItB Length Polymorphism AT Kalghatgi,* AK Praharaj,+ AK Sahni,# DF Pradhan,** S Kumaravelu,++ PL Prasad,## and A Nagendra, (Retd)*** O TY RS I IV E UN xxii CHAPTER ONE INTRODUCTION 1.1 Pseudomonas aeruginosa as a nosocomial pathogen Pseudomonas aeruginosa is a Gram-negative opportunist pathogen which tak es advantage of the breach in the host immune system to cause infection. This Yis the reason P. aeruginosa is commonly implicated in deadly bacterial AinfeRctions of immunodepressed people but rarely causes disease in healthy individRuals (Dean et al., 2008). P. aeruginosa produces a high mortality rateI Bof over 70% in immunocompromised individuals (Mackie and MacCartne y,L 1996). P. aeruginosa is the main nosocomial pathogen which causes about 10%N of infections in most hospitals followed by Staphylococcus aureus (Fazeli et al., 2A012). The US Centre for Disease Control and Prevention (2014) estimated thAat DP. aeruginosa accounted for 51,000 healthcare-associated infections and over 6,000 (13%) are caused by multidrug resistant P. aeruginosa with about 40 0I dBeaths occurring every year in the USA. P. aeruginosa is the second and thirdF most frequently isolated nosocomial pathogen in respiratory tract infection (17% O), urinary tract and surgical establish infections (11%), respectively (Richards et al., 1999). TY 1.2 HealthcarSe-AIssociated Infections (HAIs) Healthcare asRsociated infections otherwise known as nosocomial infections are infections Ethat were not present prior to patients‟ admission but acquired within 72 hours IoVf hospitalisation (World Health Organisation, 2002). Patients that are admitted inN the sickbay, specifically those in the critical care units are liable to contract Uinfections from hospital staff or infected medical devices (Khan et al., 2017). Nosocomial infections are caused by bacteria, fungi and viruses. However, bacteria especially Gram-negative bacteria, are more often than not, connected with nosocomial infections (Ige et al., 2011; Khan et al., 2017). HAIs contribute to death in infected individuals because majority of these infections are produced by microbes that are not sensitive to many antibiotics (Ige et al., 2011). Nosocomial infections majorly accompany the use of invasive devices and mostly affect sick individuals in critical 1 care divisions (Dia et al., 2008). Surgical-site infection was the principal infection in hospitals in both advanced and unindustrialised countries (Atif et al., 2006; Ige et al., 2011; Nejad et al., 2011). Frequently reported nosocomial infections include bloodstream, catheter-related urinary tract, surgical site and ventilator-linked pneumonia (Khan et al., 2017). HAI is a global challenge but the burden of HAIs is considered highest in developing countries (Nejad et al., 2011). Most hospitals in developing countries are overcrowd ed and understaffed leading to poor infection control measures. Moreover, absenYce of infection control strategies and skillful personnel also worsen the issue (NeRjad et al., 2011). There is little statistics on the epidemiology of nosocomial infectioAns in African countries (Vincent et al., 1995; Nejad et al., 2011). The pooIlBed Rpredominance of nosocomial infection from some countries in Africa was 15 ·5L% which was more than rate reported from the USA and Europe (Nejad et al., 20N11). Occurrence rate of 2.5%, 10.3%, 10.9%, 14.8% were observed from Algeria,A Ethiopia, Senegal and Tanzania, respectively (Gosling et al., 2003; Atif et al., 200D6; Dia et al., 2008; Mulu et al., 2012). Prevalence of HAIs in a tertiary hospital in IbAadan increased slowly from 2.4% in 2005 to 3.1% in 2008 (Ige et al., 2011). Oth eIr Bstudies from Tanzania and Nigeria showed a reduction in the prevalence of HAI Ffrom 9.5 in 2001 to 4% in 2005, and 5.8% in 2003 to 2.8% in 2006, respectivel y Owhen an infection control practice was implemented (Atif et al., 2006; AbubakaYr, 2007). 1.3 PseudomoSnasI a Teruginosa Type III Secretion System (T3SS) Type III secrRetion systems (T3SS) are virulence factors found in Gram-negative bacteria whEich enable them to bypass the extracellular milieu by injecting bacterial effectoIrV proteins straight into the cytoplasm of the infected individual cell (Coburn et alN., 2007). Many P. aeruginosa strains produce an extremely harmful T3SS upon Ucontact with human. In P. aeruginosa, there are four known type III effector toxins that function as a virulence factor. Over 80% of P. aeruginosa isolates from serious infections express these toxins (Hauser et al., 2002). The complement of effectors differs among strains of P. aeruginosa. Nevertheless, most strains have exoT and exoY. exoU and exoS are not usually present together in one strain; that is the presence of one may mean the absence of the other and vice versa (Roy-Burman et al., 2001). Exotoxin U is the most destructive of the type III effector proteins with phospholipase A2 activity which is only expressed by a few hospital isolates (Sato et al., 2003). 2 Strains expressing exoU are highly cytotoxic. Exotoxin Y is an adenylate cyclase and has a moderate impact on virulence. Exotoxin S and exotoxin T are double-functional enzymes, though effects of exoT in overall virulence were not as impressive as exoS. Exotoxin S adds to the capacity of P. aeruginosa to pass over the epithelial barricade (Soong et al., 2008). T3SS has been linked with additional severe proven illness in mortal patients (Sato et al., 2003). For instance, the existence of efficient T3SS has been linked with poor aftermath of infection in ventilator-related pneumonia patients (Hauser et al., 2002). It has been established that type III toxins are virulen ce determinants that aid the spread of P. aeruginosa from burn wounds (NiRcas Yet al., 1985). A study that examined the contribution of a lively T3SS in PA. aeruginosa infection has established that great relationship exists between T3SS mRanifestation and mortality in patients infected with P. aeruginosa (Roy-Burman eIt Bal., 2001). L 1.4 Carbapenem resistance Resistance of bacteria to antibiotics is no longer aA neNw burden worldwide. As new antibiotics come into clinical use, bacteria aDlso devise means of surviving the antibacterial action of antibiotics. P. aeruginoAsa is notorious having resistance to a lot of antibiotics intrinsically with abilityI tBo obtain and disseminate resistance genes among themselves and other bacteriFal g enera. Their ability to form biofilm and survive in low nutrient environments suOch as in catheters also aids the spread of this organism in the hospital settings (Mayhall, 1996). Multidrug resistant P. aeruginosa are becoming difficult to ItrTeat Y(Bassetti et al., 2018). Therefore, P. aeruginosa infection in critically ill patienSts is a great concern. CarbapenemEs aRre used in the treatment of infections caused by Gram-negative bacteria that pIroVduce expanded spectrum beta-lactamase as well as P. aeruginosa. However, reNsistance to carbapenems is being witnessed in these isolates. Carbapenem resistance Uin bacteria is a serious threat to the healthcare system because carbapenem resistant bacteria are well-known to be resistant to several beta lactam drugs and other antibiotics, however only susceptible to polymyxins, leaving physicians with few or no treatment options (Varaiya et al., 2008). Carbapenem resistance among P. aeruginosa has been shown to be mediated by acquired carbapenemases most especially the metallo-beta lactamases (MBLs) (Villegas et al., 2007). However, the presence of intrinsic resistance devices such as upregulation of efflux pump systems, exceeding manifestation of chromosomal cephalosporinase and decreased outer membrane 3 permeability has also been observed. There is paucity of information on the genetic basis of carbapenem resistance in P. aeruginosa from Nigeria. Therefore, the present study will be conducted to molecularly characterise carbapenem resistance genes in clinical isolates of P. aeruginosa from Southwest Nigeria. 1.5 Justification for this study Pseudomonas aeruginosa is an emerging pathogen and the frequent source of urinary tract and surgical site infections in patients in critical care units. It also plays a serio us part in producing persistent respiratory infections in cystic fibrosis patientRs (MYackie and MacCartney, 1996; Gaynes and Edwards, 2005). P. aeruginosa has mortality rate in immunosuppressed individuals particularly patients with severe burAns or cancer (Mackie and MacCartney, 1996). Carbapenems are potent beta-laBctaRm antibiotics that are used in serious nosocomial infections particularly those LthaIt are triggered off by Gram-negative bacteria that produce extended spectrum beta -lactamases owing to their broad spectrum of action and steadiness to hydrolysAis bNy majority of beta-lactamases (Gupta, 2008; Armand-Lefèvre et al., 2013).D However, increased prevalence of resistance to carbapenems is being observed Aamongst Gram-negative organisms more frequently in non-fermenters compris inIgB P. aeruginosa and Acinetobacter species (Lolans et al., 2005; Mohammed anFd Raafat, 2011). The genes for carbapenemY res is Otance are typically carried on mobile genomic elements such as plasmids and tTransferred among bacterial genera and species (Nicasio et al., 2008). The ability Sto aIcquire resistance genes has made P. aeruginosa resistant to most beta-lactam anRtibiotics and to develop resistance to many other antibiotics, bringing about veryE limited therapeutic possibilities (Poulakou et al., 2014). There have been reportIs Von outbreak of pandrug-resistant P. aeruginosa. For instance, in Belgium, an ouNtbreak of pan-resistant P. aeruginosa (resistant to all antibiotics except the Upolymyxins) having upregulation of efflux pump systems, chromosomal ampC cephalosporinase and deficient in outer membrane porins was reported (Deplano et al., 2005). An epidemic of pan-drug resistant P. aeruginosa, which produced blaVIM-2 metallo-beta lactamase, chromosomal ampC beta-lactamase and aminoglycoside modifying enzymes (aacA7 and aaC-A5) in a critical care unit in Chicago has also been documented (Lolans et al., 2005). Also, in Lithuania, increased prevalence of MBL-producing P. aeruginosa from 15.8% in 2003 to 61.9% in 2008 has also been 4 documented (Vitkauskiene et al., 2011). It is crucial to monitor dissemination of P. aeruginosa and identify carbapenem resistance encoding genes to aid infection control. An increase in knowledge about the predominance and tools of resistance to carbapenems in P. aeruginosa may be a valuable guide in the selection of antibiotics and possibly help to prevent associated morbidity and mortality (Meradji et al., 2015). There are few antimicrobial options for multi-drug resistant organisms in clinical trials; hence, regular monitoring of this pathogen is highly essential. The genomic basis for carbapenem resistance in P. aeruginosa has been intenYsively studied in developed countries, but there are limited data from sub-SaharRan Africa (Cholley et al., 2014), including Nigeria. Some studies from Nigeria havAe reported the occurrence of MBL in Gram-negative bacteria by phenotypic BmeRthods, while the presence of blaKPC, blaNDM and blaVIM type carbapenemase Lin Icarbapenem resistance Enterobacteriaceae from Northern, Nigeria have also beeNn d ocumented (Mohammed et al., 2015; Abdullahi et al., 2017). There is scarcity Aof information on the prevalence and tool of resistance to carbapenems in clinDical isolates of P. aeruginosa from Nigeria. Understanding of the prevalence aAnd tools of carbapenem resistance in P. aeruginosa is of paramount importanc eI iBn order to aid infection control and monitor nosocomial spread. F 1.6 Hypotheses O Resistance to carbapenemYs in P. aeruginosa has been known to result from outer membrane protein loIssT, efflux systems, enzyme production (MBL) and mutations among which RMBSL production is of great clinical concern because of its broad hydrolysis Eprofile and rapid rate of dissemination. The research hypotheses are: 1. IVMBL genes have high prevalence among carbapenem-resistant P. aeruginosa N from Southwest Nigeria hospitals U 2. Occurrence of MBL resistance genes is closely related to the occurrence of efflux pump genes in carbapenem-resistant clinical isolates of P. aeruginosa 3. Occurrence of metallo-beta lactamases and integrons in carbapenem-resistant P. aeruginosa from Southwest Nigeria are not mutually exclusive 4. Carbapenem-resistant clinical isolates of P. aeruginosa from Southwestern Nigeria hospitals are clonally related 5 1.7 General objective To determine the prevalence of carbapenem resistance and characterise carbapenem- resistance encoding genes in clinical isolates of P. aeruginosa from Southwestern Nigeria. Specific objectives 1. To determine the antibiotic susceptibility profile of clinical isolates of P. aeruginosa to antibiotics and relate the antibiotic susceptibility pattern of MBL-positive to MBL-negative P. aeruginosa clinical isolates Y 2. To characterise carbapenemase genes and determine expression lAeveRl of efflux pump systems, AmpC chromosomal cephalosporinase and Router membrane porin (OprD) in carbapenem-resistant P. aeruginosa B 3. To detect the presence of resistance-plasmids amongL MIBL-producing isolates and determine the transferability of carbapenemN-res istance encoding genes by transformation experiment 4. To determine the molecular epidemDioloAgy of carbapenem-resistant P. aeruginosa and establish the mechaAnisms for carbapenem resistance in P. aeruginosa from Southwest Nig F eIriBa. O TY RS I E NI V U 6 CHAPTER TWO LITERATURE REVIEW 2.1 The test organism: Pseudomonas aeruginosa Pseudomonas aeruginosa is a rod-formed Gram-negative bacterium which hYas its place in Pseudomonads family. P. aeruginosa is an opportunistic nosocomialR pathogen and the major cause of depression and death in immunosuppressed host Aor those with comorbid illness such as prolonged renal failure or cystic fibrosis (CRhastre and Fagon, 2002). P. aeruginosa causes infection in various parts of the bodIyB, including ear (otitis externa), eye (corneal ulcers and keratitis), heart (endocardit isL), central nervous system (meningitis) and infections of the urinary tract, burns, bNone, joints, wound as well as, infection of the respiratory tract (Bielecki et al.A, 2008; Galle et al., 2012). P. aeruginosa possess a large collection of virulAencDe features that permit them to flourish in an animate domicile and escape host Bresponse (Reinhart and Oglesby-Sherrouse, 2016). Virulence factors play a criti cIal role in establishment of infections by P. aeruginosa. Virulence factors inclFude endotoxin (lipopolysaccharides), pili, capsule and several toxins including t yOpe III secretion system (T3SS) toxins which makes it easy for P. aeruginosa to eYscape host defence (Davies and Bilton, 2009). The existence of biofilm also allowITs P. aeruginosa to stick to diverse medical tools such as indwelling catheteSr and the air route of cystic fibrosis patients and heavily contributes to wound rEelatRed morbidity and mortality globally (Mayhall, 1996). Figure 2.1 shows virulencVe factors produced by P. aeruginosa. 2.2N IThe Pseudomonas aeruginosa type III secretion system (T3SS) UType III secretion system is found in majority of Gram-negative bacteria. It is a main virulence factors that maneuver eukaryotic host cell response (Galle et al., 2012). T3SS introduces effector toxins straight into the mammalian host cell in a very well- coordinated manner (Galle et al., 2012). T3SS and its effectors were first detected in P. aeruginosa in 1996 and are the foremost virulence determinants of P. aeruginosa (Galle et al., 2012). The T3SS of P. aeruginosa comprises of many proteins that form compound which is structurally and functionally conserved (Moraes et al., 2008). Four 7 Y R RA LI B N DA BA Figure 2.1. A typical image of P. aerFug inIosa with its virulence factors Source: Sadikot et al. (2005). O Y IT S VE R I UN 8 effector toxins identified in P. aeruginosa thus far include: exoY, exoT, exoU and exoS. Not all the four effector toxins can be found in one strain, though exoT and exoY may be existing in all strains but dissimilar strains have one or the other of the exoS and exoU gene that is, exoS and exoU are mutually exclusive in P. aeruginosa so that the populace is at present set apart into a main exoS-containing strain, a negligible exoU- containing strain, while both exoS and exoU T3SS may be lacking in some strains of P. aeruginosa (Freschi et al., 2015). Exotoxin U is extremely cytotoxic and more destructive than exoS. The expression of T3SS effector proteins correspond with po or outcome of a disease (Hauser, 2009). RY Apart from the four recognised effector toxins, additional two effeRctoAr toxins have been established in P. aeruginosa; pemA and pemB but ItBhese genes are not indispensable machineries of T3SS. The effectors are typic aLlly found in all strains of P. aerginosa as with exoY and exoT (Burstein et al., 20N15). The two effector cells fail to stimulate a toxic reaction in Saccharomyces cerevisiae. Therefore, Burstein and his fellow worker concluded that lack of toxicity obsDerveAd in their study may stipulate that the two genes may be toxic on lesser collectAion of hosts or could be activated by yet undiscovered signals. The effectors mayI aBlso be host particular, in other words, the two effector genes may be toxic in anothe r host. On the other hand, these effectors may interrelate with the host inhere nOt pr Fotection or normalise definite passage-ways for the advantage of the bacteria. This has opened way for additional investigation into translating the roles anIdT deYvices of these novel effectors (Burstein et al., 2015). 2.3 PseudomoSnas aeruginosa as a nosocomial pathogen HospitaliseEd pRatients (especially prolonged ICU stay) stand a great danger of being inhabiteVd and infected with P. aeruginosa and causes between 10 to 20% of noNsocIomial infections (Fazeli et al., 2012). Amongst the family of Pseudomonads, P. Uaeruginosa is the greatest regularly isolated nosocomial pathogen in urinary tract infection, ventilator associated pneumonia and bacteraemia (Elkhatib et al., 2008). It is the chief cause of the long-lasting debilitating lung infection, which is the primary basis of death in those suffering from cystic fibrosis (Govan and Nelson, 1992), and the second best regular pathogen of the lower respiratory tract infections (Jones, 2001). 9 2.4 Common Hospital-Acquire Infections (HAIs) caused by P. aeruginosa 2.4.1 Ventilator-Associated Pneumonia (VAP) An infection which progresses in the lung of patient who is on a ventilator is referred to as ventilator-associated pneumonia (CDC, 2012). P. aeruginosa has a high mortality in causing ventilator-associated pneumonia (Chastre and Fragon, 2002). P. aeruginosa is the most important bacterial source of serious ventilator-related pneumonia and protracted lung contaminations in patients with cystic fibrosis (Goldberg, 2010). The properties of the P. aeruginosa strains connected to VAP and critical respirato ry letdown are very unique from those related to P. aeruginosa strains that resRide Yin the air routes of patients with cystic fibrosis (Sadikot et al., 2005). PulmonAary infections caused by exoU secreting P. aeruginosa are usually more severe. ERxotoxin U toxin secretion in isolates obtained from patients with VAP is a syImBbol for exceedingly contagious P. aeruginosa and is linked with reduced clinica l Loutcome in patients with VAP (Hauser et al., 2002). N 2.4.2 Catheter-Associated Urinary Tract InfeDctioAns (CAUTIs) About 15 to 25% of patients admitted to hAospitals receive urinary catheters during hospitalisation (CDC, 2017). Relative lyI, B75% of nosocomial urinary tract infections (UTIs) are allied with a urinary catheter (CDC, 2017). Lengthy use of urinary catheter is the greatest essential reason whFich predisposes patients to catheter-linked urinary tract infections. Catheter bru isOes the natural barrier and damages the mucosal layer thereby allowing bacIteTriaYl establishment. CAUTI is the major root of nosocomial infection and P. aeruginosa is the third chief pathogen of CAUTIs accounting for 40% of all nosocomRial Sinfections. UTIs due to P. aeruginosa infections are connected with high death Ein hospitalised patients, especially those with underlying diseases (Lamas- FerreiIroV et al., 2017). U2.4N.3 Central line-Associated Bloodstream Infections (CLABSIs) Bloodstream infections (BSI) are infections that result when there is presence of viable bacteria or fungi in the bloodstream (Viscol, 2016). When a central-line catheter was in place at the period of or within 48 hours earlier, the commencement of an infection is referred to as CLABSIs. CLABSI is also known as primary bloodstream infection because infection does not result from other parts of the body. BSIs are a foremost cause of death universally, particularly those connected with patients on admission in intensive care units has fatality rate of 35-50% (Timsit and Laupland, 2012). P. 10 aeruginosa is the third best common Gram-negative organism identified in CLABSI. P. aeruginosa bloodstream infection is connected with great hospital death and is topmost among patients getting improper preliminary antimicrobial therapy (Micek et al., 2005). 2.4.4 Surgical Site /Wound Infections Surgical site infections can occur on the surface of skin or encompass muscles beneath the membrane, structures or inserted material (CDC, 2012). Skin infections such as boils or abscesses which develop at sites other than the operation site is a sign thYat the infection was acquired in the ward. Primary surgical site infections are Rinfections caused by patients‟ normal flora and some other environmental soAurces in the operating theater. It becomes evident within 5–7 days after surgery anRd is usually more serious. The deep-rooted sepsis that occurs within 30 days after IsuBrgery and before the wound is dressed is also an indication of theatre infectio nL. Surgical site or wound infections specifically caused by multidrug-resistanNt P. aeruginosa isolates are connected with amplified illness and death. TDhe aAbility of P. aeruginosa to form biofilms makes them persist in hospital wateAr systems and serve as potential reservoir for Pseudomonas surgical site infection sI (BFalkinham et al., 2015). 2.5 Other infections caused byF Pseudomonas aeruginosa 2.51 Gastrointestinal infectiOons The gastrointestinal tract Y(GI ) is thought-out to be the best essential source of P. aeruginosa. GI tract isI aTlso an essential route of entry in Pseudomonas septicaemia and bacteraemia. The rSate of mortality of patients travailing from bowel colonisation by P. aeruginosa is Rundoubtedly advanced than that of patients lacking establishment by P. aeruginVosaE in critical care unit (Marshall et al., 1993). 2.5N.2 ICentral nervous system (CNS) infections UPseudomonas aeruginosa rarely cause infection in the CNS except in conditions such as postneurosurgical procedures or head trauma to cause meningitis or brain abscesses (Juhi et al., 2009). Previous studies showed that Pseudomonas meningitis which results after surgical operations is always accompanied with pronounced mortality (Huang et al., 2007; Juhi et al., 2009). P. aeruginosa was accountable for 8.3% to 10.7% of meningitis in postneurosurgical patients (Erdem et al., 2008). Postoperative Pseudomonas meningitis is linked with increased mortality (Huang et al., 2007; Juhi et al., 2009). 11 2.5.3 Ocular infections Pseudomonas aeruginosa is a chief cause of bacterial keratitis. Infections of the eye have been reported infrequently in individuals wearing contact lenses for extended period. The most common form of infection is corneal ulcer, which spreads with frightening rapidity to panophthalmitis. The infection often follows penetration of the cornea by a foreign body which becomes embedded there. In a lot of cases, loss of vision in the affected eye ensues or evisceration may be necessary. Most infections of the eye with P. aeruginosa occur as a result of chemical, mechanical or thermal Yinju ry to the eye (Spencer, 1953). R 2.5.4 Endocarditis A Pseudomonas aeruginosa infect the endocardial surface of the heart Rthrough invasion from the blood stream and establish itself. If left untreated, PI. Baeruginosa invasion may result to incurable heart failure and myocardiNal a b Lscesses (Brusch, 2017). Infective endocarditis (IE) caused by P. aeruginosa is also rare accounting for roughly 3% of all patients with infective endocarditisD. AAlmost 90% of patients with P. aeruginosa endocarditis use intravenous druAgs. Most of these reports reveal that infective endocarditis caused by P. aerIugBinosa may be life threatening and has also been linked with high mortality rateF (L in et al., 2016). 2.5.5 Ear infections O Pseudomonas aeruginosa Yis the most principal bacterial pathogen of external otitis and the major cause of earI iTnfections in children. Otitis media is an indigenous infection of the exterior ear paSssage often associated with warm humid conditions after exposure to inadequateEly cRhlorinated hot tubs or swimming pools (Havelaar et al., 1983). 2.5.6 IVSkin infections PsNeudomonas aeruginosa can cause both localised and diffuse skin infections. Skin Uand soft tissue infections with P. aeruginosa result when the normal skin is bridged as in burns wound, surgical wound, trauma or dermatitis. P. aeruginosa is among the most frequent causes of burn wound infection and a main cause of death in burn patients (Lari et al., 1998). 12 2.5.7 Bone and joint infections Pseudomonas aeruginosa accounts for up to 20% of Gram-negative bacteria osteomyelitis infection and is linked with the use of orthopaedic devices and complicates the surgical management of this infection (Rodríguez-Pardo et al., 2014). 2.6 Antibiotics The word antibiotic was devised from the name “antibiosis” which exactly means “against life”. In the previous years, antibiotics were deliberated to be orga nic compounds produced by one microorganism which are lethal to other microorgaYnisms (Russell, 2004). As a result of this conception, an antibiotic was initiallRy, largely defined as a substance, produced by one microorganism, or of biologRical Asource which at low concentrations can inhibit the growth of, or are lethal to other microorganisms (Russell, 2004). On the other hand, this description has been modIiBfied in contemporary times, to comprise antimicrobials that are also manufac tuLred partially or entirely through synthetic means (Etebu and Arikekpar, 2016)N. In September 1928, late Sir Alexander Fleming, an English Bacteriologist byD chaAnce discovered penicillin, the first antibiotic produced by Penicillium notatum, aA soil residing mould (Russell, 2004). The discovery of penicillin has continued toI tBransform the management and battle against bacterial infections (Etebu and AFrik ekpar, 2016). Antibiotic may be expansive-spectrum or constricted-spectrum. Expansive-spectrum antibiotics are active against both Gram-positive and GYram -n Oegative bacteria while constricted-spectrum antibiotics are active on either IGTram-positive or Gram-negative bacteria. Antibiotic could be bactericidal or baScteriostatic in action. Bactericidal antibiotics act by killing the bacteria and Rtheir action is irreversible while bacteriostatic antibiotics work by hindering Ethe growth of bacteria and the action is reversible. Antibacterial agents implicIaVted in the treatment of bacterial infections are considered in accordance with thNeir tool of exploit (Figure 2.2). U 13 Y RA R LI B N DA IB A F O Figure 2.2. MechanisImTs oYf action of antibacterial drugs Source: healthloveS.in/antibiotics/how-do-antibiotics-work-mode-mechanism-of-action- of-antibiotiEcs R IV UN 14 2.7 Beta lactam antibiotics The manner of exploit of beta-lactam agents is closely connected to the arrangement and biosynthesis of the bacterial cell wall (Williamson et al., 1986). Beta-lactam antibiotics are composed of four-membered beta-lactam ring at the centre of their organisation, which is indispensable to their manner of exploit. Beta-lactam antibiotics could be bacteriostatic or bactericidal in action. The bacteriostatic effect of beta-lactam antibiotics ensue as a consequence of inhibition of transpeptidase and carboxypeptidase enzymes implicated in the coming together of the bacterial ceYll w all and consequent inhibition of cell splitting up and development (Lee et al., 20R01). Beta- lactam antibiotics include: A  Penicillins: ampicillin, carbenicillin and amoxicillin R  Cephalosporins: ceftazidime, cefotaxime and cefepime IB  Cephamycins: cefoxitin, cefmetazole and cefotetan L  Monobactams: aztreonam N  Carbapenems: imipenem, meropenem, doDripeAnem, ertapenem, panipenem and biapenem 2.7.1 Carbapenems BAI Carbapenems are potent member Fof beta-lactam antibiotics which are structurally related to penicillins but have extended activity profile against most beta-lactamases for instance the extended spec trOum beta lactamases (Walsh et al., 2005). Carbapenems utilise their exploit by ThinYdering transpeptidase enzyme situated on the exterior of the cytoplasm (HuangS etI al., 1995). Carbapenems which are in clinical use include doripenem, imRipenem, ertapenem, panipenem, meropenem and biapenem (Codjoe and Donkor, 20E18). Broad spectrum of activity is found among meropenem, doripenem, panipeInVem and ertapenem but the effectiveness of panipenem against Gram-negative baNcterial strains is negligible (Codjoe and Donkor, 2018). U2.7.1.1 Imipenem Imipenem was the principal antibiotic in carbapenem family to be accepted for the management of infections initiated by multiple drug resistant bacteria. Imipenem exhibited great attraction for penicillin binding proteins (PBPs) and is stable towards most beta-lactamases. However, both imipenem and panipenem can be easily degraded by dehydropeptidase I (DHP-I) enzyme, present in human renal tubules (Hikida et al., 1992). 15 AR Y R LI B N AD A F I B Y O Figure 2.3. StructuresI oTf beta-lactam antibiotics Source: http://dragSon.klte.hu/~gundat/betalaca.htm R IV E UN 16 Therefore, it requires coadministration with an inhibitor, cilastatin or betamipron (Hikida et al., 1992). 2.7.1.2 Meropenem Meropenem demonstrated increased stability to enzyme dehydropeptidase-1 (DHP-1) and does not require to be administered with dehydropeptidase-1 inhibitor (cilastatin) hence, its being administered without a DHP-1 inhibitor (Zhanel et al., 2007). 2.7.1.3 Doripenem Y Doripenem does not require beta-lactamase inhibitor and has potent activity Rtowards a varied collection of Gram-positive and Gram-negative clinical isolates A(Mushtaq et al., 2004). Relatively, amongst the carbapenems, doripenemB isR steady towards degradation by majority of beta-lactamases and is not inactivIated by human renal dehydropeptidase I as against imipenem (Mushtaq et al., 20 0L4). 2.7.1.4 Ertapenem AN The efficacy of ertapenem against Gram-negativeD non-fermenters is low in comparison with other carbapenems (Papp-Wallace et al., BA 2011). 2.8 Carbapenem resistance I Normally, Gram-negative bacOteriaF are impervious towards many antibiotics than Gram-positive bacteria in lin e with the dissimilarities in their cell wall physiology (Armand-Lefèvre et alT., 2Y013). Bacteria elude the antibacterial action of beta-lactam antibiotics through fouIr mechanisms including: 1. reducedR heSights of drug buildup due to excessive manifestation of efflux- pumEps 2. IVdecreased or loss of the outermembrane porin N3. alteration in the active site of penicillin-binding proteins (PBPs) U 4. production of beta-lactamases such as ampC hyperproduction, metallo beta- lactamase (Taneja et al., 2010). 17 Y AR LIB R AN BA D F I Figure 2.4. Mechanisms of an tiObiotic resistance in bacteria https://www.easynotecardsY.com/print_list/59058 T RS I VE UN I 18 2.8.1 Impermeability of antibiotics mediated by overexpression of efflux pump systems Efflux pumps are carriage proteins that are implicated in the ejection of noxious substances (together with almost entire groups of clinically important antibiotics) from inside cells into the exterior milieu (Bambeke et al., 2000). These proteins are found in very nearly all bacterial species including Gram-positive and Gram-negative bacteria. The genetic material encoding efflux pump proteins can be sited on plasmids or chromosomes. Bacterial efflux pumps are categorised into five major families 1. Small multidrug resistance family (SMR) Y 2. Major facilitator superfamily (MFS) AR 3. Resistance nodulation division family (RND) R 4. Multidrug and toxic compound extrusion family (MATE)I B 5. Adenosine triphosphate-binding cassette superfamily (LABC) With exception of the ABC family, which hydolysAe ANTP to export substrates, other efflux systems use proton motive force as theiDr basis of drive. The RND family is connected widely with clinically importantA bacteria and is present only in Gram- negative bacteria while other families IwBere established in both Gram-negative and Gram-positive bacteria (Handzlik etF al. , 2013). Four efflux pumps have been Oseen as the most clinically relevant out of the twelve efflux pumps of RND famYily identified so far in the genome of P. aeruginosa. These include: MexEF-OprNI,T MexAB-OprM, MexCD-OprJ and MexXY-OprM, and are well distinguished as aSntibiotic carriers (Mesaros et al., 2007). Inherent resistance of P. aeruginosaE to Rmost antibiotics was first believed to occur due to impermeability of the cell mVembrane but is at the present established to consist of interaction of imNpenIetrability with multiple drug resistance efflux, controlled primarily by MexA-UMexB-OprM. Inherent expressions of efflux pumps in „wild-type‟ strains perform a meaningful function in the comparatively reduced susceptibility of P. aeruginosa to antibiotics (Lister, 2002). However, overproduction of these genes in mutant strains leads to high-level resistance to multiple antibiotics (Lister, 2002). Although efflux typically confers a modest degree of resistance, its role in the chemotherapy of infectious diseases could be relevant, because clinical concentration of antibiotic at the target site of action could not be achieved leading to cross-resistance to unrelated 19 antibiotic classes (Lister, 2002). Upregulation of efflux system in P. aeruginosa affects the activity of meropenem more than imipenem (Walsh, 2000). 2.8.2 Down-regulation of outer membrane porin (oprD) The oprD is a specific porin with a unique function in the take-up of positively charged amino acids like arginine, glutamine, histidine (Huang et al., 1995). The oprD function as the main porin for the entrance of carbapenems in P. aeruginosa because it allows the entry of carbapenems in the bacterial cells and its diminished manifestation is usually observed in carbapenem-resistant isolates (Livermore, 1992). DeficiencyY in the regulation of outer membrane porins (oprD) as a result of mutations Rhas been recognised to convene a minor level of resistance to meropenem and dAoripenem but majorly to imipenem (Quale et al., 2006). The lack of a 45-49 kilodRalton (kDa) oprD was described as the peculiarity of imipenem-resistant P. aLeruIg Binosa (Huang et al., 1995). N The oprD protein shared between 41% and 58% simAilarity with members of the porin family and have the overall topography of oAtherD porin. Normal porin comprise a 16-strand transmembrane β-barrel arrangement which composed of 7 slight-turn sequences on the periplasmic side, whi chI Bacts as a link which connect the outer surface of the 8 loop structures to form a tFhin channel, causing reduced penetrability through the outward membrane in P . Oaeruginosa as compared with Escherichia coli ompF porin (Huang et al., 1995).Y 2.8.3 OverexpreSssioIn T of ampC cephalosporinase The ampC beRta-lactamases have its place in class C group of beta lactamases. ChromosomEal ampC cephalosporinase production poses a grave danger to the effectiIvVeness of beta-lactam antibiotics (Lister, 2002). The ampC beta-lactamase was deNtected in 1940 in Escherichia coli as the major beta-lactamase which hydrolyse Upenicillin. The expression of ampC is low in majority of Gram-negative bacilli, but inducible in reaction to contact by beta-lactam and is typically associated with multidrug resistance (Jacoby, 2009). Antibiotics like aztreonam, oxacillin and cloxacillin are able to repress ampC beta-lactamases but beta-lactamase inhibitors and ethylenediamine tetraacetic acid (EDTA) do not have activity against them (Bush et al., 1995). 20 The tool by which these organisms induce ampC expression is complex (Jacobs et al., 1997). The mutation in ampD gene resulting in ampC constitutive hyperproduction is most frequent cause of ampC overexpression in clinical isolates (Schmidtke and Hanson, 2006). When ampC expression increases through mutational events, overexpression of this cephalosporinase provides high-level resistance to all beta- lactams except the carbapenems (Lister, 2002). The ampR modifications are unusual but can also lead to highly-constitutive or hyperinducible phenotypes (Hanson and Sanders, 1999). Y P. aeruginosa PAO1 contains three ampD genes, describing theR gradual overexpression of ampC observed by means of subsequent triggering ofA every single ampD gene (Juan et al., 2006). The presence of numerous loci in aRmpD adds to the virulence ability of P. aeruginosa, which subsequently resultIs Bin a derepressed P. aeruginosa strain following inactivation of one ampD allele wLhile twofold or threefold mutation in ampD gene results in the inability of ampDN mutant to strive in a typical infection with mouse. The regulation of ampC iDn PA. aeruginosa is more multifaceted than the regulation observed in EnterobacterAiaceae (Moya et al., 2008). Beta-lactam antibiotics differ in their inducing po teInBtials and the prompting effect of clavulanic acid is markedly essential for P. aeruginosa. For instance, when ampC expression is induced, clinically reached co ncentr Fations of clavulanic acid have been documented to upset antibacterial activity Yof tic Oarcillin (Lister et al., 1999). 2.8.4 Mechanism oIf Taction and classification of Beta lactamases The chief tool forS resistance to beta-lactam drugs in Gram-negative bacteria is the generation Eof bReta-lactamases. Beta-lactamases are degradative enzymes that dislocate the amiVde connection of the four-membered beta-lactam ring characteristic of beta-lacNtamI antibiotics making the efficacy of beta-lactams useless. Beta-lactamases have Ubeen described numerous times in many bacterial species and may be present on the chromosome or plasmid and repeatedly connected with moveable genomic features such as transposons and integrons (Nijssen et al., 2005). Beta-lactamase produced by different bacteria holds a spectrum of physical, biochemical and practical assets that are unique for particular beta lactam antibiotics just as penicillinases for penicillins, cephalosporinases for cephalosporins, and carbapenemases for carbapenems while some of this beta-lactamases may have affinity for one or more additional groups (Livermore and Brown, 2001). Grouping of beta-lactamases has conventionally 21 remained grounded on either the primary structure or operative features of the enzymes (Bush et al., 1995). The straightforward classification is by protein sequence which is grounded on the amino acid motifs and then grouped into four classes. Class A, C, and D enzymes employ serine for beta-lactam hydrolysis and class B enzymes are those that have need of divalent zinc ions for hydrolysis (Bush and Jacoby, 2010). Carbapenemases have their place in the molecular classes A, B, and D. Production of carbapenemase especially the class B (acquired MBLs) is a key tool through which Gram-negative bacteria confer resistance to carbapenems. Y 2.8.4.1.1 Class A carbapenemases R Class A enzymes have comprehensive spectrum of hydrolysis towardAs most beta- lactams like the penicillins, aztreonam, first group cephalosporIinBs a Rnd carbapenems. Examples of class A enzymes that are existing in the geno mLe of bacteria are NmcA (not metalloenzyme carbapenemase A), SME (SerratiaN marcescens enzyme), IMI-1 (Imipenem-hydrolysing beta-lactamase), SFC-1 (SAerratia fonticola carbapenemase-1) (Codjoe and Donkor, 2018). SME-1 was Doriginally detected in two Serratia marcescens from England in 1982. The AIMI and NMC-A were identified in Enterobacter cloacae from hospital, wiIthB amino acid similarity to 97% and 70% to SME (Nordmann et al., 2011). DNA segment coding IMI-2 beta-lactamases were detected on plasmids in EnterobaFcter species. They are resistant to carbapenem, susceptible to third group of c eOphalosporins and are partially hindered by clavulanic acid (Garcia et al., 20I1T3).Y Plasmid encoded Sclass A carbapenemases include KPC (Klebsiella pneumoniae carbapenemEasRe) (KPC-2 to KPC-13) and GES (Guiana extended spectrum). These membeVrs of class A carbapenemases are also partly hindered by clavulanic acid and hyNdroIlyse carbapenems (Nordmann et al., 2011). KPC is the most dominant class A Ucarbapenemase present mainly on plasmids (Nordmann et al., 2011). There has been report of KPC in other regions after its rapid spread in the United States (Queenan and Bush, 2007). Concomitant with the increasing reports of KPC-2, KPC-3 which is a sole amino acid substitute of KPC-2 was described in New York in a 2000 to 2001 epidemic of K. pneumoniae. The GES and IBC (integron-borne cephalosporinase) groups are not frequently come across and were principally described in 2000. IBC-1 was identified in Greece from E. cloacae while GES-1 was discovered from France in a K. pneumoniae isolate (Poirel et al., 2001). The arrangement of their amino acid 22 confirmed that they are remotely linked to KPC-2, SME-1 and NMC-A with identities of 36%, 35% and 31%, respectively (Poirel et al., 2001). These enzymes were at the start grouped as ESBLs but further categorised to include imipenem hydrolysis when a clinical isolate of P. aeruginosa was identified to contain GES-2 which inactivates imipenem (Poirel et al., 2001). Although GES enzymes are not common, they had been identified worldwide in developed countries and have not been connected with outbreaks. Nevertheless, P. aeruginosa strains carrying GES-2 have been linked to minor nosocomial epidemic in eight patients (Poirel et al., 2002). In Africa, KP Cs were famous to be widespread among Enterobacteriaceae but later discoverRed iYn a P. aeruginosa isolate from a teaching hospital in northwestern, Tanzania ( RA Mushi et al., 2014). B 2.8.4.2 Class D carbapenemase LI The OXA beta-lactamase with carbapenemase fuNncti on was first found in Acinetobacter baumannii in Scotland. These caArbapenemases are of the OXA (oxacillin-hydrolyzing beta-lactamases) enzymeD type. They are poorly repressed by EDTA and clavulanic acid and have weak acAtivity against carbapenems (Codjoe and Donkor, 2018). In the later 1970s an dI Binitial 1980s, OXA was among the utmost dominant plasmid-encoded beta-laFctamase. OXA carbapenemases have ability to rapidly mutate which results iOn large variability in their amino acid sequences with extended spectrum of actYivity (Codjoe and Donkor, 2018). Recurrent discovery of class D carbapenemaIseTs among the clique of Enterobacteriaceae has been reported. OXA-48 was alsoS reported in P. aeruginosa from Tanzania (Mushi et al., 2014). EndemicityE ofR OXA-48 containing Enterobacteriaceae in northern part of Africa has been reVported (Manenzhe et al., 2015). OXA-48, OXA-58, OXA-51-like, OXA-181, OXNA-I163 producing Enterobacteriaceae were reported from Tunisia, Egypt and UMorocco. Non-fermenters are also found to contain class D beta-lactamases in Africa (Manenzhe et al., 2015). Outburst of OXA-23 containing A. baumannii was reported from Tunisia while A. baumannii isolates from Egypt and Algeria have been documented to carry OXA-24-like enzymes. Most remarkable is the report of OXA-48 gene in P. aeruginosa. This enzyme is known to be widespread in Enterobacteriaceae but now described in non-fermenters suggesting horizontal transfer of this gene (Mushi et al., 2014). 23 2.8.4.3 Molecular Class B Carbapenemases (Metallo beta-lactamases) Metallo beta-lactamases (MBLs) have its place in class B of Ambler grouping system. 2+ MBLs are group of beta-lactamase that requires divalent cation usually Zn as a cofactor for the enzyme activity (Queenan and Bush, 2007). They are categorised by 2+ the capability to inactivate carbapenems and are inhibited by EDTA chelators of Zn (Queenan and Bush, 2007). MBLs are of great clinical importance because they are capable of hydrolyzing virtually all beta-lactam antibiotics with the exclusion of aztreonam and most clinically existing inhibitors such as clavulanate, sulbactam a nd tazobactam do not hinder their activity (Bonomo and Szabo, 2006). On the gRrouYnds of amino acid relatedness, ten MBL types and different variants of these Afamilies have been recognised (Pollini et al., 2013). R MBL genes could exist on the chromosome or plasmid mLedIia Bted. Acquired MBL genetic factors are situated on the integron configuraNtions that exist on moveable genomic features like plasmids or transposons, aAnd can rapidly spread to other bacterial species (Walsh et al., 2005). MBLs haDve been identified in countless Gram- negative bacteria and are linked with clonal sApread (Bush et al., 1995). The blaIMP and blaVIM types are the most widely repor teIdB except in Brazil where blaSPM was the most prevalent enzyme (Lucena et al., 2F014). Both blaIMP and blaVIM possess the broadest hydrolysis profile (Queenan aOnd Bush, 2007). P. aeruginosa isolates that produce MBLs are more prevalentY in Brazil and Italy and accounted for 43.9% and 39.1%, respectively (TolemanI Tet al., 2005). The blaVIM-2 has also been described in West and Central Africa (ChSolley et al., 2014). Comprehensive evidence on the predominance of carbapenem reRsistance and molecular characterisation of carbapenem-resistant clinical isolates of EP. aeruginosa from Nigeria are lacking. 2.8.5 IVN TYPES OF MBL U2.8.5.1 Imipenem (IMP) type MBL IMP (active on imipenem) was the first MBL known in imipenem-resistant P. aeruginosa isolate from Japan in 1988 (Watanabe et al., 1991). IMP was present on a transferable conjugative plasmid. After three years, the existence of IMP gene was also established in Serratia marcescens from Japan. MBLs were primarily restricted to South-east Asia, but have extended quickly to Europe, North America, Oceania, including Africa especially after 2000 (Walsh et al., 2005; Manenzhe et al., 2015). 24 2.8.5.2 Veronese Imipenemase (VIM) type MBL Veronese imipenemase (VIM-1) was first recognised in P. aeruginosa in a hospital in Verona, Italy (Lauretti et al., 1999). Later a blaVIM-1 genetic material was found in Achromobacter xyloxidans in similar hospital. The blaVIM-1 gene was also found on gene cassette integrated into class 1 integron (Lauretti et al., 1999). The blaVIM-1 are predominant in Pseudomonas aeruginosa strains but have been noticed in additional Gram-negative bacteria. In 1996, blaVIM-2 closely interconnected to blaVIM-1 was first identified in P. aeruginosa from France (Poirel et al., 2000). The existence of blaVI M-2 has been acknowledged in many countries presenting its circulation uRnivYersally (Mohammed et al., 2015; Zubair and Iregbu, 2018). A novel blaVIM typeA identified in clinical isolates of P. aeruginosa in Japan was recently reported by RHishinuma et al. (2019). This variant was designated blaVIM-60 with two aminoI Bacid substitutions at positions 228 and 252 compared with blaVIM-2 L 2.8.5.3 Sao Paulo Imipenemase (SPM-1) type MBL N A novel gene blaSPM was first identified in P. aerDugiAnosa in Sao Paulo, Brazil in 1997. The blaSPM-1 showed 35.5% identity to IMP-A1 (Toleman et al., 2005). The blaSPM-1 is linked with common region elements bIuBt not integrons or transposons. The blaSPM-1 can inactivate the entire classes of bFeta-lactams but does not hydrolyse clavulanic acid and aztreonam like blaIMP-1 an d OblaVIM-1 (Murphy et al., 2003). 2.8.5.4 Germany ImipeneYmase (GIM-1) type MBL The blaGIM-1 is the fouIrTth MBL enzyme to be identified. It was first detected in 2004 in five identical P. Saeruginosa isolates from a German hospital. The isolates were susceptibleE to Rpolymyxin only (Walsh et al., 2005). The blaGIM-1 showed highest similariVty with IMP-types. Reports have shown that blaGIM do not inactivate azlocillin anNd azItreonam but are able to inactivate beta-lactamase inhibitors as several MBLs do U(Gupta, 2008). 2.8.5.5 Seoul Imipenemase (SIM-1) type MBL The blaSIM-1 was principally discovered in multidrug resistant Acinetobacter baumannii and exhibits 64–69% similarity with the blaIMP and can hydrolyse majorly all beta-lactam antibiotics (Poirel et al., 2000). The blaSIM have not been detected in P. aeruginosa. 25 2.8.5.6 Adelaide Imipenemase (AIM-1) type The blaAIM-1 was discovered in P. aeruginosa from Australia (Yong et al., 2012). The blaAIM-1 is found on two ISCR elements not on integrons as most MBL-genes. The presence of ISCR 10 adjacent to blaAIM-1 showed that ISCR elements are implicated in its mobility (Yong et al., 2012). 2.8.5.7 Kyorin Hospital Imipenemase (KHM) In 1997, blaKHM-1 was identified in clinical isolate of Citrobacter freundii from a patient with catheter-linked urinary tract infection in Japan (Sekiguchi et al., 200Y8). R 2.8.5.8 New Delhi metallo-beta lactamase (NDM) A The blaNDM-1 was majorly recognised in Klebsiella pneumoniae iRn 2008 but later detected in E. coli from the same patient (Yong et al., 2009). TIhBe blaNDM-1 encoding gene is found on diverse big plasmids which also convey g eLnes that confer resistance to practically all antibiotics; hence generating quick diNstribution in clinically related bacteria poses a severe danger for therapy. The blaANDM-1 parts very slight uniqueness with other MBLs. The blaVIM-1/blaVIM-2 is tAhe Dmost identical with 32.4% similarity (Pitout et al., 2005). After the first discBovery of blaNDM-1, infections with blaNDM-1 containing Escherichia coli, Klebsi elIla pneumoniae, Enterobacter cloacae and Acinebacter baumannii were also idFentified in New Delhi. 2.8.5.9 Dutch ImipenemaYse ( D OIM) DIM-1 (Dutch imipeneTmase) was principally recognised in Pseudomonas stutzeri in the Netherlands. DSIMI-1 was remotely interconnected to other class B beta -lactamases, having maximRum amino acid resemblance of 52% with blaGIM-1, 48% with blaKHM-1, 45% with EblaIMP and merely 30% uniqueness with blaVIM. The blaDIM-1, like other MBLsI dVid not hydrolyse aztreonam (Walsh et al., 2005; Sekiguchi et al., 2008). U2.8N.5.10 Florence Imipenemase (FIM) The blaFIM-1 was recognised in a Pseudomonas aeruginosa strain from hospital in Italy (Pollini et al., 2013). The blaFIM-1 enzyme exhibited the uppermost amino acid connection of 40% with blaNDM-type. The blaFIM-1 genetic factor was supposedly present on the chromosome and connected with ISCR19-like features which may be involved in its incorporation and mobilisation (Pollini et al., 2013). Analysis of the kinetic factors displayed that blaFIM-1 has expanded substrate specificity and do not hydrolyse aztreonam like other MBLs (Pollini et al., 2013). 26 2.9 Clinical significance of carbapenem-resistant Gram-negative bacteria The risk of acquiring carbapenem resistant Gram-negative bacteria is greater in the unindustrialised world especially in sub-Saharan Africa where irrational use of antibiotics predominates (Donkor et al., 2012). Carbapenem resistant Gram-negative bacilli are mainly involved in infections in patients with prolonged length of hospital stay or critical care units, extensive antibiotic use, lengthy usage of indwelling therapeutic procedures such as catheter, or immunocompromising conditions. They are mostly associated with healthcare-acquired infections. For instance, blaNDM-produci ng Enterobacteriaceae were identified most commonly among hospitalised pRatieYnts in critical care units and is linked with high mortality (Barguigua et al., 201A2; Gniadek et al., 2016). High death rate in patients carrying isolates that are resistaRnt to carbapenem has been documented (Meradji et al., 2015). The potential for IspBread of carbapenem resistant genes to other bacterial species is worrisome. F oLr instance, blaIMP-4 was discovered firstly in P. aeruginosa from Australia andN later described in Aeromonas junii from this country (Peleg et al., 2006). NosoAcomial outbreaks of carbapenem resistant Gram-negative bacilli have persistAentlDy produced evidence that significant outbreaks are occurring (Gniadek et al., B2016). For instance, a study from Tunisian university hospital described an outbur stI of blaVIM-4-producing K. pneumoniae isolates (Ktari et al., 2006). In additionF, a clonal outbreak of carbapenem-resistant P. aeruginosa strains in hospitali seOd patients in Algeria have also been reported (Meradji et al 2015). Likewise, a reYport from the Netherlands described a nosocomial explosion of blaVIM-2-containingI PT. aeruginosa isolates. Multifocal outbreaks of MBL-containing P. aeruginosa froSm Japan have also been described (Senda et al., 1996). All these reports havEe sRhown that MBLs are disseminated through horizontal transfer of these genes. IV 2.1N0 Mobile genetic elements U2.10.1 Integrons Integrons were first described in the 1980s as mobile genetic element that can capture and incorporate small mobile elements known as gene cassettes, especially gene responsible for antibiotic resistance through site-specific recombination event (Hall and Collis, 1995). The antibiotic resistance genes are located on gene cassettes and can exist as free circular DNA. More than one gene cassette can be inserted into an integron (Hall and Collis, 1995). Integrons consist of an intI gene that is required 27 within the integron for site specific recombination, an attI site that is identified by the integrase and a Pc promoter necessary for gene cassette expression (Boucher et al., 2007). Integrons were majorly clustered into few classes centered on genetic relatedness of intI gene but class 1, 2 and 3 were the most recognised. Later, over 90 different variants of intI gene were identified (Boucher et al., 2007). Integrons may also be present on the chromosome, called chromosomal integrons (CI). Chromosomal integrons are thought to be motionless but movement of gene cassettes has been reported (Cambray et al., 2010). They can carry varying number (from zero to hundreds) of gene cassettes that are not particularly involved in antibiotic RresiYstance (Cambray et al., 2010). A Approximately 10% of bacterial chromosomes consist of intIeBgron Rs that are often connected to MGEs specifically the class 1 integrons (Domi ngLues et al., 2012). Class 1 integrons are the most widespread, basically in medical settings and contribute majorly to widespread antibiotic resistance (Cambray et alA., 2N010). The genesis of Class 1 integrons is not clear, that they were believed to Dbe in existence in bacteria prior to the advent of antibiotics and are implicated in glAobal transmission of antibiotic resistance genetic material (Cambray et al., 2010 ).I CBlass 1 integrons have been reported in both Gram-positive and Gram-negativFe bacteria. In Gram-positive bacteria, class 1 integrons are identified mainl y Ofrom the clinical isolates while class 1 integrons have been isolated from diverseY sources like soil, water, hospitals, animals, food and so on in Gram-negative bactIeTria (Domingues et al., 2012). Typical class 1R intSegrons basically consist of a crucial adjustable area where the gene cassettes arEe placed and two highly conserved regions; the 5′-preserved section (5′-CS) and thIeV 3′-preserved section (3′-CS). The 5′-CS consists of the integrase gene (intI1), thNe recombination site attI1 and the promoter Pc. A growing number of class 1 Uintegrons with structure different from the typical class 1 integrons have also been documented (Toleman et al., 2005). They are associated with insertion sequence common region 1 (ISCR1) and are often referred to as composite class 1 integrons. Majority of class 1 integrons are naturally connected or inserted in mobile genetic elements like Miniature Inverted-repeat Transposable Elements (MITEs), insertion sequences (ISs), transposons (Tns), genomic islands (GIs) and plasmids, thus promoting the transfer of integrons within and between bacterial genome (Domingues 28 et al., 2012). The same class 1 integron identified in different bacterial host has been reported from different geographical locations suggesting horizontal dissemination. 2.10.2 Plasmids Plasmids are extrachromosomal spherical DNA particles that reproduce freely of the bacterial genetic material. Plasmids perform a major role in horizontal spread of genetic material especially the antimicrobial resistance genes and occur in many bacterial strains. One strain can harbor multiple plasmids (Bennett, 2008). A resistan ce plasmid (R-plasmid) is every plasmid that conveys one or additional antYibiotic resistance DNA segment. A R-plasmid may perhaps be a metabolic plaRsmid if it encrypts a degradative role or a virulence plasmid, if it owns virulence geAnetic material (Bennett, 2008). Plasmids may be conjugative or mobilisable. It isR conjugative if it codes for task required for its movement or mobilisable if it relieIsB on conjugative ones to become moveable. Plasmids are the key route for an tiLbiotic resistance spread, including resistance genes encoded by class 1 inteAgroNns (Hu and Zhao, 2009). For instance, the plasmid-mediated transfer of claDss 1 integron from clinical isolates of Serratia marcescens to Escherichia coli thArough conjugation has been reported (Hu and Zhao, 2009). IB 2.10.3 Miniature Inverted-repeatF Transposable Elements (MITEs) MITEs are dependent move aOble genetic elements that consist of small repeat sequences that do not codYe for any protein but are found at different locations in the chromosome of manyI bTacteria. For instance, two copies of a 288 bp MITEs have been detected in a clSass 1 integron present in the plasmid of a clinical isolate of EnterobaEcteRr cloacae (Poirel et al., 2009). 2.10.4I VGenomic islands (GIs) GeNnomic islands are large chromosomal regions which contain genes that encode Udifferent functions. They are referred to as resistance islands when they encode resistance genes. Genomic islands are usually incorporated on the chromosome of many bacteria at specific positions. Genomic islands harbouring class 1 integrons in Gram-negative bacteria have been acknowledged (Douard et al., 2010). 2.10.5 Insertion sequences (ISs) ISs are smallest mobile elements surrounded by terminal inverted repeats which are composed of a middle region that code for a transposase, which accounts for its 29 mobility and are essential for transfer of resistance genes (Domingues et al., 2012). When a DNA fragment is flanked on both sides by the one and the same IS, it is called complex transposon. There have been reports of class 1 integron whose sides are flanked by IS as found in IS6 family that are found in transposons (Doublet et al., 2009). 2.10.6 Transposons Transposons are referred to as jumping genes. Resistance transposons incorporate o ne or more resistance gene inside the element (Bennett, 2004). They have the abiYlity to jump from one plasmid to another or from bacterial chromosome to a plasmRid or vice versa (Bennett, 2008). These mechanisms do not necessarily need correspAonding DNA between the sequence and the positions of attachment but on rare ocRcasion, a specific transposon may have solid preference for a specific nucleIotBide sequence at an attachment position. Some transposons are mobilisable wh ilLe others are conjugative. There is also information on other mobile transposonsN containing class 1 integrons (Petrova et al., 2011). DA 2.12 Molecular typing A The development of multiple drug rIeBsistance in Pseudomonas aeruginosa has strengthened the requirement Ffor epidemiological studies, unfolding their predominance and resistance Oprofile. Numerous numbers of techniques including phenotypic and genotypic Yapproaches have been established for typing P. aeruginosa, the selection of each tIecThnique depend on the capacity to type large number of strains. Such a method sShould also have good discernment with capability to identify a sensible nuEmbRer of types, a method that will not be costly and be reproducible over an extendeVd period (Olive and Bean, 1999). Genotypic methods that have been succesIsfully used to type P. aeruginosa include: Pulsed-field gel electrophoresis, UmNultilocus sequence typing and PCR-based typing among others. 2.12.1 Pulsed-field gel electrophoresis (PFGE) Of all the molecular typing methods, PFGE is frequently regarded as the ‘gold standard’. In this procedure, overnight culture of bacterial cells implanted into melted agarose are lysed in situ and treated with restriction enzymes to obtain an agarose plug comprising whole DNA digested with varying fragment sizes. Electrophoresis is done after the bacterial plug is implanted into electrophoretic tool whose current 30 polarity is changed at programmed period of time. Subsequently, gels are visualised for pattern of arrangement of DNA which is separated according to size after staining with fluorescent dye (Olive and Bean, 1999). This technique permits perfect split-up of bulky DNA fragments of between 10 and 800 kb. Snapped gels can then be analysed with accessible software bundles. PFGE is acknowledged to be better than other methods for molecular typing. It has displayed worthy discriminatory supremacy over repetitive extragenic palindromic polymerase chain reaction (REP-PCR) in distinguishing strains of P. aeruginosa (Grundmann et al., 1995). Distinguishi ng capability of PFGE can be improved through the use of two or more RrestrYiction endonucleases (Olive and Bean, 1999). A The most important rewards of this method are great degrIeeB of R replication and discernment while its greatest disadvantages lie in the us e Lof costly apparatus and timewasting procedure that requires 2-3 days subject toN specific protocols (Olive and Bean, 1999). A 2.12.2 Multi Locus Sequence Typing (MLST) D The procedure of MLST is grounded on the eAxplicit categorisation of bacterial isolates and other organisms in a consistent rep rIoBducible and convenient way using nucleotide sequence (Urwin and Maiden, 2F003). MLST adopted the recognised ideas of multilocus enzyme electrop hOoresis (MLEE) but contrary to relying on the electrophoretic movementY of the loci products, MLST is established on direct sequencing of the lociI pTroducts to define the alleles at each locus (Urwin and Maiden, 2003). MLST clasSsically denotes logical sequencing of five to six greatly-preserved housekeepiEng Rgenes within the bacterial genomic DNA. Alteration in the nucleotide sequencVe of every locus is categorised and documented and a sequence type or anNcestIry is assigned by comparing the sequence obtained with sequence of other Uisolates in the databank (Maiden, 2006). Many typing systems have been suggested for typing P. aeruginosa with PFGE being the standard against which other techniques are justified. The inordinate need for clear-cut and reproducible worldwide procedure for typing P. aeruginosa led to the development of MLST system for P. aeruginosa typing and has provided improved information on epidemiology of P. aeruginosa (Curran et al., 2004). A lot of studies have proven that MLST satisfies these standards (Curran et al., 2004). 31 MLST has lot of benefits above other techniques of typing in the sense that it can identify variations in nucleotide sequence, that is, at the level of DNA that are not obvious by phenotypic techniques. Moreover, the method is reproducible and did not necessitate the use of living bacterial cells or standard-quality genomic DNA, thereby preventing problems accompanying transportation and handling of pathogens. Records obtained through MLST can be accessed in research laboratories all over the world via internet (Maiden, 2006). 2.12.3 Repetitive element sequence-based polymerase chain reaction (rep-PCRY) Repetitive element sequence-based polymerase chain reaction (rep-PCR) iRs a typing technique that permits the creation of DNA fingerprinting. It involveAs the use of definite primers for PCR magnification of scattered repetitive DNA elRements present at different positions in prokaryotic genomic constituent (Hie tLt aInd B Seal, 2009). These elements seem to be well preserved amongst various bacterial species (Versalovic et al., 1991). The rep-PCR method involves the use of isolNated DNA from bacteria or the use of whole cells in PCR amplification of eithDer aA single primer or combination of primers succeeded by gel electrophoresis aftAer staining with ethidium bromide. When the amplified DNA fragments are par teIdB by electrophoresis, a genomic pattern that can be employed for strain represenFtation and discernment is obtained (Hiett and Seal, 2009). O Repetitive extragenic palYindromic PCR (REP-PCR) and enterobacterial repetitive intragenic consensus IseTquence (ERIC-PCR) are the frequently used repetitive element. REP comprises of Ssix degenerate locations with 38 bp arrangements and flanked by a 5 bp variabEle lRoop on each side of a preserved palindromic structure (Versalovic et al., 1991). VREP is best described among the repetitive sequences in bacterial genome and arNoundI 500-1000 replicas of these elements are found in the genome of E. coli and USalmonella typhimurium (Hiett and Seal, 2009). ERIC is the second PCR-based typing method that has been used effectively for differentiating bacterial strains. ERIC sequences are situated in the extragenic areas of the bacterial genome and composed of 126 bp elements of exceedingly preserved indispensable reversed replication (Sharples and Lloyd, 1990). Several studies have successfully typed P. aeruginosa with REP- and ERIC-PCR; while ERIC-PCR has been reported to show equal discriminatory potential with PFGE and MLST (Kidd et al., 2011; Faridi and Javidpour, 2015). Both REP and ERIC sequences have been described in many 32 bacteria. REP patterns are extra challenging than ERIC patterns; nonetheless the combination of both REP-PCR and ERIC-PCR has proven to augment the discriminatory control as compared to when the methods are used independently (Olive and Bean, 1999). While REP- and ERIC-PCR are the best recurrently used targets for DNA typing, BOX elements was primarily believed to be exclusive to Streptococcus pneumoniae but is at present established in many bacterial species as with REP and ERIC elements (Kw on et al., 1998). Sequences of BOX elements are not in any way related to the twYo rep- PCR methods above. Wolska and coworkers (2012) demonstrated that BOXR-PCR is a pronounced investigational strategy useful for categorising P. aerugiRnosaA because it is quick, extremely replicated and with good ability to distinguish bacterial strains. Rep- PCR is extremely relevant and largely more appropriate thaLn gIeBnomic fingerprinting and plasmid analysis and has shown worthy connectio n with PFGE with less differentiating ability (Liu and Wu, 1997). AN 2.12.4 Restriction Fragment Length PolymorphDism (RFLP) This method is centered on the belief that vaArious restriction enzyme recognition sites in a certain genetic region of interest m aIyB possibly vary and be located at diverse sites in the chromosome of different bFacteria which invariably give different banding arrangement. In this method , Ochromosomal DNA is broken down with restriction enzyme and then analysedY on agarose gel through electrophoresis, the gel is observed for discrepancy in rIesTtriction pattern which signifies strain disparity. RFLP may involve the transfSer of electrophoretic products to any of nylon or nitrocellulose membrane EusiRng a probe that is homologous to the target site in a practice called SoutherVn blotting or by amplifying the chromosome with a primer specific for the gene of intIerest and subjected to gel electrophoresis to obtain a banding arrangement UpaNrticular for a certain strain in the process named PCR-based RFLP (Olive and Bean, 1999). 2.12.5 Randomly amplified polymorphic DNA (RAPD) This method depends on the use of short primer of between 6-10 base pair of random sequence that hybridise to DNA at low annealing temperature. The quantity and position of these arbitrary primer sites differ for diverse strains of a bacterial species. A PCR product with size equivalent to the distance amidst the two primers will form if 33 the two primers anneal within limited kilobases of one another. Therefore, subsequent segregation of PCR products on agarose gel produces a band arrangement which is perculiar to specific bacterial strain (Olive and Bean, 1999). RAPD is able to differentiate strains than RFLP but with not as much of discrimination as rep-PCR. In RFLP, since primer is not directed towards a precise locus, many of the annealing procedures result from defective alignment between the target position and the primer. Also, amplification procedure is exceedingly sensitive to minor variation in the annealing temperature which brings about inconsistency in banding arrangemenYts. RA R LIB AN AD F I B O SI TY VE R I UN 34 CHAPTER THREE MATERIALS AND METHODS 3.1 Materials 3.1.1 Equipment and glassware All equipment and glassware used in this study are listed below: Y NanoDrop® spectrophotometer (Thermo Scientific), Hot air oven A(GRallenkamp England), Autoclave (Healthquip England), Incubator (GallenRkamp England), Centrifuge (Merck), Microscope (Olympus England), Q125 SIoBnicator (QSONICA), Weighing balance (Ohaus London), Vortex mixer (Gr ifLfin & George Britain), o o Refrigerator (-20 C and -80 C), Digital water bath (TNhomas Scientific), Water bath shaker (Bionics Scientific), Ice machine (fisher AScientific), Thermal cycler (AB Applied Biosystem), Step One® real time polAymDerase chain reaction (RT-PCR) system (AB Applied Biosystem), Gel DocumentaBtion and Analysis System (GenoSens 1560), Gel casting trays and combs, Electroph oIresis tank and Electrophoresis Power supply. Glassware F Beakers, Conical flasks, MYea su Oring cylinder, Test tube, Petri dish, Inoculating loop, Bunsen burner and MIicTropipettes 3.1.2 Media, BuSffers, Chemicals All media, EantiRbiotics, buffers and chemicals used in this study are listed in Appendix I 3.1.3 IVDNA isolation and purification kits, enzymes, molecular weight markers N and Master mix UDNA isolation and purifica tion kits, enzymes, molecular weight markers and Master mix are listed in Appendix II. 3.1.4 List of primers Primer sequences used for polymerase chain reaction (PCR) and reverse transcriptase quantitative polymerase chain reaction (RT-qPCR) and sequencing are listed in Table 3.1 35 3.1.5 Study design and place of study Study design: This work is a cross-sectional study Place of study: University Teaching Hospitals and Federal Medical Centers in five states, southwest Nigeria (Ekiti, Ogun, Ondo, Osun and Oyo states). These included Obafemi Awolowo University Teaching Hospital Complex, Ile-Ife, Osun state (OTHI), University College Hospital, Ibadan, Oyo state (UCH), Ladoke Akintola University Teaching Hospital, Osogbo, Osun state (LTHO), Federal Medical Centre, Abeokuta, Ogun state (FMCA), Olabisi Onabanjo University Teaching Hospital (OTHS) Sagam u, Ogun state, Federal Medical Centre, Owo, Ondo state (FMCO), FederaRl MYedical Centre, Ido-Ekiti, Ekiti state (FMCI).. A 3.1.6 Bacterial isolates R Four hundred and forty-seven (447) P. aeruginosa isolates recIoBvered were collected from Microbiology units of seven tertiary hospitals from LSeptember 2014 to June 2017. The isolates were obtained from urine, pus, catheNter tips, wound swabs, wound biopsy, sputum, eye swab, cerebrospinal fluid (ACSF), tracheal aspirate, genital discharge, blood and sputum. D IB A O F SI TY VE R NIU 36 Y R Table 3.1. Primers used for molecular screening in this study A Primer Sequence (5ʹ to 3ʹ) Gene Amplicon size, bp RReference OprI-F ATGAACAACGTTCTGAAATTCTCTGCT oprI 249 De Vos et al., 1992 OprI-R CTTGCGGCTGGCTTTTTCCAG IB OprL-F ATGGAAATGCTGAAATTCGGC oprL 504 L Lim et al., 1997 OprL-R CTTCTTCAGCTCGACGCGACG SME-F ACTTTGATGGGAGGATTGGC blaSME 551 N Hong et al., 2012 SME-R ACGAATTCGAGCATCACCAG A GES-F GCTTCATTCACGCACTATT blaGES D323 Hong et al., 2012 GES-R CGATGCTAGAAACCGCTC NMC-A F TGCGGTCGATTGGAGATAAA IBbla A NMC-A 399 Hong et al., 2012 NMC-A R CGATTCTTGAAGCTTCTGCG BIC-1 F TATGCAGCTCCTTTAAGGGC blaBIC-1 537 Hong et al., 2012 BIC-1 R TCATTGGCGGTGCCGTACAC F IMP-F GGAATAGAGTGGCTTAAYTCTC O blaIMP 188 Ellington et al., 2007 IMP-R CCAAACYACTASGTTATCT VIM-F GATGGTGTTTGGTCGCIATTAY blaVIM 390 Ellington et al., 2007 VIM-R CGAATGCGCAGCACCAG SPM-F AAAATCTGGGTACGCAAACG blaSPM 271 Ellington et al., 2007 SPM-R ACATTATCCRGCTGSGAACAGG GIM-F TCGACAECACCTTGGTCTGAA blaGIM 477 Ellington et al., 2007 V I 37 UN Y R Table 3.1. Primers used for molecular screening in this study (cont’d) A Primer Sequence (5ʹ to 3ʹ) Gene Amplicon size, bp Reference GIM-R AACTTCCAACTTTGCCATGC R SIM-F TACAAGGGATTCGGCATCG blaSIM 570 ELllinIgtBon et al., 2007 SIM-R TAATGGCCTGTTCCCATGTG AIM-F TGAGAAATGGCTACGCACTG blaAIM 1241 Yong et al., 2007 AIM-R GTACGGAAAACTCAGCACCC N DIM-F GCTTGTCTTCGCTTGCTAACG blaDIM 699 A Poirel et al., 2011 DIM-R CGTTCGGCTGGATTGATTTG D NDM-F GGGCAGTCGCTTCCAAGGT blaNDM A475 Deshpande et al., 2010 NDM-R GTAGTGCTCAGTGTCGGCAT OXA-48A TTGGTGGCATCGATTATCGG blaIOBXA 744 Poirel et al., 2012 OXA-48B GAGCACTTCTTTTGTGATGGC F OXA-58A CGATCAGAATGTTCAAGCGC O blaOXA 529 Poirel and Nordmann, 2006 OXA-58B ACGATTCTCCCCTCTGCGC ampC-PA-A CTTCCACACTGCTGTTCGCC ampC 1063 Rodriguez-Martinez et al., 2009 ampC-PA-B TTGGCCAGGATCACICATGTYCC OprD –F GCTCGACCTCGAGGCAGGCCA oprD 242 Gutiérrez et al., 2007 OprD-R CCAGCGATTGGSTCGGATGCCA MexB-F CAAGGGCGRTCGGTGACTTCCAG mexB 272 El Amin et al., 2005 MexB-R ACCTGEGGAACCGTCGGGATTGA IV 38 UN RY Table 3.1. Primers used for molecular screening in this study (cont’d) A Primer Sequence (5ʹ to 3ʹ) Gene Amplicon Reference size, bp R MexY-F GGACCACGCCGAAACCGAACG mexY 522 El ALminI etB al., 2005 MexY-R CGCCGCAACTGACCCGCTACA MexD-F GGACGGCTCGCTGGTCCGGCT mexD 236 NRo driguez-Martinez et al., 2009 MexD-R CGACGAAGCGCGAGGTGTCGT MexF-F CGCCTGGTCACCGAGGAAGAGT mexF 255 A El Amin et al., 2005 MexF-R CGCCTGGTCACCGAGGAAGAGT D rpsL-F GCAAGCGCATGGTCGACAAGA rpsL A201 Dumas et al., 2006 rpsL-R CGCTGTGCTCTTGCAGGTTGTGA MexA-F GGCGACAACGCGGCGAAGG mIeBxA 203 Tomás et al., 2010 MexA-R CCTTCTGCTTGACGCCTTCCTGC MexC-F GCAATAGGAAGGATCGGGGCGTTGOG F mexC 102 Tomás et al., 2010 MexC-R CCTCCACCGGCAACACCATTTCG MexE-F TCATCCCACTTCTCCTGGCGCTAC C mexE 150 Tomás et al., 2010 MexE-R CGTCCCACTCGTTCAGCIGGTTYGTTCGATG MexX-F AATCGAGGGACACCCATTGCACATCC mexX 82 Tomás et al., 2010 MexX-R CCCAGCAGGAATASGGGCGACCAG ER NI V 39 U RY A R Table 3.1. Primers used for molecular screening in this study (cont’d) IB Primer Sequence (5ʹ to 3ʹ) Gene Amplicon Lsize, Reference bp hep35 TGCGGGTYAARGATBTKGATTT int 1, 2 3 A491 N White et al., 2000 hep36 CARCACATGCGTRTARAT hep58 TCATGGCTTGTTATGACTGT intI cassette DVariable White et al., 2001 hep59 GTAGGGCTTATTATGCACGC ExoS MP5, GCGAGGTCAGCAGAGTATCG exoSI B A 118 Ajayi et al., 2003 MP3, TTCGGCGTCACTGTGGATGC ExoT MP5, AATCGCCGTCCAACTGCATGCG ex oT 152 Ajayi et al., 2003 MP3, TGTTCGCCGAGGTACTGCTC F ExoU MP5, CCGTTGTGGTGCCGTTGAA GO exoU 134 Ajayi et al., 2003 MP3, CCAGATGTTCACCGACYTCGC ExoY MP5, CGGATTCTATGGICATGGGAGG exoY 289 Ajayi et al., 2003 MP3, GCCCTTGATGCACTCGACCA where B = C or G or T; K = G or TS; R = A or G; Y = C or T; D = A or G or T VE R I 40 UN Table 3.2. Primers used for typing in this study Primer Sequence (5ʹ to 3ʹ) Reference REP1R-I IIIICGICGICATCIGGC Versalovic et al., 1991 REP2-I ICGICTTATCIGGCCTAC ERIC1R ATGTAAGCTCCTGGGGATTCAC Versalovic et al., 1991 ERIC2 AAGTAAGTGACTGGGTGAGCG BOX-A1R CTACGGCAAGGCGACGCTGAC Versalovic et al., Y199 4 RA R LI B N AD A B F I O SI TY ER IV UN 41 3.2 Methods 3.2.1 Phenotypic identification of the study isolates Pseudomonas aeruginosa was identified phenotypically by Gram staining, growth on cetrimide agar, motility, pigment production, oxidase reaction, gelatin liquefaction, catalase production, citrate utilisation, urea hydrolysis, oxidative utilisation of sugars. 3.2.1.1. Cultural identification of isolates Clinical isolates collected on agar slants were inoculated on Pseudomonas cetrim ide o agar (Oxoid) and incubated at 37 C for 24 - 48 hours to observe for growtYh and pigment production. AR 3.2.1.2 Gram’s staining R A smear of the cell culture was made on a clean grease-free micrIoBscope slide with a bit of colony from 18 - 24 hour old culture grown on Pse uLdomonas cetrimide agar (Oxoid). The smear was air dried and passed through theN Bunsen burner flame thrice to heat fix the smear. Crystal violet (primary stain) wasA applied on the smear and left for 60 seconds. The smear was washed under thDe tap for 3 seconds to remove the remaining unbound crystal violet. Lugol‟s ioAdine used as a mordant, was added to the smear to fix the crystal violet, for 60 se cIonBds. The slides were gently washed under the tap, rinsed with alcohol for 3 secondFs and then rinsed again with water. Carbol fuschin (secondary stain) was added t oO the slide for 30 seconds and washed with water for 5 seconds, allowed to dry anYd then observed under oil immersion x100 objective lens of light microscope. ThIe TGram reaction that showed red rod-shaped cells under the microscope was iSndicative of P. aeruginosa. P. aeruginosa ATCC 27853 served as control straEin. R 3.2.1.3I VOxidase test OxNidase test strip (Oxoid™) was put on a sterile Petri dish with the use of sterile pair Uof forceps, 3 - 4 well isolated colonies of bacteria from 18 hour culture on Pseudomonas cetrimide agar were smeared onto the strip with sterile cotton swab. Presence of a blue or purple colour within 30 seconds indicates a positive test characteristic of P. aeruginosa. P. aeruginosa ATCC 27853 served as control strain. 42 3.2.1.4 Gelatin liquefaction test A heavy inoculum of 18-24 hour old culture of each isolate was stab inoculated on the tube containing nutrient gelatin medium. The tubes were inoculated at 37°C for 24 - 48 hours along with uninoculated control tube. Thereafter, the tubes were placed in refrigerator for 1 hr and then examined. Liquefaction of the test while the control remains solidified showed a positive test, characteristic of P. aeruginosa. P. aeruginosa ATCC 27853 served as control strain. 3.2.1.5 Catalase test Y Well isolated colony of the test organism from overnight culture is placed onRto a clean microscope slide. Two drops of 3% hydrogen peroxide (H A2O2) solution was placed separately on a slide, a colony of the test isolate was placed on oInBe o Rf the H2O2 smear. Immediate bubbles observed by rapid evolution of oxyge nL (O2) indicate a positive result. P. aeruginosa ATCC 27853 served as control straiNn. 2H2O2 2H2DO + AO2 3.2.1.6 Urease test A Overnight cultures of the isolates wer eI oBbtained in 5mL nutrient broth and used to inoculate Motility Indole Urea (MIUF) medium. With an inoculating wire, the centre of the tube was stabbed to about oOne-half its length. The tubes were incubated aerobically o with caps loosened at 37 CY fo r 24 hours. Intense pink-red colour indicates a positive result while colour chIaTnge to pale yellow indicates a negative result. P. aeruginosa ATCC 27853 serveSd as control strain. R 3.2.1.7 HydErogen sulphide production SulphIidVe indole motility medium (SIM) tubes were stab inoculated with overnight o cuNlture of the test isolates. Tubes were incubated at 37 C for 24 hours. Hydrogen Usulphide will react with sodium thiosulphate and ferric ammonium citrate in the medium if hydrogen sulphide is present, to give ferrous sulphide which turns the medium into a black colour. The presence of hydrogen sulphide means that the bacteria produce the enzyme cysteine in the medium to give hydrogen sulphide. P. aeruginosa does not produce hydrogen sulphide. P. aeruginosa ATCC 27853 served as control strain. 43 3.2.1.8 Citrate test Overnight broth culture of the test organism was obtained in 5ml nutrient broth. The overnight culture was used to inoculate sterile 5mL Koser‟s citrate broth. The o inoculated tubes and an uninoculated control tube were incubated at 37 C for 72 hours. The medium was later examined for colour change. Colour change from green to blue specifies a positive test whereas no colour change specifies a negative result. P. aeruginosa ATCC 27853 served as control strain. Y 3.2.1.9 Oxidative utilisation of sugars Nutrient broths containing each of glucose, sucrose or mannitol (AppendixR II) were o inoculated with overnight culture of the test isolates, followed by incubAation at 37 C for 24-72 hours. If acid is produced, colour changes from red to yBelloRw, while bubbles in the inverted Durham tube indicate production of gas. P. LaerIuginosa ATCC 27853 served as control strain. N 3.2.2 Antimicrobial susceptibility testing (AST) A 3.2.2.1 Disc diffusion AD Antimicrobial susceptibility testing was IdBone to determine the antibiotic susceptibility profile of the isolates to various antips eudomonad antibiotics using Kirby-Bauer disc diffusion procedure. OvernighOt nuFtrient broth culture of the test bacterium was 8calibrated to achieve McFYarl and turbidity standard of 0.5, equivalent to 1.5 x 10 cells/mL. Then, 0.1 mILT of the suspension was added to 20 mL molten Mueller Hinton oagar and cooled toS about 45 C. The mixture was swirled gently and then poured into sterile Petri-dishes. The plates were left on the bench for 15 minutes and dried for 20 o minutes atE 37RC in an oven. Antibiotic discs were positioned on the surface of the plates. VThe plates were left on the bench for 1 hour preincubation period and then o inNcubaIted at 37 C for 24 hours. Thereafter, zones of growth inhibition were then Udetermined and reported in millimeter (mm). The result was interpreted as resistant, sensitive or intermediate according to CLSI (2015). P. aeruginosa ATCC 27853 served as control strain. The following antibiotics were tested against the isolates: amoxicillin clavunate (30 µg), piperacillin (100 µg), ticarcillin (75 µg, piperacillin tazobactam (100/10 µg), cephalothin (30 µg), cefuroxime (30 µg), ceftazidime (30 µg), ceftriaxone (30 µg), cefoperazone (30 µg), cefepime (30 µg), imipenem (10 µg), meropenem (10 µg), 44 doripenem (10 µg), aztreonam (30 µg), amikacin (30 µg), gentamicin (10 µg), tobramycin (30 µg), ciprofloxacin (5 µg), levofloxacin (10 µg), ofloxacin (5 µg), colistin sulphate (10 µg), polymyxin B (300 unit). MDR was defined as resistance to at least one antimicrobial agent in three or more antimicrobial classes listed while XDR was defined as non-susceptible to at least one agent in all but two antimicrobial classes listed (Magiorakos et al., 2012). Antimicrobial classes include: 1. Aminoglycosides: Gentamicin, Tobramycin, Amikacin Y 2. Antipseudomonal cephalosporins: Ceftazidime, Cefepime R 3. Antipseudomonal fluoroquinolones: Levofloxacin, Ciprofloxacin A 4. Antipseudomonal carbapenems: Meropenem, Imipenem anBd DRoripenem 5. Antipseudomonal penicillins plus beta-lactamas eL iInhibitors: Ticarcillin clavulanic acid, Piperacillin-tazobactam 6. Monobactams: Aztreonam N 7. Polymyxins: Colistin, Polymyxin B DA 3.2.2.2 Determination of minimum inhIibBito Ary concentrations (MICs) This was carried out to look at the bac teriostatic activity of selected antibiotics on the test isolates. Minimum inhibitory cFoncentrations of selected antibiotics against the P. aeruginosa were determined b yO broth microdilution (CLSI, 2015). In a 96 well plate, 100 TL Yof double-strength Mueller-Hinton broth was introduced into each well with thSe aIid of multi-channel pipette followed by 50 L of x4 strength antibiotic dilutRions (Appendix I) and 50 L of the test organism suspension at a 6 concentVratiEon of 2x10 /mL. The contents of each well was mixed and incubated at 37C for 24I hours. Two control wells one without antibiotic and the other without test organism weNre set up. Standard strain (P. aeruginosa ATCC 27853) was run with the test to Ucheck the reagents and conditions. The lowest concentration showing inhibition of growth assessed by lack of turbidity in the well was considered the MIC of the organism. P. aeruginosa ATCC 27853 served as control strain. 3.2.3.1 Detection of beta-lactamase production Beta-lactamase detection in clinical strains of P. aeruginosa was performed with Iodometric cell suspension method as used by Adeleke and Odelola (2000). Bacterial 45 colonies were emulsified into a 0.5 mL of freshly prepared Penicillin G phosphate buffer to form suspension of McFarland standard no 6 and the suspension was homogenised. o The tubes were incubated for 1 hour at 28 C (room temperature). Then, 2 drops of newly processed starch suspension (1%) were introduced into the mixture and shaken smoothly. One drop of iodine solution was introduced gently into the mix and allowed to stay at o 28 C for 10 minutes to observe for colour conversion from blue-black to white if beta- lactamase is produced. P. aeruginosa ATCC 27853 and Escherichia coli ATCC 36318 served as controls. Y 3.2.3.2 Screening of pathogen for carbapenemase production R Well isolated colonies of E. coli ATCC 25922 from overnight culture Aemulsified in normal saline (0.5 McFarland standard) was swabbed on the surface oRf Mueller Hinton agar plate. Meropenem (10 µg) disc was placed at the centre IofB the Mueller Hinton agar plate aseptically. Well isolated colonies of the test oLrganism from overnight culture emulsified in normal saline (0.5 McFarland turNbidity standard) was smoothly lined from the end of the meropenem disc to thDe enAd of the culture plate. The plates were kept warm for 24 hrs at 37ºC. AfterwaArds, the plates were inspected for spread- out at the point of intersection of the t esIt Borganism and Escherichia coli ATCC 25922 within the inhibition region of theF meropenem antibiotic disc. P. aeruginosa ATCC 27853 served as control strain ( WOalsh et al., 2011). 3.2.3.3 Combined discT teYst for phenotypic detection of metallo-beta-lactamases (MBLs) I Test organism RequSivalent to 0.5 McFarland turbidity standards was inoculated on plate of Muller EHinton agar. Two imipenem (10 µg) and meropenem (10 µg) discs were positioInVed on inoculated plates and 5 µL of 0.5M EDTA solution was added to one of eaNch imipenem and meropenem disc. After 18-24 hour incubation at 37ºC, the region Uthat does not permit the growth of the test organism all over the single imipenem and meropenem discs and the one having EDTA was noted and compared. An escalation in size of at least 7 mm at the area surrounding the imipenem-EDTA disc and meropenem-EDTA discs where no growth was observed reflected a positive result. P. aeruginosa ATCC 27853 served as control strain (Walsh et al., 2005). 46 3.2.3.4 Curing of antibiotic resistance The genetic basis of resistance to carbapenems in selected carbapenem-resistant P. aeruginosa was investigated by subjecting them to plasmid curing experiment. This was done to determine the location of the resistant marker in the bacteria whether it is chromosomal or plasmid mediated. Curing was carried out according to the method described by Obaseiki-Ebor (1988) with additional adjustment in the use of more than one concentration of ethidium bromide. Ten carbapenem-resistant P. aeruginosa strains were randomly selected. Overnight culture of each resistant strain was obtain ed in 5 mL nutrient broth containing 12.5, 25, 50 and 100 μg/mL of ethidium bromiYde and o -2 R incubated at 37 C for 24 hours. Then, 10 dilution of mutagen-expoAsed overnight o culture was inoculated on Mueller Hinton agar and incubated at 37RC for 24 hours. Afterwards, four colonies were selected from each of the IpBlates which showed dispersed colonies. Each colony was incubated similarly Lafter inocution in 5 mL nutrient broth and diluted to obtain 0.5 McFarland standNard. Thereafter, 0.1 mL of the inoculum was introduced into 20 mL molten MuelleAr Hinton agar, whirled to mix, and then transferred into sterile Petri dish and allAoweDd to set. A cork-borer of 8 mm width was used to bore wells in the set agar medium; wells were then occupied with different concentrations of the imipenem and mIeBropenem. After a pre-incubation circulation o period of 2 hours, plates were incuFbated at 37 C for 24 hours, and then observed for zones of growth inhibition. OPresence of inhibition zone on agar medium was suggestive of plasmid-medYiated resistance. T 3.2.4. MolecularS mIethods 3.2.4.1 IsolatioRn of plasmid DNA The WizPrEep™ Plasmid DNA Miniprep purification Kit (Wizbiosolutions) was used for plaIsVmid DNA extraction in accordance with Manufacturer‟s instruction. Bacterial ceNlls grown overnight in Luria Bertani broth were transferred into 1.5 mL spin tube Uand spinned for 60 seconds at 13000 rpm. The supernatant was gently discarded. Two hundred microliters of PD1 buffer containing RNase A was introduced into the Eppendorf tube and vortexed to resuspend the cell pellet. Then, 200 µL of PD2 Buffer was introduced into the suspension and slightly turned upside down for 10 times. The tube was allowed to stand on the table for 2 minutes or until the lysate was mixed. The mixture was not vortexed o avoid shearing the plasmid DNA. Three hundred microliter of PD3 Buffer was then introduced into the suspension and immediately mixed by 47 inverting the tubes ten times and spinned for 10 minutes at 13000 rpm. Spin Column (SC) was retained in a 2 mL Collation Tube. The supernatant was poured into the SC and spinned for 1 minute at 13000 rpm. The tube containing clear liquid was discarded and the SC was placed back into the 2 mL Collation Tube. At this point, 600 µL of Wash Buffer containing ethanol was put into the DNA SC and spinned for 1 minute at 13,000 rpm. The tube holding clear liquid was discarded and the DNA SC was placed back in the 2 mL Collation Tube. Centrifugation process was repeated for 2 minutes at 13000 rpm to make sure that the matrix is totally dried. The dried SC was transferr ed to a new micro tube. Then, 50 μL of elution buffer was introduced to the cenRtreY of the column matrix and allowed to stand for 3 minutes and then spinned foAr 1 minute at 13,000 rpm to elute the DNA. BR 3.2.4.2 Extraction of genomic DNA I For routine isolation of genomic DNA from P. aeruginosa, WLizPrep™ genomic DNA purification kit was used in accordance with ManufactuNrer‟s instruction. Cells grown 9 overnight in 5 mL LB broth (<2x10 ) were spDinneAd for 10 min at 7,500 rpm. The supernatant was carefully separated fromA the pellet and bacterial pellet was resuspended in 180 μL of GT1 Buffe rI. BTwo hundred microliters (200 μL) of GT2 Buffer and 20 μL of Proteinase K wFas introduced into the solution and vortexed. The mixture was incubated at 56 ℃O for 10 minutes and the tube was inverted every 5 minutes. Two hundred miYcroliters of 100% ethanol was added to the sample lysates and mixed by vortexiIngT briefly. The suspension was poured to the Spin Column (SC) and spinned for 1S min. at 13,000 rpm. The tube containing clear liquid was thrown away and re-joRined with the SC. Five hundred microliters (500 μL) of W1 Buffer was introduced Einto the SC and centrifuged for 1min. at 13,000 rpm. The flow-through was throwIn-Vaway and rejoined with the SC. Seven hundred microliter (700 μL) of W2 BuNffer (ethanol added) was added in the Spin Column and spinned for 1 min. at 13,000 Urpm. The tube containing clear liquid was discarded and rejoined with the SC and spinned for 2 min. at 13,000 rpm. The SC was connected to a new 1.5 mL tube; 100μL of Elution Buffer was introduced into the middle of the SC and let stand on the table for 1 min. and centrifuged for 1 min. at 13,000 rpm. The SC was discontinued and eluted DNA was kept at –20°C for a few days or –70°C for extended period of storage. 48 3.2.4.3 Extraction of total RNA The PureLink TM total RNA purification kit (Invitrogen) was used to extract total RNA from carbapenem-resistant Pseudomonas aeruginosa (CRPA) in accordance with Manufacturer‟s instruction. The bacterial cells were harvested from 5 mL volume of mid-log-phase culture in LB medium and collected into 1.5 mL eppendorf tube by o spinning at 5000 x g for 5 minutes at 4 C and the supernatant was discarded. Then, 10 mg/mL of lysozyme was introduced into TE buffer (10mM Tris-HCl + 0.1mM EDTA, pH 8.0), followed by addition of 0.5 µL of 10% sodium dodecyl sulphate. T he suspension was thoroughly mixed and incubated for 5 seconds (room temRperYature). Then, 350 µL of RNA Lysis solution prepared by adding 1% (v/v) 2-meArcaptoethanol was introduced into the mixture and vortexed vigorously. The lysate wRas transferred to a 15 mL round-bottomed tube and homogenised using a rotor-IsBtator homogeniser at maximum speed for at least 45 seconds and spinned at room tLemperature for 5 minutes at 2,600 × g. The supernatant was decanted into a microNcentrifuge tube free of RNase. To the lysate was added 250 µL of 100% ethanol andA vortexed thoroughly. D The homogenised lysate was poured into tAhe RNA spin-catridge pre-inserted in a collection tube and then spinned at roo mI Btemperature for 15 seconds at 12,000 × g. To a clean, RNase-free microcentrifuge tube, 70 µL of DNase buffer plus 10 µL of DNase 1 (1 unit/µL) was mixed sli ghOtly Fby turning the mixture upside down and briefly spinned to bring together tYhe contents of the tube. The solution was then transferred to the middle of the spiInT cartridge and incubated at room temperature for 15 minutes. Then, 350 µL of WSash Buffer I was introduced into the spin cartridge (SC) and then spinned at 120R00 x g for 15 seconds at room temperature and the liquid collected inside the Ecollection tube was discarded. Thereafter, 700 µL of wash buffer 1 was introdIuVced into the SC and spinned at room temperature for 15 seconds at 12000 x g. ThNen 500 µL of wash buffer II was introduced into a clean RNA Wash tube containing Uthe SC and spinned at 12000 x g for 15 seconds (room temperature) to wash off the column-bound RNA. Previous step was repeated. To dry the spin-column membrane with adhered RNA completely; the SC was spinned at 12,000 × g for 1 minute at room temperature. Then, 50 µL of RNase-free water was introduced to the middle of the SC inside the RNA recovery tube and incubated for 1 minute (room temperature). RNA was eluted by centrifuging the SC for 2 minutes at 12000 x g at room temperature. The o eluted RNA was kept at -20 C in 10 µL aliquots until used. 49 3.2.4.4 Quantification of DNA /RNA The purity and concentration of DNA and RNA were checked by quantifying the absorption of ultraviolet light at 260 and 280 nm with a NanoDrop® spectrophotometer. 3.2.4.5 Gel preparation and electrophoresis Pre-weighed agarose was suspended in 1x TAE buffer (Appendix II) to achieve 0.8% and 1.5% concentrations. The mixture was heated to melt the agarose completely. T he o ® suspension was cooled to about 54 C and ethidium bromide (CARL ROTHY) was introduced to achieve a final concentration of 0.5 µg/mL. A gel tray was taRped at the ends with glue tape and appropriate comb was inserted. The gel was Apoured to the thickness preferred and the gel was allowed to set and tIhBen Rimmersed in an electrophoretic tank containing 1X TAE buffer. L 3.2.4.6 Preparation of sample for agarose gel electropNhoresis The sample DNA /RNA (5 µL) was mixed withA 6x gel-loading buffer (Thermo Scientific) in a 5:1 (vol/vol) proportion pArecDeeding filling into wells. For PCR products, DNA loading dye was not mixed with the product prior to filling into wells because it already contains gel-loading bIuBffer. DNA molecular weight marker (100 bp or 1 kb plus ladders) was includedF in a well in each gel to determine the size of the amplicons. Electrophoresis wa sO carried out until the dye in the loading buffer migrated an appropriate distance (4Y5 – 60 minutes) with voltage of 90 – 100 on 0.8 - 1.5% agarose. IT 3.2.4.7 VisualRisatiSon of DNA The ethidiuEm bromide stained gel was visualised using a UV transilluminator. The picturIesV were captured saved electronically with the Gel Documentation and Analysis SyNstem (GenoSens 1560). Electronic images were edited using Windows Photo Viewer Uand images were labeled with Microsoft word. 3.2.4.8 Molecular identification of Pseudomonas aeruginosa Confirmation of identity of P. aeruginosa strains was done by PCR using oprI and oprL primers which anneal to the begining and the end of the exposed reading frame of the oprI and oprL genes, respectively (De Vos et al., 1992; Lim et al., 1997). A total 25 μL reaction mixture comprising 12.5 μL WizPure™ PCR 2X Master (Wizbiosolutions, Korea South), 1 µL of 10 μmol/L forward and reverse primers of 50 oprI and oprL (Alpha DNA) genes in a singleplex reaction, 1 µL of genomic DNA and 9.5 µL DNase/RNase-Free Distilled Water (Invitrogen) was prepared in a 0.2 mL PCR tube. The tubes were loaded into Thermal Cycler (Applied Biosystems). PCR Conditions: DNA was first denatured at 94°C for 5 minutes then followed by30 cycles of denaturation at 94 °C for 30 s, annealing at 55°C for 40 s, and extension at 72°C for 50 s, and a final extension step at 72°C for 5 minutes. A positive control with P. aeruginosa ATCC 27853 DNA and a negative control without DNA were included. Gel electrophoresis was performed on all amplified DNA fragments with 1.5% agaro se and estimated with 100 bp plus ladder as explained in sections 3.2.4.5 to 3.2.4R.7. Y 3.2.5 Detection of carbapenemase encoding genes by PCR A 3.2.5.1 Class A carbapenemases BR GES, NMC-A, BIC-1 and SME: Multiplex PCR was c aLrriIed out in a Thermal Cycler (Applied Biosystems) with a total volume of 25 µL containing 1µL of total DNA isolated in section 3.2.2.1, 12.5 μL of 2× MyTaq NRed Mix (Bioline, London), 1 µL each 10 μmol/L of both forward and reverse pDrimAers and 9.5 µl DNase/RNase-Free Distilled Water (Invitrogen). The followingA thermal cycling conditions were used: DNA was first denatured at 94°C for 5 mIiBnutes and followed by 25 cycles of 94 °C for 30 s, 50°C for 30 s and 72°C for 1 Fminute of denaturing, annealing and extension, and a final extension step at 72°C Ofor 7 minutes. A positive control with P. aeruginosa ATCC 27853 DNA and aY negative control lacking DNA were included (Hong et al., 2012). T Gel electrophoresiSs wIas performed on all amplified DNA fragments with 1.5% agarose and estimated Rwith 100 bp plus ladder as explained in sections 3.2.4.5 to 3.2.4.7. 3.2.5.2I VCla Ess B carbapenemases (MBLs) A N25 µl volume containing 1µL of total DNA isolated in section 3.2.2.1, 12.5 μL of UWizPure™ PCR 2X Master mix (Wizbiosolutions, Korea South), 1 µL each 10 μmol/L of both forward and reverse primers and 9.5 µL distilled water devoid DNase / RNase (Invitrogen) were mixed in 0,2 mL tubes. The tubes were loaded into a Thermal Cycler instrument (Applied Biosystems). For the detection of blaIMP, blaVIM and blaSPM, the following thermal cycling conditions were used: DNA was first denatured at 94°C for 5 minutes then followed by 36 cycles of denaturation at 94 °C for 30 s, annealing at specific temperatures (blaIMP, blaSPM - 51 52°C and blaVIM -55°C) for 40 s and extension at 72°C for 50 s, with a final extension for 5 minutes at 72°C (Poirel et al., 2011). For the detection of blaNDM, the following thermal cycling conditions were used: DNA was first denatured at 95°C for 5 minutes then followed by 30 cycles of denaturation at o o 95 °C for 30 s, annealing at 60 C for 30 s, and extension at 72 C for 30 s, with a single final extension for 3 minutes at 72°C. For the detection of blaAIM, blaDIM, blaSIM and blaGIM, overall capacity of 2Y5 µL containing 1µL of total DNA isolated in section 3.2.2.1, 12.5 μL PCR 2X MRaster mix (Bioline, London; Wizbiosolutions, Korea South), 1 µL each of bothA forward and reverse primers (10 μmol/L), 3 μL of dimethyl sulfoxide and 6B.5 µRL distilled water devoid DNase / RNase (Invitrogen) was used. The followIing thermal cycling o conditions were used for amplification: 10 minutes at 9L4 C with 36 cycles of o o o denaturation at 94 C for 30 seconds, 52 C at 40 secondsN and 72 C at 50 seconds, and a single finishing step of 5 minutes at 72°C (PoirelD et aAl., 2011). Gel electrophoresis was performed on all amplified DNA fragments with 1.5% agarose and estimated with 100 bp plus ladder oBr 1AI kb plus ladder as explained in sections 3.2.4.5 to 3.2.4.7. F 3.2.5.3 Class D carbapenemasOes For the detection of blaOXY A-48 and blaOXA-58, a 25 µL volume containing 1µL of total DNA isolated in seIctTion 3.2.2.1, 12.5 μL of WizPure™ PCR 2X Master mix (Wizbiosolutions, SKorea South), 1 µL each 10 μmol/L of both forward and reverse primers and 9.R5 µL distilled water devoid DNase / RNase (Invitrogen). The tubes were loaded VintoE a Thermal Cycler instrument (Applied Biosystems). PCNR IConditions: The following thermal cycling conditions were used: DNA was first U odenatured at 94°C for 5 minutes then followed by 30 cycles at 94 °C for 35 s, 60 C for o 35 s, 72 C for 30 s, for denaturing, annealing and extension, respectively and finally a single extension for 6 minutes at 72°C. Gel electrophoresis was performed on all amplified DNA fragments with 1.5% agarose and estimated with 100 bp plus ladder as explained in sections 3.2.4.5 to 3.2.4.7. 52 3.2.6.1 Characterisation of class 1, 2 and 3 integron One microliter of entire DNA was exposed to PCR in a 20 μL reaction capacity. The mixture contained 10 μL of 2× MyTaq Red Mix (Bioline, London; Wizbiosolutions, Korea South), 10 μmol/L of each primer and 7 µL distilled water devoid of DNase and RNase (Invitrogen). The tubes were loaded into Thermal Cycler (Applied Biosystems). PCR conditions: The following thermal cycling conditions were used: DNA was fi rst denatured at 94°C for 5 minutes and then succeeded by 30 cycles at 94 °C forY 30 s, o o 55 C for 30 s, 72 C for 45 s, of denaturing, annealing and extension, respectRively with a final extension for 10 minutes at 72°C (White et al., 2001). A Gel electrophoresis was performed on all amplified DNA fragments wRith 1.5% agarose and estimated with 100 bp plus or 1 kb plus ladders as expl aiLnedI Bin sections 3.2.4.5 to 3.2.4.7. N 3.2.6.2 Restriction Fragment Length PolymorDphisAm (RFLP) for differentiation of integrons In strains that were positive for integroInB, in Ategrase PCR products were analysed to determine the class of the integronF p ossessed by each strain by digesting the PCR products with RsaI (Thermo Scientific) restriction enzyme. PCR products (10 µL) were digested with 1 µL of Rs aOI by adding 2 µL of 10x buffer and 7 μL of PCR water (UltraPure, Invitrogen) to Yobtain the overall capacity of 20 μL. The mix was incubated o o at 37 C for 3 hours. TIoT inactivate RsaI enzyme, the mix was further incubated at 80 C for 20 minutesR. ThSe digested amplicons were electrophoresed for 45 minutes at 90 V on 1.5% aEgarose. After digestion of PCR products with RsaI restriction enzyme, integrIasVe I will give only one amplicon size of 491 bp, integrase II will give two amNplicon sizes of 334 bp and 157 bp, while integrase III will give rise to three Ufragment sizes of 97 bp, 104 bp and 290 bp if they were present (White et al., 2001). All amplified DNA fragments were estimated alongside with a DNA molecular weight marker (100 bp plus and 1 kb plus ladders) as explained in sections 3.2.4.5 to 3.2.4.7. 3.2.6.3 Characterisation of cassette arrays Strains that gave amplicon size of 491 bp were subjected to another PCR using hep58 and hep59 primers to detect class 1 gene cassette arrays (White et al., 2001). 53 PCR Conditions: the initial denaturation temperature was set at 94°C for 5 minutes, succeeded by 30 cycles of 94°C for 30 seconds, annealing at 55°C for 30 seconds followed by 72°C for 4 minutes (White et al., 2000). All amplified DNA fragments were estimated alongside with a DNA molecular weight marker (1 kb plus ladder) as explained in sections 3.2.4.5 to 3.2.4.7. 3.2.7 Purification and Sequencing of PCR products 3.2.7.1 Purification of PCR products For DNA purification purposes, a WizPrep Gel/PCR purification kit (WizbioRsoluYtions) was used in accordance with Manufacturer‟s instruction. The PCR product was transferred into 1.5ml eppendorf tube. To 1 volume of the sample, 5 vAolume of GP Buffer was introduced and mixed thoroughly. The mixture was rBemRoved into a DNA Spin Column (SC), spinned at 13,000 rpm for 19 minute and LtheI filtrate was discarded. Then, 700 μL of Wash buffer containing ethanol was introduced into the DNA SC, spinned at 13,000 rpm for 30 seconds and the filtrateA waNs discarded. The Spin Column was spinned for an additional 1 minute and the DDNA Spin Column was transferred to a new 1.5 ml tube. Then, 50 μL of Elution BAuffer was introduced to the centre of the column matrix and allowed to stand fo r I1B minute and spin for 1 minute at 13,000 rpm to elute the DNA. F 3.2.7.2 Sequencing of PCR p roOducts The purified PCR produTctsY for blaVIM and blaNDM were sent to First Base Laboratory in Malaysia for sequenIcing using both forward and reverse primers. Sequencing reactions wereR doSne with BigDye Terminator v3.1 Cycle Sequencing Kit on ABI PRISM 373E0xi Genetic analyser (Applied Biosystems, USA). The resulting sequences were IcVompared with the sequences in the NCBI GenBank database through the BLNAST program (http://blast.ncbi.nlm.nih.gov/Blast.cgi). U3.2.8 Transformation experiment Preparation of competent cell: A single colony of Escherichia coli (DH5α) from 18- 24 hour old culture was added into LB medium and incubated overnight at 37°C with shaking at 250 rpm. LB medium (100 mL) was inoculated with 1ml of the E. coli (DH5α) culture and incubated at 37°C for 2-3 hours with agitation until the OD reached 600 nm. The culture was put in an ice bath for 10 minutes and transferred into 50 mL falcon tubes. The pre-chilled culture was spinned at 2700x g for 10 minutes at 54 4°C and the supernatant was discarded. The cell pellet was mixed in 1.6 ml of cold CaCl2 and incubated on ice for 30 minutes. The cell solution was spinned at 2700x g for 10 minutes at 4°C and the supernatant removed. Ice cold 100 mM CaCl2 was used to liquefy the pellet and incubated on ice for 20 minutes. Then, 0.5 mL of ice cold 80% glycerol was put and dispensed in 100 µL aliquots and stored at -80°C (Chang et al., 2017) Transformation protocol: DH5α cells (chemically competent) were defrosted on ice for 30 minutes. At this juncture, 2 uL of plasmid DNA was added and the cellYs were o heat-shocked at 42 C for 60 seconds. The cells were placed on ice for 5R minutes. o Thereafter, 900 uL of LB broth was introduced and incubated at 37 C fAor 2 hours at 125 rpm. The cell was centrifuged at 5000 rpm for 45 seconIdBs. L RB (600 µL) was removed and the suspension was resuspended. Then, 100 μL Lwas spread onto the plate comprising LB agar and 100 μg/mL of ampicillin and incubated overnight. Transformants were selected from colonies grown on NLuria Bertani agar containing 100 µg/mL of ampicillin. Plasmid DNA was eDxtraActed from the transformant with DNA extraction kit (Wizbiosolutions) and suAbjected to PCR using PCR conditions for blaVIM and blaNDM detection as explaine dI iBn section 3.2.5.2. 3.2.9 Statistical analysis F Sensitivity and specificityY, ne g Oative predictive values (NPV), and positive predictive values (PPV) were aTlso calculated. Fisher‟s exact test was used to define the connection betweSen Iintegrons and MBL(s); association between exoU and exoS; association of Rclinical source with T3SS and association of T3SS and clonality. One-way analysEis of variance (ANOVA) was used to determine the association between antibioItVic sensitivity profile of MBL+ve and MBL-ve strains. A P value of < 0.05 was coNnsidered to be significant. U3.2.10 Quantification of gene expression 3.2.10.1 Removing the co-purified contaminating DNA and synthesis of First strand complementary DNA (cDNA) To remove any trace of genomic DNA co-purified along with total RNA prior to RT- PCR. DNase 1 (Thermo Scientific) was used following manufacturer‟s instruction. The following components were assembled in a thin-walled 0.2 ml tube, on ice: 0.5 - 4 55 µL (1 µg) of RNA template, 1 µL of 10x DNase 1 buffer, 1 µL DNase 1, RNase-free water to make up 10 µL. The mixture was incubated for 30 minutes at 37°C. cDNA was synthesised by means of the WizScript™ cDNA Synthesis Kit (Wizbiosolutions, Korea) according to the manufacturers instruction. The following components were added to the purified RNA. 1 µL of EDTA (50mM), 2 µL of Random hexamer and 1 µL of dNTP mix (2.5 mM). The mixture was incubated at 65°C for 10min and then cooled immediately on ice. Thereafter, 2 µL of 10x buffer, 1 µL of Reverse transcriptase (200 U/µL), 0.5 YµL of RNase Inhibitor (40U/µL), 1 µL of DTT (100mM) and RNase free water Ato mRake up to 20 µL were added to the template RNA and primer mix. The reacRtion mixture was mixed slightly and incubated at 37°C for 60 min. Reverse transcIriBptase was inactivated by incubation at 70°C for 10 min and cooling on ice. Synth esLised cDNA was kept at - 20°C and used as template for PCR N 3.2.10.2 Expression levels quantification ofD efAflux pumps, ampC and oprD transcripts A The expression levels of two genes in IeBach of the four major P. aeruginosa efflux pumps; (mexA, mexB, mexC, mexDF, mexE, mexF, mexX and mexY), ampC and oprD genes were analysed with Ste pO One ® real-time reverse transcription-PCR (Applied Biosystem). Y The primers for the PICTR amplification of cDNA were shown in Table 3.1. Primer conditions weRre oSptimised with PCR. Real time PCR was accomplished in a set of three per saEmple. A total capacity of 20 µL comprised of 10 µL of SYBR green /ROX qPCRI mVaster mix (Thermo Scientific), 0.5 µL (10 pmol) each of onward and backward prNimers, 1 µL of the cDNA, 8 µL of PCR water, was used. PCR conditions include Uinitial denaturation temperature at 95°C for 10 min to trigger the modified Taq polymerase, followed by 40 cycles of 20s at 95°C, 30s at 66°C (mexA, mexB, mexC, mexX and mexY), 30s at 67°C (mexE), 30s at 68°C (mexF), 30s at 64°C (mexD), 30s at 62°C (ampC), 30s at 62°C (oprD), 30s at 62°C (rpsL) and 30s at 72°C were performed. Data acquisition was performed at 72°C. At the end of the 40 cycles, a melt curve was determined to look for the existence of a single PCR product. The mean threshold cycle (ct) value of triplicate sample was taken. 56 To regulate the expression levels of mRNA, housekeeping gene (rpsL gene) was included and outcomes were referenced against P. aeruginosa ATCC 27853 expressions. When the equivalent mRNA level was at least 2-fold (mexA, mexB), 4- fold (mexX and mexY) or 100-fold (mexC, mexD, mexE and mexF) higher than that for P. aeruginosa ATCC 27853, strains were considered to be overexpressed (Hocquet et al., 2006). Control reaction lacking reverse transcriptase was set up to check for contaminating genomic DNA. Fold expression level was calculated using the formula below: Y TE – HE = ΔCTE R TC – HC = ΔCTC A ΔCTE – ΔCTC = ΔΔCt R -ΔΔCt Fold change = 2^ B LI Where: N TE = Threshold cycle value obtained from amplDificaAtion of test organism cDNA with resistance gene primer A HE = Threshold cycle value obtained frIomB amplification of test organism cDNA with housekeeping gene primer F TC = Threshold cycle value obOtained from amplification of control organism cDNA with resistance gene priTmeYr I HC = Threshold cSycle value obtained from amplification of control organism cDNA with housekeeRping gene primer E ΔCTEI =V Delta Ct housekeeping gene UΔCNTC = Delta Ct control 3.2.11 Detection of type III secretion system Multiplex PCR was done to detect type III secretion systems in carbapenem non- susceptible strains of P. aeruginosa using primers which amplify the conserved regions of exoS, exoU, exoT, and exoY genes (Ajayi et al., 2003). PCR set up included 1 µL of DNA template, 0.5 µL of each primers (Macrogen), 12.5 µL of WizPure™ PCR 2X Master mix (Wizbiosolutions, Korea South), and 7.5 µL of 57 DNase/RNase-Free Distilled Water (Invitrogen). P. aeruginosa ATCC 27853 served as a control. PCR Conditions: The following thermal cycling conditions were used: DNA was first o denatured at 94°C for 2 minutes then followed by 36 cycles at 94 °C for 30 s, 58 C for o 30 s, and 68 C for 1 minute, of denaturing, annealing and extension, respectively with a single final extension at 68°C for 7 minutes (Ajayi et al., 2003). Gel electrophoresis was performed on all amplified DNA fragments with 1.5% agaro se and estimated with 100 bp plus and 1kb plus ladders as explained in sectionsR 3.2Y.4.5 to 3.2.4.7. A 3.2.12 PCR-based genotyping BR For genotyping of CRPA, three PCR-based genotyping met hLodsI were employed. PCR set up included 1 µL of DNA template, 0.5 µL of each Nprimer (Macrogen), 12.5 µL of WizPure™ PCR 2X Master mix (Wizbiosolutions, Korea South), and 7.5 µL of distilled water devoid of DNase and RNase (InvitDrogAen). 3.2.12.1 Repetitive extragenic palindr AIomBic PCR (REP PCR) Two primers were used; REP1R-I and REP2-I (Table 3.1). PCR Conditions: Preliminary denFaturation temperature was set at 95°C for 2 min; succeeded by 35 cycles of 95° CO for 30 s, 38°C for 1 minute, and 72°C for 2 minutes of denaturation, annealingT anYd extension, respectively. A final extension step was set at 72°C for 16 minutesI. The arrangement of amplified DNA fragments was scored alongside with a SDNA molecular weight marker (1 kb plus ladder) as explained in sections 3.E2.4.R5 to 3.2.4.7. Data matrices were computed based on the presence or absenceV of bands at definite positions. REP-PCR profiles were compared with dice coNeffiIcient method and clustered with Unweighted Pair Group Method with Arithmetic UMean (UPGMA) using Phylotree software. 3.2.12.2 Enterobacterial Repetitive Intragenic Consensus-PCR (ERIC PCR) Two primers were used; ERIC1R and ERIC2 (Table 3.1). PCR Conditions: Preliminary denaturation temperature of 94°C for 2 min; followed by 35 cycles of 94°C for 45 s, 52°C for 45 s, and 72°C for 2 minutes of denaturation, annealing and extension, respectively. A final extension step was set at 72°C for 10 minutes (Pinna et al., 2009). The arrangement of amplified DNA fragments was scored 58 alongside with a DNA molecular weight marker (1 kb plus ladder) as explained in sections 3.2.4.5 to 3.2.4.7. Data matrices were computed based on the presence or absence of bands at definite positions. ERIC-PCR profiles were compared with dice coefficient method and clustered with Unweighted Pair Group Method with Arithmetic Mean (UPGMA) using Phylotree software. 3.2.12.3 BOX PCR One primer (BOX-A1R) was used for BOX-PCR typing of P. aeruginosa (Table 3.1 ). PCR Conditions: Preliminary denaturation temperature of 94°C for 2 min; folYlowed by 35 cycles of 94°C for 45 s, 52°C for 45 s, and 72°C for 2 minutes of denRaturation, annealing and extension, respectively. A final extension step set at 72°C fAor 10 minutes was used (Pinna et al., 2009). The arrangement of amplifiedI DBNA R fragments was scored alongside with a DNA molecular weight marker (1 k bL plus ladder) as explained in sections 3.2.4.5 to 3.2.4.7. Data matrices were computed based on the presence or absence of bands at definite positions. BOX-PCR pArofNiles were compared with dice coefficient method and clustered with UnweighteDd Pair Group Method with Arithmetic Mean (UPGMA) using Phylotree software. A F I B O SI TY ERIV UN 59 CHAPTER FOUR RESULTS 4.1 Identification of clinical isolates of Pseudomonas aeruginosa Four hundred and forty-seven isolates were collected on nutrient agar slants fro m seven tertiary hospitals in five southwestern states of Nigeria and transported Yto the Department of Pharmaceutical Microbiology laboratory, University oAf IbRadan, for further identification. Ninety (90) out of 92 isolates collected from URniversity College Hospital, Ibadan (UCHI) were identified as P. aeruginosaI. BAmong 97 isolates collected from Ogun state, 47 and 50 isolates from Ola bLisi Onabanjo University Teaching Hospital, Sagamu (OTHS) and Federal MedicNal Centre Abeokuta (FMCA), respectively, 45 isolates each were identified as PA. aeruginosa. Out of 48 isolates collected from Ladoke Akintola University TeaDching Hospital, Osogbo (LTHO), 45 were identified as P. aeruginosa, 45 isolAates collected from Obafemi Awolowo University Teaching Hospital, Ile-Ife ( OITBHI) were identified as P. aeruginosa; of the 94 isolates collected from Federal FMedical Centre, Owo (FMCO), 90 were identified as P. aeruginosa. Of the 72 is oOlates collected from Federal Medical Centre, Ido-Ekiti (FMCI), 70 were recogniYsed as P. aeruginosa. the identification centered on their growth on PseudomoInaTs cetrimide agar, pigment production, motility, production of cytochrome c oxSidase, catalase, urease positive, inability to produce hydrogen sulphide, aEbiliRty to utilise citrate as their only carbon source, oxidative utilisation of glucoseV. Further confirmation at the molecular level was done at the Department of BiosciIences, COMSATS University, Islamabad, Pakistan. UN4.2 Distribution of clinical isolates of P. aeruginosa according to site of isolation in relation to hospital Figure 4.1 illustrates the distribution of clinical isolates of P. aeruginosa according to their clinical sources. Out of 430 P. aeruginosa isolates, 298 (69.3%) isolates were from wound swab, 5 (1.2%) from wound biopsy, 49 (11.4%) from ear, 43 (10.0%) from urine, 10 (2.3%) from eye, 5 (1.2%) from ear pus, 5 (1.2%) from sputum, 60 4 (0.9%) from blood, 3 (0.7%) from tracheal aspirate, 3 (0.7%) from cerebrospinal fluid, 4 (0.9%) from catheter tips and 1 (0.2%) from genital discharge. Table 4.1 presents the dissemination of clinical isolates of P. aeruginosa according to hospital with respect to their clinical sources. Out of the 90 isolates from UCHI, 37 (41.1%) were isolated from wound swab, followed by17 (18.9%) from ear, 16 (17.8%) from urine, 5 (5.6%) from wound biopsy, 4 (4.4%) from catheter tips, 3 (3.3%) each from eye, tracheal aspirate and blood, 2 (2.2%) each from sputum and cerebrospinal fluid, and 1 (1.1%) from genital discharge and ear pus. Among the FMCA isolates, 32 (71.1%) were isolated from wound followed by 5 (11.1%) from urine, 3 (6R.7%Y) from eye, 2 (4.4%) from ear, 2 (2.2%) from catheter tips and 1 (2.2%) fArom ear pus. Twenty-eight (62.2%), out of forty-five (45) isolates from OTHS wRere isolated from wound followed by 7 (15.6%) from urine, 5 (11.1%) from ear,I 2B (4.4%) from eye, 2 (4.4%) from sputum and 1 (2.2%) from ear pus. Out of 4 5L isolates from LTHO, 37 (82.2%) were from wound, 2 (4.4%) isolates each were fNrom urine, ear, ear pus while 1 (2.2%) each were from cerebrospinal fluid and blooAd. Out of 45 isolates from OTHI, 24 (53.3%) were from wound, 11 (24.4%) AwerDe from urine, 6 (13.3%) from ear, 2 (4.4%) were from wound biopsy, while 1 (2.2%) each were from eye and sputum. Ninety (90) isolates collected from FM ICBO was from wound 79 (87.8%) and ear 11 (12.2%) only. Out of 70 isolates oFf P. aeruginosa collected from FMCI, 71 (87.1%) were isolated from wound, 6 ( 8O.6%) from ear, 2 (2.9%) from urine and 1 (1.4%) from eye. TY RS I E IV U N 61 Y RA R B N LI DA IB A OF Figure 4.1. Distribution ofY the isolates according to clinical source TSI ER NI V U 62 Table 4.1. Occurrence of P. aeruginosa strains according to hospital with respect to their clinical sources Site of isolation UCHI FMCA OTHS LTHO OTHI FMCO FMCI (90) (45) (45) (45) (45) (90) (70) Wound (298) 37 32 28 37 24 79 61 Wound biopsy (5) 3 - - - 2 - - Ear (49) 17 2 5 2 6 11 6 Eye (10) 3 3 2 - 1 - 1 Catheter tips (4) 2 2 - - - - RY- Ear pus (5) 1 1 1 2 - - A - Urine (43) 16 5 7 2 11 R- 2 CSF (3) 2 - - 1 - IB - - Sputum (5) 2 - 2 - L1 - - Tracheal aspirate (3) 3 - - - N - - - Genital discharge (1) 1 - - -A - - - Blood (4) 3 - - AD1 - - - Key: IB UCHI = University College HospitaFl, I badan FMCA = Federal Medical Cen trOe, Abeokuta OTHS = Olabisi OnabanjoY University Teaching Hospital, Sagamu LTHO = Ladoke AkinItoTla University Teaching Hospital, Osogbo OTHI = Obafemi ASwolowo University Teaching Hospital Complex, Ile-Ife FMCO = FedeRral Medical Centre, Owo FMCI = FeEderal Medical Centre, Ido-Ekiti V U N I 63 4.3 Antibiotic susceptibility profile of clinical isolates of Pseudomonas aeruginosa Table 4.2 is the antibiotic susceptibility profile of 430 clinical isolates of P. aeruginosa against twenty-three selected antibiotics. One hundred percent (100%) of the isolates were resistant to ampicillin, cephalothin and cefuroxime followed by amoxicillin clavulanate (99.5%), ticarcillin (64.4%), ceftriaxone (54.4%), cefepime (45.8%), gentamicin (43.3%), ofloxacin (43.3%), ciprofloxacin (39.3%), ceftazidime (38.1%), levofloxacin (38.1%), cefoperaxone (33.0%), tobramycin (32.6%), amikacin (27.4% ), meropenem (18.8%), piperacillin (18.6%), doripenem (17.8%), piperacillin taRzobYactam (17.6%), aztreonam (16.7%), imipenem (15.8%), colistin sulphate A(5.1%) and polymyxin B (3.7%). R Percentage antibiotic resistance among isolates from the sevenI Bhospitals is given in figure 4.2. An equivalent proportion (100%) was LNobs erved to be resistant to amoxicillin clavulanate in all the hospitals. Highest proportion of resistance to cefepime (65.6%), ofloxacin (62.2%), levofloxDacinA (56.7%), ciprofloxacin (56.7%), gentamicin (56.7%), tobramycin (51.1%), Apiperacillin (27.8%), amikacin (42.2%), imipenem (27.8%), colistin sulphate ( 8I.9B%) and polymyxin B (7.8%) was observed among isolates from UCHI. Among isolates from the seven hospitals, OTHI isolates showed highest proportion of F Oresistance to ticarcillin (84.4%), ceftriaxone (71.1%), cefoperaxone (66.7%), ceftazidime (60%), piperacillin tazobactam (31.1%), meropenem and aztreoTnamY (28.9%). Isolates from FMCO gave least resistance to ciprofloxacin (28.9%I) and gentamicin (35.6%), while resistance to tobramycin (13.3%), amikaciSn (11.1%), imipenem (2.2%), ceftriaxone (32.7%), ticarcillin (48.9%), piEperRacillin (8.9%) and piperacillin tazobactam (4.4) was least in isolate from FMCAI.V UTaNble 4.3 shows the antibiotic resistance profile of P. aeruginosa isolates. According to the criteria for classification of multidrug resistant P. aeruginosa described by Magiorakos and colleagues (2012), 135 (31.4%) and 50 (11.6%) of P. aeruginosa isolates were multidrug resistant (MDR) and extensive drug resistant (XDR), respectively. Multidrug resistance was highest among isolates from UCHI 33/90 (36.7%) followed by OTHI 16/45 (35.6%), FMCI 31/90 (34.4%), FMCA 15/45 (33.3%), FMCO 25/90 (27.8%), OTHS 12/45 (26.7%) and LTHO 10/45 (22.2%). The occurrence of extensively drug resistance (XDR) among the clinical isolates according 64 to the hospital was in the following descending order: UCHI 20 (22.2%) > OTHI 8 (17.8%) > LTHO 7 (15.6%) > FMCI 5 (7.1%) > FMCO 6 (6.7%) = OTHS 3 (6.7%) > FMCA 1 (2.2%). 4.4. Minimum inhibitory concentrations (MICs) of selected antibiotics against clinical isolates of P. aeruginosa The MIC of five selected antibiotics together with their MIC50 and MIC90 against the 430 clinical isolates of P. aeruginosa is given in Table 4.4. MIC of imipenem a nd meropenem against all the isolates ranged from 0.12 to >64 µg/mL and 0.06 tYo >64 µg/mL, respectively with MIC50 and MIC90 of 1 and >64 µg/mL. AR The MIC of imipenem and meropenem against clinical isolates froRm UCHI, FMCO and FMCI ranged from 0.12 to >64 µg/mL, each. MIC of imipIeBnem and meropenem against clinical isolates from FMCA and OTHS state rang edL from 0.12 to 32 µg/mL and 0.06 to 64 µg/mL, respectively while LTHO and ONTHI isolates had MIC ranged from 0.12 to 64 µg/mL for imipenem and meropAenem. MIC of cefepime against clinical isolates from UCHI, FMCA and OTHAS rDanged from 0.5 to >128 µg/mL while that of FMCO and FMCI ranged from 1B to >128 µg/mL. MIC of cefepime against isolates from LTHO and OTHI ranged fIrom 1 to 64 µg/mL. UCHI and FMCI isolates had the same MIC ranged from 1 Fto >256 µg/mL against ceftazidime with FMCA, OTHS, LTHO, OTHI and FM COO isolates having the same MIC ranged from 0.5 to >256 µg/mL. The MIC Yof ciprofloxacin against clinical isolates from the seven hospitals ranged fromI 128 µg/mL ER S V UN I 65 AR Y R LI B AN Plate 4.1. Zone of growth inhibition of P. aeArugDinosa by aztreonam (ATM), cefepime (FEP) and colistin sulphate (CT), ticarciIllBin clavulanic acid (TIM) OF ITY RS VE I U N 66 Table 4.2. Antibiotic susceptibility profile of clinical isolates of P. aeruginosa Antibiotics Sensitive (%) Intermediate (%) Resistance (%) Ceftazidime 214 (49.8) 47 (10.9) 164 (38.1) Amoxicillin clavulanate 2 (0.5) 0 (0) 428 (99.5) Piperacillin 299 (69.5) 51 (11.9) 80 (18.6) Ticarcillin 32 (7.4) 121 (28.1) 277 (64.4) Cephalothin 0 (0) 0 (0) 430 (100) Cefoperazone 211 (49.1) 77 (17.9) 142 (33Y.0) Piperacillin/ tazobactam 303 (70.5) 51 (11.9) 76 (R17.7) Gentamicin 214 (49.8) 30 (7.0) 1A86 (43.3) Cefuroxime 0 (0) 0 (0) BR430 (100) Ampicillin 0 (0) 0 (0) LI 430 (100) Imipenem 360 (83.7) 2 (0.5) 68 (15.8) Aztreonam 212 (49.3) 146 (A34.N0) 72 (16.7) Ceftriaxone 63 (14.7) 133 (30.9) 234 (54.4) Tobramycin 284 (66.1) A6D (1.4) 140 (32.6) Meropenem 341 (79.3) IB 8 (1.9) 81 (18.8) Amikacin 29 (69.5F) 13 (3.0) 118 (27.4) Levofloxacin 259 (60.2) 7 (1.6) 164 (38.1) Doripenem Y347 O(80.7) 7 (1.6) 76 (17.8) Cefepime 203 (47.2) 30 (7.0) 197 (45.8) Polymixin B TSI 414 (96.3) 0 (0) 16 (3.7) Ciprofloxacin R 257 (59.8) 4 (0.9) 169 (39.3) Colistin suElphate 408 (94.9) 0 (0) 22 (5.1) OfloxIacVin 231 (53.7) 13 (3.0) 16 (3.7) UN 67 PB CT OFX LEV CIP TOB CN Y AK ATM ARFMCI (n = 70)DOR R FMCO (n = 90)MEM B OTHI (n = 45) IMP LI LTHO (n = 45)FEP N OTHS (n = 45)CRO A FMCA (n = 45)CFP AD UCHI (n = 90)CAZ TIC TZP F I B PRL AMC O CXM KF SI TY 0R 20 40 60 80 100E Percentage (%) of resistant P. aeruginosa V FiNgurIe 4.2. Percentage antibiotic resistance of 430 isolates of P. aeruginosa from Useven hospitals Key: AMC=Amoxicillin clavulanate, PRL=Piperacillin, TZP=Piperacillin-tazobactam, TIC=Ticarcillin, CAZ=Ceftazidime, PB=Polymyxin B, CFP=Cefoperazone, CRO=Ceftriaxone, FEP=Cefepime, IPM=Imipenem, CT=colistin sulphate, MEM=Meropenem, DOR=Doripenem, ATM=Aztreonam, AK=Amikacin, CN=Gentamicin, CXM=Cefuroxime, TOB=Tobramycin, KF=Cephalothin, CIP=Ciprofloxacin, LEV=Levofloxacin, OFX=Ofloxacin 68 Antibiotics Table 4.3. Classification of clinical isolates of P. aeruginosa based on antibiotic resistance profiles Hospital MDR (no (%)) XDR (no (%)) Non-MDR (no (%)) UCHI (n = 90) 33 (36.7) 20 (22.2) 37 (41.1) FMCA (n = 45) 15 (33.3) 1 (2.2) 29 (64.4) OTHS (n = 45) 12 (26.7) 3 (6.7) 30 (66.7) RY OTHI (n = 45) 16 (35.6) 8 (17.8) 21 (46.7) RA LTHO (n = 45) 10 (22.2) 7 (15.6) 28 (62.2) LIB FMCO (n = 90) 25 (27.8) 6 (6.7) N59 ( 65.6) FMCI (n = 70) 24 (26.7) 5 (7.1) DA 41 (58.6) Key: A MDR = Multidrug resistant B XDR = Extensive drug resistant I UCHI = University College HosOpitaF l, Ibadan FMCA = Federal Medical YCentre, Abeokuta OTHS = Olabisi OnabIaTnjo University Teaching Hospital, Sagamu LTHO = Ladoke ASkintola University Teaching Hospital, Osogbo OTHI = ObafeRmi Awolowo University Teaching Hospital Complex, Ile-Ife FMCO = FEederal Medical Centre, Owo FMCII =V Federal Medical Centre, Ido-Ekiti UN 69 Y R Table 4.4. Antibiotic susceptibility rates of 430 clinical isolates of P. aeruginosa at indicated MIC in µg/mL A Isolates number at indicated MIC (µg/mL) R Antibiotics 0.0 0.06 .12 0.25 0.5 1 2 4 8 16 32 64 12 25IBBreak % MIC50/MIC3 8 L6 points susceptibl 90 (µg/mL) e c Imipenem 43 65 76 97 54 16 16 16 9 36 2/8 76.7 1/>64 c Meropenem 2 38 73 71 88 49 20 16 11 6 A48N 2/8 74.7 1/>64 cCefepime 3 51 68 73 82 49 D22 31 39 8/32 64.4 8/>128 cCeftazidime 15 49 53 61 48 A42 24 13 25 85 8/32 52.6 8/>256 cCiprofloxaci b54 39 61 37 22 33 25 21 1B5 15 25 4 68 1/4 57.2 1/>128 n I Key: F a The bolded fonts indicate the susceptible strains O b MIC ≤ the value indicated Y c MIC ≥ the value indicated IT Breakpoints according to CLSI, 20S15 (susceptible ≤/resistant>). R IV E 70 UN 4.5 Phenotypic detection of carbapenemase, beta lactamase and metallo beta- lactamase (MBL) Phenotypic detection of carbapenemase using Modified Hodges test showed negative result for all the isolates (Plate 4.2). Highest prevalence of beta lactamase was observed in isolates from UCHI (81.1%) followed by OTHS (80.0%), OTHI (77.7%), FMCO (75.6%), LTHO (75.5%), FMCA (64.4%) and lastly FMCI (61.4%). Based on phenotypic MBL detection using combined disc method with EDTA soaked imipenem and meropenem discs, the prevalence of MBL in clinical isolates of P. aeruginosa w as established to be 15.4%. These isolates exhibited a significant zone size enhRancYement (≥ 7mm) with the EDTA soaked imipenem and meropenem discs when cAompared with the plain imipenem and meropenem discs. UCHI had the highest Rprevalence of 22 (24.4%) followed by OTHI, LTHO, FMCO, OTHS, FMCI and IFBMCA states with the prevalence of 10 (22.2%), 8 (17.8%), 14 (15.6%), 5 (11.1% )L, 6 (8.6%) and 1 (2.2%), respectively (Table 4.5). Plates 4.3a and 4.3b shows Nthe zone of growth inhibition between the EDTA soaked imipenem and meropAenem discs and with the plain imipenem and meropenem discs. AD Figure 4.3 compares the antibiotic sensIitBivity profile of MBL-positive P. aeruginosa isolates and MBL-negative isolateFs to different antipseudomonad antibiotics. There was significant difference i nO antibiotic sensitivity profile of MBL-positive P. aeruginosa isolates when Ycompared with MBL-negative isolates (p < 0.05). MBL-negative P. aeruginosIaT isolates were more sensitive to majority of the antibiotics (17 out of 23) than MBSL-positive strains. MBL-positive strains exhibit 0.0% sensitivity to ticarcillin, cefRtazidime, cefoperazone, ceftriaxone, doripenem and meropenem as compared tEo MBL-negative strains with percentage sensitivity of 8.8, 58.8, 58.0, 17.3, 100 aIndV 100, respectively. MBL-positive strains were more sensitive to only colistin suNlphate and polymyxin B (97%) while MBL-negative strains had 96.2% and 94.5% Usensitivity to colistin sulphate and polymyxin B, respectively. 4.6 Curing of antibiotic resistance in carbapenem-resistant P. aeruginosa Table 4.6a and 4.6b shows the effect of different concentrations of ethidium bromide on the susceptibility of carbapenem-resistant strains of P. aeruginosa to imipenem and meropenem. After subjecting the ten carbapenem-resistant strains isolates to 12.5, 25, 50 and 100 μg/mL concentrations of ethidium bromide, the carbapenem-resistant strains showed varying degree of sensitivity to imipenem and meropenem at 0.5, 1, 2, 71 4 and 8 μg/mL concentrations after treatment with 50 and 100 μg/mL ethidium bromide while resistance was observed with untreated strains. RY A LIB R AN D BA F I O SI TY VE R U N I 72 Table 4.5. Prevalence of beta lactamase and MBL in clinical isolates of CRPA using phenotypic method Hospital Beta-lactamase MBL UCHI (n = 90) 81.1% (73) 24.4% (22) FMCA (n = 45) 64.4% (29) 2.2% (1) OTHS (n = 45) 80.0% (36) 11.1% (5) Y OTHI (n = 45) 77.7% (35) 22.2% (10) AR LTHO (n = 45) 75.5% (34) 17.8% (8) BRI FMCO (n = 90) 75.6% (68) 15.6% (14) L FMCI (n = 70) 61.4% (43) 8.6%A (6)N Total (n = 430) 74.0% (318) A1D5.4% (66) Key: IB UCHI = University College HospitaFl, Ibadan FMCA = Federal Medical Cen trOe, Abeokuta OTHS = Olabisi OnabanjoY University Teaching Hospital, Sagamu LTHO = Ladoke AkinItoTla University Teaching Hospital, Osogbo OTHI = Obafemi ASwolowo University Teaching Hospital Complex, Ile-Ife FMCO = FedeRral Medical Centre, Owo FMCI =V FeEderal Medical Centre, Ido-Ekiti I U N 73 Y AR LIB R AN AD Plate 4.2. Modified Hodges test for IdeBtection of carbapenemases. There was no indentation or flattening at the inteFrsect of control organism (E. coli ATCC 25922) with the test organism (P. aeru gOinosa) indicating a negative result ITY RS IV E C N D U A B 74 MEM disc + EDTA RY MEM disc RA IPM disc + EDTA LI B IPM disc DA N A IB Plate 4.3a. No inhibition zone sizeF au gmentation with EDTA impregnated imipenem and meropenem discs in comparison with the plain imipenem and meropenem discs (Phenotypic MBL negative). O ITY ER S IV UN 75 IPM disc MEM disc Y AR IB R IPM disc + EDTA MEM disc + ED L NTA DA BA F I Plate 4.3b. Inhibition zone si zOe augmentation with the EDTA soaked imipenem and meropenem discs in comYparison with the plain imipenem and meropenem discs (Phenotypic MBL posIitTive) S VE R NIU 76 100 MBL +ve (n=66) MBL –ve (n=364) 90 80 70 60 Y 50 40 RA R 30 LIB 20 N 10 DA 0 BA F I O Antibiotics ITY Figure 4.3. ComSparison of antibiotic sensitivity among MBL-positive and MBL-negative P. aerRuginosa isolates Key: E * N= Ip V<0.05, KF = Cephalothin, CXM = Cefuroxime, AMP = Ampicillin, UAMC=Amoxicillin clavulanate, PRL* = Piperacillin, TZP* = Piperacillin-tazobactam, TIC* = Ticarcillin, CAZ* = Ceftazidime, PB = Polymyxin B, CFP* = Cefoperazone, CRO* = Ceftriaxone, FEP* = Cefepime, IPM* = Imipenem, CT = colistin sulphate, MEM* = Meropenem, DOR* = Doripenem, ATM* = Aztreonam, AK* = Amikacin, CN* = Gentamicin, TOB* = Tobramycin, CIP* = Ciprofloxacin, LEV* = Levofloxacin, OFX* = Ofloxacin. 77 Percentage sensitive KF CXM AMP AMC PRL* TZP* TIC* CAZ* CFP* CRO* FEP* IMP* MEM* DOR* ATM* AK* CN* TOB* CIP* LEV* OFX* CT PB Table 4.6a. Effect of 100 μg/mL ethidium bromide on the susceptibility of carbapenem-resistant strains of P. aeruginosa to imipenem and meropenem Treated strains of P. Imipenem (μg/mL) Meropenem (μg/mL) aeruginosa 0.5 1 2 4 8 0.5 1 2 4 8 PS093 Colony 1 8 11 14 18 22 8 13 17 19 20 Colony 2 8 11 14 20 22 8 13 17 19 21 Colony 3 8 16 18 20 22 8 14 17 18 20 Colony 4 8 17 20 21 23 8 14 17 19 23 Control 8 8 8 8 8 8 8 8 8 R8 Y PS146 Colony 1 8 12 18 20 23 8 13 16 19A 20 Colony 2 8 12 20 20 23 8 13 16 R19 20 Colony 3 8 13 19 22 24 8 13 I1B6 19 21 Colony 4 8 12 19 22 24 8 13L 15 15 20 Control 8 8 8 8 8 8 N 8 8 8 8 PS150 A Colony 1 8 14 16 20 2D4 8 15 20 22 23 Colony 2 8 13 16 20 A25 8 15 20 19 21 Colony 3 8 12 16 IB16 23 8 16 20 18 21 Colony 4 8 12 F16 18 24 8 16 19 19 22 Control 8 8 8 8 8 8 8 8 8 8 PS154 O Colony 1 8 Y 13 15 18 26 8 12 14 17 21 Colony 2 T8 13 15 19 27 8 13 14 16 21 Colony 3 SI 8 14 16 19 26 8 13 15 16 24 Colony 4 R 8 13 16 20 26 8 12 14 17 23 Control E 8 8 8 8 8 8 8 8 8 8 PS168I V CoNlony 1 8 13 18 20 24 8 14 18 20 23 Colony 2 8 13 18 20 25 8 14 17 20 23 UColony 3 8 13 17 19 25 8 13 17 20 24 Colony 4 8 13 17 20 24 8 14 18 21 24 Control 8 8 8 8 8 8 8 8 8 8 Key: Diameter of cork borer = 8 mm 78 Table 4.6a. Effect of 100 μg/mL ethidium bromide on the susceptibility of carbapenem-resistant strains of P. aeruginosa to imipenem and meropenem (cont‟d) Treated strains of P. Imipenem (μg/mL) Meropenem (μg/mL) aeruginosa 0.5 1 2 4 8 0.5 1 2 4 8 PS184 Colony 1 8 13 16 18 22 8 14 17 19 23 Colony 2 8 14 16 18 23 8 13 17 19 24 Colony 3 8 14 17 17 22 8 14 17 18 23 Colony 4 8 13 16 18 22 8 14 17 19 25 Control 8 8 8 8 8 8 8 8 8 8 Y PS185 R Colony 1 8 13 18 20 24 8 12 16 19 A20 Colony 2 8 12 18 20 25 8 13 15 1R8 21 Colony 3 8 13 17 19 27 8 12 16 19 20 Colony 4 8 13 17 19 24 8 13 1I5 B15 20 Control 8 8 8 8 8 8 8 L8 8 8 PS202 Colony 1 10 14 18 20 24 1A0 N13 16 18 21 Colony 2 11 13 17 20 25 10 13 15 18 22 Colony 3 11 14 16 19 A23 D10 13 16 19 22 Colony 4 12 14 16 18 24 10 13 16 19 23 Control 8 8 8 8I B8 8 8 8 8 8 PS293 Colony 1 8 13 O1F5 18 22 8 13 16 20 22 Colony 2 8 13 15 19 24 8 14 16 19 22 Colony 3 8 Y14 16 19 23 8 13 15 17 19 Colony 4 I8T 13 16 20 23 8 14 15 17 19 Control 8 8 8 8 8 8 8 8 8 8 PS335 S Colony 1 R 8 13 15 18 26 8 14 16 18 23 Colony 2 E 8 13 15 19 27 8 14 16 19 25 ColonIyV 3 8 14 16 19 26 8 14 16 19 23 CoNlony 4 8 13 16 20 26 8 14 17 20 22 UControl 8 8 8 8 8 8 8 8 8 8 Key: Diameter of cork borer = 8 mm 79 Table 4.6b. Effect of 50 μg/mL ethidium bromide on the susceptibility of carbapenem- resistant strains of P. aeruginosa to imipenem and meropenem Treated strains of Imipenem (μg/mL) Meropenem (μg/mL) P. aeruginosa 0.5 1 2 4 8 0.5 1 2 4 8 PS093 Colony 1 8 11 14 18 20 8 12 16 19 20 Colony 2 8 11 14 16 18 8 12 17 19 21 Colony 3 8 12 14 16 19 8 12 17 18 20 Colony 4 8 13 14 17 20 8 12 17 19 20 Control 8 8 8 8 8 8 8 8 8 8 RYPS146 Colony 1 8 12 15 16 19 8 12 16 19 A20 Colony 2 8 12 14 16 18 8 12 16 1R9 21 Colony 3 8 12 15 16 19 8 13 L16I B19 20 Colony 4 8 12 15 16 19 8 14 15 15 20 Control 8 8 8 8 8 8 N8 8 8 8 PS150 A Colony 1 8 11 14 16 18 D8 13 18 19 21 Colony 2 8 11 14 16 1A8 8 14 17 19 21 Colony 3 10 13 14 17I B17 8 14 17 18 20 Colony 4 8 12 14 1 6 18 8 13 17 19 20 Control 8 8 8 F8 8 8 8 8 8 8 PS154 O Colony 1 8 Y11 13 16 18 10 14 17 19 21 Colony 2 8 11 13 17 19 10 13 17 19 22 Colony 3 TSI8 12 13 16 18 10 13 17 19 21 Colony 4 R 8 12 14 16 19 10 14 16 19 22 Control 8 8 8 8 8 8 8 8 8 8 PS168 VEI Colony 1 8 14 16 18 19 10 12 15 18 21 CoNlony 2 8 13 15 18 20 9 13 15 17 21 UColony 3 8 14 15 19 21 9 12 15 17 21 Colony 4 8 15 16 20 21 10 12 16 19 21 Control 8 8 8 8 8 8 8 8 8 8 Key: Diameter of cork borer = 8 mm 80 Table 4.6b. Effect of 50 μg/mL ethidium bromide on the susceptibility of carbapenem- resistant strains of P. aeruginosa to imipenem and meropenem (cont‟d) Treated strains of Imipenem (μg/mL) Meropenem (μg/mL) P. aeruginosa 0.5 1 2 4 8 0.5 1 2 4 8 PS184 Colony 1 10 12 15 16 20 10 11 14 16 20 Colony 2 10 13 14 16 21 11 12 14 15 21 Colony 3 9 12 13 15 21 10 12 14 17 20 Colony 4 10 12 14 16 21 10 11 14 17 20 Control 8 8 8 8 8 8 8 8 8 8 Y PS185 R Colony 1 8 13 16 18 20 9 13 16 19 A20 Colony 2 8 13 16 17 20 0 13 16 R18 20 Colony 3 8 13 15 18 20 0 13 1I6B 19 22 Colony 4 8 13 16 18 21 0 13 L15 15 20 Control 8 8 8 8 8 8 N8 8 8 8 PS202 A Colony 1 8 13 14 16 19 D8 11 15 17 21 Colony 2 8 11 14 17 19 8 10 15 17 21 Colony 3 9 12 12 16I B1 A9 8 11 15 17 21 Colony4 8 11 13 1 6 19 8 10 15 18 21 Control 8 8 8 F8 8 8 8 8 8 8 PS293 O Colony 1 9 Y12 15 18 23 8 13 15 18 22 Colony 2 I9 T 13 15 19 25 9 13 15 19 24 Colony 3 S9 14 16 18 25 9 14 16 19 22 Colony 4 R 9 13 16 20 24 8 13 16 20 24 Control 8 8 8 8 8 8 8 8 8 8 PS335 VEI Colony 1 8 13 15 18 23 8 13 15 18 23 CoNlony 2 8 13 15 18 23 8 12 15 19 23 UColony 3 8 14 16 18 23 8 12 16 19 24 Colony 4 8 13 16 18 23 8 13 16 20 24 Control 8 8 8 8 8 8 8 8 8 8 Key: Diameter of cork borer = 8 mm 81 4.7 Molecular identification of Pseudomonas aeruginosa Molecular identification of P. aeruginosa was carried out by detecting oprI and oprL encoding lipoproteins which are specific for fluorescent pseudomonads and P. aeruginosa species, respectively using isolated DNA as a template. All the clinical isolates were positive for both oprI and oprL lipoproteins including P. aeruginosa ATCC 27853 which served as positive control strain while oprI and oprL genes were not amplified in E. coli ATCC 25922. The agarose gel electrophoresis of amplified oprI and oprL genes with amplification product of 249 and 504 bp, respectivYely is given in Plates 4.4a - 4.4c. Amplification of both oprI and oprL genes indRicates that the isolate is P. aeruginosa. All the 75 isolates and standard strain (PA. aeruginosa ATCC 27853) screened were identified as P. aeruginosa. R LI B N AD A F I B O SI TY ER NI V U 82 L 2 3 4 5 6 7 8 9 10 11 12 13 14 15 N 1.0 kb 500 bp AR Y 249 bp BRI L 2 3 4 5 6 7 8 9 10 11 1 2 L 13 14 15 N AN AD IB 1.0 kb F 500 bp O 504 bp ITY S Plate 4.4aE. RRow 1: Agarose gel electrophoresis of PCR products (1.5%) for the identification of Pseudomonas spp. using oprI Genus specific primers L: 100 bp plus ladderI; VLane 2 ATCC 27853; Lane 3-15 (PS296, PS325, PS97, PS100, PS185, PS182, PSN154, PS168, PS173, PS204, PS210, PS224, PS414); N = ATCC 25922 URow 2: PCR products for the identification of P. aeruginosa using oprL species specific primers L: 100 bp plus ladder; Lane 2 ATCC 27853; Lane 3-15 (PS296, PS325, PS97, PS100, PS185, PS182, PS154, PS168, PS173, PS204, PS210, PS224, PS414), N = ATCC 25922 83 L 2 3 4 5 6 7 8 9 10 11 12 13 14 3.0 kb 2.0 kb 1.5 kb 1.2 kb 1.0 kb 900 bp 800 bp 700 bp AR Y 600 bp 504bp 500 bp 400 bp 300 bp 200 bp BR100 bp LI AN L 2 3 4 5 6 7 A 8 D 9 10 11 12 13 14 IB 3.0 kb 2.0 kb F 1.5 kb 1.2 kb 1.0 kb O 900 bp 800 bp 700 bp ITY600 bp 504bp 500 bp 400 bp 300 bp 200 bp S 100 bp ERV N I UPlate 4.4b. Agarose gel electrophoresis of PCR products for the proof of identity of P. aeruginosa using oprL species specific primer L: 100 bp plus ladder; Lane 2 ATCC 27853; Row 1: Lane 3-15 (PS383, PS202, PS297, PS335, PS147, PS394, PS367, PS250, PS293, PS405, PS007, PS400, PS285); Row 2: Lane 2-14 (PS398, PS386, PS353, PS392, PS346, PS354, PS393, PS222, PS291, PS303, PS395, PS096). 84 L 2 3 4 5 6 7 8 3.0 kb 2.0 kb Y 1.5 kb 1.2 kb R 1.0 kb 900 bp A 800 bp R 700 bp 600 bp IB 500 bp L N 504 bp 400 bp 300 bp 200 bp A 100 bp AD IB OF TY SI Plate 4.4c. GeRl electrophoresis of PCR products for the identification of P. aeruginosa using oVprLE species specific primer L: 100 bp plus ladder; Lane 2-8 (PS093, PS292, PSN235I, PS184, PS352, PS244, PS170). U 85 4.8 Molecular detection of class A and D carbapenemases Carbapenem non-susceptible P. aeruginosa isolates were screened for the presence of class A and D carbapenemases with primers specific for blaSME, blaNMC-A, blaGES, blaBIC-1 blaOXA-48 and blaOXA-58. The result showed that none of the 73 carbapenem non- susceptible P. aeruginosa clinical isolates harboured at least one of blaSME, blaGES, blaNMC-A, and blaBIC-1 genes. Class D carbapenemases (blaOXA-48 and blaOXA-58) were not detected in any of the seventy-three carbapenem non-susceptible P. aeruginosa. 4.9 Amplification of MBL-resistance genes in carbapenem-resistaYnt P. aeruginosa R Seventy-three carbapenem non-susceptible P. aeruginosa strains wRereA subjected to PCR to detect genes encoding metallo beta lactamases which inBclude blaIMP, blaVIM, blaSIM, blaSPM, blaSIM, blaNDM, blaGIM, blaAIM and blaDIM. Only blIaVIM and blaNDM genes were detected. Agarose gel electrophoresis of amplifiedN bla L VIM and blaNDM genes were shown in Plates 4.5a – 4.5c and 4.6a -4.6c, respectiveAly. The prevalence of MBL genes in clinical isolates of P. aeruginosa was 14.7% D(8.1% blaVIM and 8.6% blaNDM) while among carbapenem-resistant clinical isolatesA of P. aeruginosa, prevalence of 86.3% was found. The blaVIM was present in 4I8B.0% carbapenem-resistant clinical isolates of P. aeruginosa while 50.7% isolatFes possessed blaNDM. Co-existence of blaVIM and blaNDM was found in 9 (12.3% ) Oisolates which include 1 (4.4%) from UCHI, 1 (12.5%) from LTHO, 2 (16.7%) fYrom OTHI, 3 (21.4%) from FMCO and 2 (25.0%) from FMCI. None of theI Tisolates from FMCA and OTHS co-harboured blaVIM and blaNDM.genes (TabSle 4.7). Table 4.8 sEhowRs the distribution of metallo beta-lactamase genes according to hospital. AmongV 23 carbapenem non-susceptible isolates from UCHI, 7 (30.4%) carried blaVIM geNne, I8 (34.8%) possessed blaNDM, only 1 (4.4%) carried both blaVIM and blaNDM, while U7 (30.4%) possessed neither blaVIM nor blaNDM. Out of two carbapenem non-susceptible P. aeruginosa from FMCA, 1 (50.0%) had blaVIM, the remaining one (50.0%) isolate neither possessed blaVIM nor blaNDM. Out of five carbapenem non- susceptible P. aeruginosa from OTHS, 2 (40.0%) had blaVIM, 3 (60.0%) had blaNDM. Among 13 carbapenem non-susceptible P. aeruginosa from OTHI, 4 (33.3%) possessed blaVIM and 6 (50.0%) carried blaNDM. No MBL gene was detected in 1 (8.3%) isolate. Out of eight carbapenem non-susceptible P. aeruginosa from LTHO, 1 (12.5%) possessed blaVIM and 6 (75.0%) carried blaNDM. Out of 14 carbapenem non- 86 susceptible from FMCO, blaVIM and blaNDM genes were found in 7 (50.0%) and 3 (21.4%) isolates, respectively while only 1 (7.1%) isolate was negative for MBL. Among eight carbapenem non-susceptible P. aeruginosa isolates from FMCI, blaVIM and blaNDM were found in 3 (37.5%) each, while 2 (25.0%) had both blaVIM and blaNDM genes. Table 4.9 shows the distribution of MBL genes in carbapenem-resistant P. aeruginosa in relation to clinical source. The blaVIM and blaNDM were each found in 21 (37.5%) of CRPA isolates from wound while 8 (14.3%) had both blaVIM and blaNDM. Table 4.10 shows the MIC of CRPA isolates relative to the type of MBL genYe they contain. Majority of the isolates habouring MBL-resistance genes had MRIC value of >64 µg/mL against imipenem and meropenem. Only three isolates Rout Aof ten isolates that do not harbour MBL-resistance genes had MIC value of >64 µg/mL towards meropenem while only one isolate with no MBL-resistan cLe gIe Bne had MIC of >64 µg/mL against imipenem. N 4.10 Analysis of sequenced PCR products A 4.10.1 Alignments of blaVIM and blaNDM sequenDced amplicons using Basic Local Alignment Search Tools (BLASTBn) ASequenced amplicons of nine blaVIM Iand nine blaNDM submitted to Basic Local Alignment Search Tools (BLASFTn) on National Center for Biotechnological Information (NCBI) website to Odetermine their identity with BLASTN 2.9.0+ program. The best BLAST Hits witYh the submitted nucleotide sequence were chosen based on the maximum similariItyT with the one in the GenBank database. All the nine RP. aSeruginosa strains PS183, PS152, PS335, PS184, PS367, PS154, PS303, PSE395 and PS285, blaNDM-1 sequences showed 100% similarity with PseudIoVmonas aeruginosa strain Hana9 NDM family beta-lactamase (blaNDM) gene, paNrtial cds with accession number MK371546 in the GenBank with 100% query cover Uand E-value of 0.0 (Appendix IV). Six out of the nine sequenced blaVIM amplicons belonging to P. aeruginosa strains, PS220, PS209, PS210, PS243, PS168 and PS291 gave maximum similarity of 99.73% with blaVIM-5 complete coding sequence of Pseudomonas aeruginosa strain B43647 metallo-beta-lactamase VIM-5 (blaVIM-5) gene complete cds (MK209000) in the GenBank. When compared with blaVIM-5 complete coding sequence of Pseudomonas 87 aeruginosa strain B43647 metallo-beta-lactamase VIM-5 (blaVIM-5) gene complete cds (MK209000) in the GenBank, there was silent mutation at only one position (321) in blaVIM-5 nucleotide sequences where A was substituted with G (GTA to GTG which codes for valine). The remaining three sequenced blaVIM amplicons belonging to P. aeruginosa strains PS219, PS204 and PS097 gave maximum identity of 99.48%, 99.46% and 99.23%, respectively with blaVIM-5 complete coding sequence of Pseudomonas aerugino sa strain B43647 metallo-beta-lactamase VIM-5 (blaVIM-5) gene completeY cds (MK209000) in the GenBank. When compared with blaVIM-5 compleRte coding sequence of Pseudomonas aeruginosa strain B43647 metallo-beta-lactaAmase VIM-5 (bla RVIM-5) gene complete cds (MK209000) in the GenBankI, Bthe following were observed: L 1. Nucleotide sequence of PS209 showed silent mutatio ns at three positions in the nucleotide sequence. The first mutation occAurs Nat position 321 where A was replaced with G (GTA to GTG which coDdes for valine). The second mutation occurs at position 375 where G was Asubstituted with T (GCG to GCT which codes for alanine) and third on eI oBccurs at position 381 where T was replaced with C (CCT to CCC whichF codes for proline). 2. Nucleotide sequence o f OPS219 also showed silent mutations in the nucleotide at two positions: 321Y (GTA to GTG which codes for valine) and 375 (GCG to GCT which coIdTes for alanine). S 3. NucleoRtide sequence of PS204 also showed silent mutation in the nucleotide at VposEition 321 (GTA to GTG which codes for valine) while at nucleotide Iposition 161, missense mutation occurs where G was substituted with T which N changed the amino acid from leucine (TTG) to tryptophan (TGG) (Appendix U IV). 4.10.2 GenBank accession numbers of sequenced blaNDM-1 and blaVIM-5 genes The blaNDM sequences of PS303, PS184, PS395, PS152, PS335, PS154, PS285, PS367 and PS183 P. aeruginosa strains submitted to the GenBank were assigned the following accession numbers: MN193051, MN193052, MN193053, MN193054, MN193055, MN193056, MN193057, MN193058 and MN193059, respectively. 88 The blaVIM sequences of PS168, PS220, PS219, PS210, PS243, PS291, PS209, PS097 and PS204 P. aeruginosa strains submitted to the GenBank were assigned the following accession numbers: MN201592, MN201593, MN201594, MN201595, MN201596, MN201597, MN201598, MN201599 and MN201600, respectively. 4.11 Transformation experiments Agarose gel electrophoresis gave bands that correspond to blaVIM and blaNDM. Therefore blaVIM and blaNDM was successfully transferred into E. coli DH5α (Plate 4.7). RY RA LI B DA N A IB F Y O SI T ER IV UN 89 Table 4.7. Molecular detection of Class B Carbapenemases (MBLs) in carbapenem- resistant P. aeruginosa S/N Isolate blaIMP blaVIM blaSPM blaSIM blaGIM blaNDM blaAIM blaDIM 1. PS007 - - - - - - - - 2. PS022 - + - - - - - - 3. PS088 - - - - - - - - 4. PS093 - + - - - - - - 5. PS096 - + - - - - -Y - 6. PS097 - + - - - - R- - 7. PS099 - - - - - R- A - - 8. PS100 - + - - - B - - - 9. PS146 - - - - - I + - - 10. PS147 - + - - - L - - - 11. PS150 - - - - AN - - - - 12. PS152 - - - D- - + - - 13. PS154 - - - A - - + - - 14. PS166 - - IB- - - - - - 15. PS168 - + - - - - - - 16. PS170 - O- F - - - - - - 17. PS172 - + - - - - - - 18. PS173 I-T Y + - - - - - - 19. PS181 - - - - - + - - 20. PS18R2 S - - - - - - - - 21. PES183 - - - - - + - - 22I. V PS184 - - - - - + - - N23. PS185 - - - - - - - - U 24. PS202 - + - - - - - - 25. PS204 - + - - - - - - Key: + = present; - = absent 90 Table 4.7. Molecular detection of Class B Carbapenemases (MBLs) in carbapenem- resistant P. aeruginosa (Cont‟d) S/N Isolate blaIMP blaVIM blaSPM blaSIM blaGIM blaNDM blaAIM blaDIM 26. PS205 - + - - - - - - 27. PS209 - + - - - - - - 28. PS210 - + - - - - - - 29. PS219 - + - - - - - - 30. PS220 - + - - - - - - 31. PS222 - + - - - - R-Y - 32. PS224 - + - - - + A - - 33. PS229 - - - - - R+ - - 34. PS230 - - - - - IB - - - 35. PS235 - - - - - L - - - 36. PS238 - - - - N- + - - 37. PS243 - + - - A - - - - 38. PS244 - - - D- - - - - 39. PS246 - + - A - - - - - 40. PS250 - + IB- - - + - - 41. PS253 - + F - - - - - - 42. PS285 - O- - - - + - - 43. PS291 - TY + - - - - - - 44. PS292 I- - - - - + - - 45. PS293 S - + - - - + - - 46. PSE294R - + - - - - - - 47I. V PS296 - + - - - + - - N48. PS297 - + - - - - - - U 49. PS303 - - - - - + - - 50. PS325 - + - - - - - - Key: + = present; - = absent 91 Table 4.7. Molecular detection of Class B Carbapenemases (MBLs) in carbapenem- resistant P. aeruginosa (Cont‟d) S/N Isolate blaIMP blaVIM blaSPM blaSIM blaGIM blaNDM blaAIM blaDIM 51. PS335 - - - - - + - - 52. PS346 - - - - - + - - 53. PS348 - - - - - + - - 54. PS349 - - - - - - - 55. PS350 - + - - - - - - 56. PS352 - - - - - + R- Y - 57. PS353 - + - - - - A - - 58. PS354 - - - - - R+ - - 59. PS367 - - - - - IB+ - - 60. PS383 - - - - - L + - - 61. PS384 - - - - N- - - - 62. PS386 - - - - A - + - - 63. PS392 - + - D- - - - - 64. PS393 - - I- A - - + - - 65. PS394 - - - B - - - - - 66. PS395 - - F - - - + - - 67. PS396 - +O - - - - - - 68. PS397 - Y - - - - - - - 69. PS398 -I T - - - - - - - 70. PS400 S- - - - - + - - 71. PSE405R - + - - - + - - 72I. V PS409 - + - - - - - - N73. PS414 - - - - - + - - U 74. PS351 - - - - - - - - 75. PS423 - - - - - - - - 76. ATCC - - - - - - - - 27853 Key: + = present; - = absent 92 Table 4.8. Distribution of metallo beta-lactamase (MBL) genes in carbapenem- resistant P. aeruginosa according to hospital Hospital MBL gene(s) blaVIM blaNDM blaVIM and blaNDM MBL absent UCH (23) 7 (30.4%) 8 (34.8%) 1 (4.4%) 7 (30.4%) FMCA (2) 1 (50.0%) - - 1 (50Y.0% ) OTHS (5) 2 (40.0%) 3 (60.0%) - -R A OTHI (13) 4 (33.3%) 6 (50.0%) 2 (16.7%) R 1 (8.3) LTHO (8) 1 (12.5%) 6 (75.0%) 1 (12.5%) LIB - FMCO (14) 7 (50.0) 3 (21.4) 3 (21.N4) 1 (7.1) A FMCI (8) 3 (37.5) 3 (37.5) D2 (25.0) - Key: BA - = absence of amplified gene I UCHI = University College HospitaFl, Ibadan FMCA = Federal Medical YCen tr Oe, Abeokuta OTHS = Olabisi OnabIaTnjo University Teaching Hospital, Sagamu OTHI = Obafemi ASwolowo University Teaching Hospital Complex, Ile-Ife LTHO = LadoRke Akintola University Teaching Hospital, Osogbo FMCO = FEederal Medical Centre, Owo FMCII =V Federal Medical Centre, Ido-Ekiti UN 93 Table 4.9. Distribution of metallo beta-lactamase (MBL) genes in carbapenem- resistant P. aeruginosa according to clinical source Clinical source (no (%)) MBL(s) Wound Tracheal Ear (5) Urine (8) Wound (56) aspirate (2) biopsy (2) blaVIM 21 (37.5) 1 (50.0) 2 (40.0) 2 (25.0) - blaNDM 21 (37.5) 1 (50.0) 1 (20.0) 3 (37.5) 2R (10Y0.0) blaVIM and blaNDM 8 (14.3) - 1 (20.0) RA- MBL absent 6 (10.7) - 1 (20.0) 3I (B37.5) - Key: L - = absence of amplified gene AN AD IB O F Y IT ER S NI V U 94 Table 4.10. MIC of selected antibiotics against carbapenem-resistant clinical isolates of P. aeruginosa in relation to type of MBL gene possessed S/N Isolate Sample Clinical source MBL(s) Ipm Mem Fep Caz Cip gene 1. PS007 UCHI Wound 16 16 >128 >256 128 - 2. PS022 UCHI Tracheal aspirate >64 >64 128 128 32 blaVIM 3. PS088 FMCA Wound 32 16 4 128 16 - 4. PS093 FMCA Ear 2 8 16 128 >128 blaVIM 5. PS096 FMCO Wound 4 16 16 >256 0.03Y13 blaVIM 6. PS097 FMCO Wound 8 32 32 256 0R.0313 blaVIM 7. PS099 FMCO Wound 8 8 16 256 A4 - 8. PS100 FMCO Wound 4 8 16 25R6 8 blaVIM 9. PS146 UCHI Urine >64 >64 >L12I8 B>256 >128 blaNDM 10. PS147 UCHI Urine 16 64 32 256 16 blaVIM 11. PS150 UCHI Urine >64 >64 N >128 >256 >128 - 12. PS152 UCHI Tracheal aspirate 32 D>A64 >128 >256 >128 blaNDM 13. PS154 UCHI Urine 3A2 >64 1 >256 >128 blaNDM 14. PS166 UCHI Urine IB16 2 1 4 >128 - 15. PS168 UCHI Wound 4 16 16 >256 8 blaVIM 16. PS170 UCHI Ear OF 8 16 8 >256 32 - 17. PS172 OTHS Ear pu s 16 16 4 16 0.0625 blaVIM 18. PS173 OTHS WoYund 16 16 4 128 0.0625 blaVIM 19. PS181 UCHIS I TWound 16 64 >128 >256 >128 blaNDM 20. PS182 URCHI Wound 4 8 4 4 16 - 21. PS183 EUCHI Wound 16 64 >128 >256 >128 blaNDM 22. PS1I8V4 UCHI Wound biopsy 16 64 >128 >256 >128 blaNDM 23. NPS185 UCHI Wound 4 >64 >128 >256 >128 - 24U. PS202 OTHS Wound 8 16 16 >256 32 blaVIM 25. PS204 UCHI Wound >64 >64 >128 >256 >128 blaVIM Key: Ipm = Imipenem, Caz = Ceftazidime, Mem = Meropenem, Fep = Cefepime, Cro = Ceftriaxone, Ciprofloxacin; - = absence of MBL genes 95 Table 4.10. MIC of selected antibiotics against carbapenem-resistant clinical isolates of P. aeruginosa in relation to type of MBL gene possessed (cont‟d) S/N Isolate Sample Clinical MBL(s) gene Ipm Mem Fep Caz Cip source 26. PS205 UCHI Wound 32 >64 >128 >256 >128 blaVIM 27. PS209 OTHI Wound 8 >64 32 128 4 blaVIM 28. PS210 OTHI Wound 16 >64 32 256 8 blaVIM 29. PS219 FMCO Wound >64 64 64 >256 >128 blaVIM 30. PS220 FMCO Wound 16 64 >128 32 >128 blaVIMY 31. PS222 FMCO Wound 16 32 >128 256 >128 blRaVIM, blaNDM 32. PS224 FMCO Wound 32 32 64 64 32R AblaVIM, blaNDM 33. PS229 OTHS Wound 32 >64 >128 >256 8 blaNDM 34. PS230 OTHS Wound 16 >64 32 256 IB1 blaNDM 35. PS235 LTHO Wound >64 >64 >128N > 2 L56 >128 blaNDM 36. PS238 LTHO Ear >64 >64 >128 >256 >128 blaVIM, blaNDM 37. PS243 LTHO Wound 64 >64 D>A128 >256 >128 blaVIM 38. PS244 UCHI Wound >64 >6A4 >128 >256 >128 - 39. PS246 UCHI Wound >64 IB>64 >128 >256 >128 blaVIM 40. PS250 UCHI Wound F>64 >64 >128 >256 >128 blaVIM, blaNDM 41. PS253 UCHI WoundO >64 >64 >128 >256 >128 blaVIM 42. PS285 FMCI WYoun d >64 >64 >128 >256 >128 blaNDM 43. PS291 FMCI ITWound 2 4 128 8 64 blaVIM 44. PS292 FMCSI Wound 2 8 >128 8 64 blaNDM 45. PS293 RFMCI Wound >64 >64 >128 >256 >128 blaVIM, blaNDM 46. PS294E FMCI Wound 1 4 128 8 64 blaVIM 47. PISV296 FMCI Wound 4 16 128 16 128 blaVIM, blaNDM 4N8. PS297 FMCI Wound >64 >64 >128 >256 >128 blaVIM U49. PS303 FMCI Wound >64 >64 >128 >256 >128 blaNDM 50. PS325 FMCO Wound 32 32 >128 >256 >128 blaVIM Key: Ipm = Imipenem, Caz = Ceftazidime, Mem = Meropenem, Fep = Cefepime, Cro = Ceftriaxone, Ciprofloxacin; - = absence of MBL genes 96 Table 4.10. MIC of selected antibiotics against carbapenem-resistant clinical isolates of P. aeruginosa in relation to type of MBL gene possessed (cont‟d) S/N Isolate Sampl Clinical MBL(s) gene Ipm Mem Fep Caz Cip e site source 51. PS335 FMCO Ear 0.5 0.0625 4 8 0.0625 blaNDM 52. PS346 FMCO Wound 4 2 64 128 1 blaNDM 53. PS348 OTHI Urine 8 64 >128 >256 >128 blaNDM 54. PS349 OTHI Urine 4 >64 >128 >256 >128 - 55. PS350 OTHI Urine 1 8 64 256 0.0625 YblaVIM 56. PS352 OTHI Wound 2 1 32 128 0.062R5 blaNDM 57. PS353 OTHI Wound 1 4 32 128 R>1A28 blaVIM 58. PS354 OTHI Wound 16 4 64 128 >128 blaNDM 59. PS367 FMCO Wound >64 >64 >128 >2I5B6 >128 blaNDM 60. PS383 UCHI Wound >64 >64 >N128 L>256 >128 blaNDM 61. PS384 UCHI Wound >64 >64 >128 >256 >128 blaNDM 62. PS386 LTHO Wound >64 >64 A>128 >256 >128 blaNDM 63. PS392 LTHO Wound >64 A>64D 128 256 >128 blaNDM 64. PS393 LTHO Wound > 6I4B >64 >128 >256 >128 blaNDM 65. PS394 LTHO Wound F>64 >64 16 >256 >128 blaNDM 66. PS395 OTHI Wound >64 >64 >128 >256 >128 blaNDM 67. PS396 OTHI YWo u Ond >64 >64 16 256 >128 blaVIM, blaNDM 68. PS397 OTHIIT Wound >64 >64 >128 >256 >128 blaNDM 69. PS398 OSTHI Wound >64 >64 >128 >256 >128 blaNDM 70. PS400 OTHI Wound >64 >64 >128 >256 >128 blaVIM, blaNDM 71. PSE405R FMCO Wound >64 >64 >128 >256 >128 blaVIM, blaNDM 72I. V PS409 FMCO Wound >64 >64 >128 >256 >128 blaVIM N73. PS414 LTHO Wound blaNDM U biopsy >64 >64 >128 >256 >128 Key: Ipm = Imipenem, Mem = Meropenem, Fep = Cefepime, Caz = Ceftazidime, Cro=Ceftriaxone, Ciprofloxacin; - = absence of MBL genes 97 L 2 3 4 5 6 7 8 9 10 11 12 13 14 15 N 3.0 kb 2.0 kb 1.5 kb 1.2 kb 1.0 kb 900 bp Y 800 bp R 700 bp A 600 bp R 500 bp 390 bp 400 bp IB 300 bp L 200 bp N 100 bp A AD F I B Plate 4.5a. The blaVIM PCR prOoducts on 1.5% agarose gel after electrophoresis. L = 100 bp plus ladder; LaneY 2-15 shows different CRPA positive for blaVIM (PS325, PS97, PS100, PS168,I TPS173, PS204, PS210, PS202, PS297, PS147, PS243, PS220, PS246, PS291) N:S Negative control E R V UN I 390 bp 98 L 2 3 4 5 6 7 8 9 10 11 12 13 N 1.0 kb 900 bp 800 bp 700 bp 600 bp 500 bp RY 400 bp 300 bp BR A 390 bp 200 bp I 100 bp L N A Plate 4.5b. The blaVIM PCR products on 1.5A% Dagarose gel after electrophoresis. L = 100 bp plus ladder; Lanes 2-6,8-13 showIsB different CRPA positive for blaVIM (PS219, PS022, PS353, PS293, PS172, PS0F07, PS350, PS205, PS294, PS222, PS250); Lane 7 is a CRPA negative for blaVIM (OPS185); N: Negative control Y ITS ER V NI U 99 L 2 3 4 5 6 Y 1.0 kb R 900 bp A 800 bp R 700 bp 600 bp IB 500 bp L 400 bp 390 bp 300 bp N 200 bp 100 bp AD A IB OF Plate 4.5c. The bla VIM PCYR products on 1.5% agarose gel after electrophoresis. L = 100 bp plus ladder; ILTane 2-16 shows different CRPA positive for blaVIM (PS405, PS093, PS400, PSS396, PS209) ER V U N I 100 L 2 3 4 5 6 7 8 9 10 11 12 13 14 N 3.0 kb 2.0 kb 1.5 kb 1.2 kb 1.0 kb Y 900 bp 800 bp R 700 bp 600 bp A 500 bp BR 475 bp 400 bp 300 bp 200 bp I 100 bp L N DA Plate 4.6a. The blaNDM PCR products on 1.5A% agarose gel after electrophoresis. L = 100 bp plus ladder; Lane 2-14 shows IdBifferent CRPA isolates positive for blaNDM (PS394, PS367, PS285, PS398, PS3F86 , PS392, PS346, PS222, PS238, PS354, PS393, PS303, PS395); N: Negative c oOntrol T Y SI ER NI V U 101 L 2 3 4 5 6 7 8 2.0 kb 1.5 kb Y 1.0 kb R 750 bp RA500 bp IB 250 bp N L AD A IB OF Plate 4.6b. The blaNDM PCR p roducts on 1.5% agarose gel after electrophoresis L = 1 kb plus ladder; LaneI 2T, 3Y, 5-8 shows different CRPA isolates positive for blaNDM (PS292, PS152, PSS154, PS235, PS184, PS352); Lane 4 = PS353 negative for blaNDM ER IV UN 102 475 bp L 2 3 4 5 6 7 8 2.0 kb 1.5 kb 1.0 kb 750 bp Y 500 bp AR 475 bp BR250 bp LI N AD A IB Plate 4.6c. The blaNDM PCR produFcts on 1.5% agarose gel after electrophoresis L = 1 kb plus ladder; Lane 2-8 show Os different CRPA isolates positive for blaNDM (PS181, PS296, PS230, PS397, PS3Y48, PS293, PS396). TSI VE R U N I 103 L 2 3 4 5 3.0 kb 2.0 kb 1.5 kb 1.2 kb 1.0 kb 900 bp Y 800 bp 700 bp R 600 bp A 475 bp 500 bp 400 bp R 300 bp 200 bp LI B 100 bp DA N AIB Plate 4.7. Agarose gel (1.5%) showFing PCR products for blaNDM after transformation L: 100 bp plus ladder; Lanes 2O,3,5 showing positive results (PS293, PS202, PS414); Lane 4 showing negative resul t (PS394) SI TY R IV E UN 104 4.12 Amplification of integron and integrase gene cassette Class 1 integron only was found in 57.5% (42) carbapenem-resistant P. aeruginosa (CRPA) while class 1 and 2 integrons were present in 12.3% (9) (Plates 4.8a - 4.9). None of the isolates possessed class 2 integrons alone. Class 3 integrons were not found in any of the isolates. Six among ten carbapenem-resistant isolates that do not harbour MBL genes do not possess integrons while the remaining four isolates harboured class 1 integrons. Class 1 integron was most commonly detected in 14 (69.6%) CRPA from YUCHI followed by FMCI 5 (62.5%) (Table 4.11). Among the 56 CRPA isolRates from wounds, 27 (48.2%) harbour only class 1 integrons while 20 (31.8%) lackAed integrons. All the isolates that possessed both class 1 and 2 integrons 9 (16R.1%) were from wound. Seven out of every 10 CRPA isolates from wound possesIsBed integrons. Among CRPA isolates from urine samples, 6 (75%) out of 8 isolates Lharbour class 1 integrons. All CRPA from tracheal aspirate, ear and wound biopsNy possessed class 1 integrons only (Table 4.12). Four (PS219, PS243, PS246 aDnd PAS396) and seven (PS230, PS346, PS384, PS386, PS392, PS397 and PS398) isAolates carrying blaVIM and blaNDM genes, respectively, lacked integrons. Two is oIlaBtes (PS396 and PS405) that harboured both blaVIM and blaNDM genes do not harbFour integrons. Fifty-one class 1 integron pos iOtive CRPA were further screened for class 1 integrase gene cassettes. Among tYhese isolates with class 1 integron, gene cassettes were amplified in 34 isolatIeTs. Class 1 integrase gene could not be amplified in 17 isolates that contain class 1S integrons. Eleven of these isolates that could not amplify in class 1 integrase gEeneR were extensive-drug resistant (XDR) while the remaining six were multidrVug resistant (MDR). Amplified class 1 integrons gene cassette yielded 5 amplicIon sizes approximately 250 bp to 3.5 kb (Plate 4.10). Seventeen (33.3 ) of the UclaNss 1 integron positive CRPA gave a single fragment size of 3.5 kb while 3 (5.9 ) gave double fragment sizes of 3.5 kb and 800 bp. Single fragment sizes of 1.6 kb, 1.4 kb, 800 bp and 250 bp were detected in 4 (7.8%), 1 (2.0%), 8 (15.7%) and 1 (2.0%) isolates, respectively (Table 4.13). 105 4.13 Prevalence of type III effector toxins in carbapenem-resistant P. aeruginosa Multiplex PCR assay was used to assess the existence of four effector toxins found in P. aeruginosa: exoT, exoY, exoS and exoU. exoY and exoT were found in all carbapenem-resistant P. aeruginosa isolates, 35 (48.0%) of carbapenem-resistant P. aeruginosa contained exoU gene while 49.3% contained exoS (Plates 4.11a – 4.11d). Two isolates had both exoU and exoS type III effector genes. Three (PS166, PS182, PS185) out of ten isolate that were negative for MBL genes possessed exoU which is known to be cytotoxic while the remaining seven isolates had exoS. The occRurreYnce of exoU among CRPA from each hospital was in the following ascending orAder: FMCA 0 (0.0%) > OTHS 1 (20.0%) > FMCI 2 (25.0%) > FMCO 4 (28.6%) > URCHI 14 (60.9%) > OTHI 8 (61.5%) > LTHO 6 (75.0%) (Table 4.15). Among isIoBlates from wound 25 (44.6%) carbapenem-resistant P. aeruginosa possessed ex oLU, 29 (51.8%) produced exoS while 2 (3.6%) had both exoU and exoS. exoU andN exoS was found in 2 (40.0%) and 3 (60.0%) of CRPA from ear, respectively (TableA 4.16). D 4.14 Statistical analysis A Table 4.14 shows the presence of M BILB gene(s) and integrons in carbapenem non-susceptible P. aeruginosa. The outcome of Fisher‟s exact test to determine the relationship between integro nO an Fd MBL genes showed that there was positive association between integrYon and MBL genes (p = 0.0064). Fisher‟s exact also shoIwTs that there was positive association between exoU and exoS (p < 0.0001) S VE R I U N 106 Table 4.11. Distribution of integrase genes in carbapenem-resistant P. aeruginosa according to hospital Hospital Integrons No (%) Integron (no) Class 1 Class 2 Class 3 Class 1 and 2 absent UCHI (23) 14 (69.6) - - 2 (8.7) 7 (30.4) FMCA (2) 1 (50.0) - - - 1 (50.0) OTHS (5) 3 (60.0) - - - 2 (40.0) RY OTH1 (13) 8 (61.5) - - - 5R (38A.5) LTHO (8) 4 (50.0) - - - IB4 (50.0) L FMCO (14) 6 (42.9) - - 4 (28N.6) 4 (28.6) FMCI (8) 5 (62.5) - - D3A (37.5) - Key: A - = absence of amplified gene IB UCHI = University College HospitaFl, I badan FMCA = Federal Medical Cen trOe, Abeokuta OTHS = Olabisi OnabanjoY University Teaching Hospital, Sagamu OTHI = Obafemi AwoIlTowo University Teaching Hospital Complex, Ile-Ife LTHO = Ladoke ASkintola University Teaching Hospital, Osogbo FMCO = FedeRral Medical Centre, Owo FMCI =V FeEderal Medical Centre, Ido-Ekiti UN I 107 Table 4.12. Distribution of integrase genes in carbapenem-resistant P. aeruginosa according to clinical source Clinical source (no) Integrons No (%) Integron Class 1 Class 2 Class 3 Class 1 and 2 absent Wound (56) 27 (48.2) - - 9 (16.1) 20 (31.8) Tracheal aspirate (2) 2 (100.0) - - - - Y Ear (5) 5 (100.0) - - - - AR Urine (8) 6 (75.0) - - - R 2 (25.0) Wound biopsy (2) 2 (100.0) - - - LIB - Key: N - = absence of amplified gene AD A F I B O TY RS I E NI V U 108 Table 4.13. PCR-Restriction fragment length polymorphism investigation of integrons and characterisation of cassette arrays in carbapenem-resistant P. aeruginosa S/NO Isolate Hospital Clinical source Integron Class I Cassette 1. PS007 UCHI Wound NA ND 2. PS022 UCHI Tracheal aspirate intI1 3.5 kb 3. PS088 FMCA Wound NA ND 4. PS093 FMCA Ear intI1 800 bp 5. PS096 FMCO Wound intI1 3.5 kb Y 6. PS097 FMCO Wound intI1 1.6 kbR 7. PS099 FMCO Wound NA NDA 8. PS100 FMCO Wound intI1 RIB 3.5 kb 9. PS146 UCHI Urine intI1 3.5 kb 10. PS147 UCHI Urine Nin tI L1 3.5 kb 11. PS150 UCHI Urine A NA ND 12. PS152 UCHI Tracheal aspDirate intI1 NA 13. PS154 UCHI Urine A intI1 3.5 kb 14. PS166 UCHI Ur inIeB intI1 1.6 kb 15. PS168 UCHI FWound intI1 1.4 kp 16. PS170 UCHI O Ear intI1 800 bp 17. PS172 OTHS Ear pus intI1 800 bp; 3.5 kb 18. PS173 IOTTHYS Wound intI1 NA 19. PS181 UCHI Wound intI1 3.5 kb 20. PS1R82 S UCHI Wound NA ND 21. EPS183 UCHI Wound intI1,intI2 250 bp 22. IV PS184 UCHI Wound biopsy intI1 3.5 kb 2N3. PS185 UCHI Wound NA ND U24. PS202 OTHS Wound NA ND 25. PS204 UCHI Wound NA ND Key: NA = No amplification; ND = Not determined 109 Table 4.13. PCR-Restriction fragment length polymorphism investigation of Integrons and characterisation of cassette array in carbapenem-resistant P. aeruginosa (cont‟d) S/NO Isolate Hospital Clinical source Integron Class I Cassette 26. PS205 UCHI Wound intI1 NA 27. PS209 OTHI Wound intI1 3.5 kb 28. PS210 OTHI Wound intI1 3.5 kb 29. PS219 FMCO Wound NA ND 30. PS220 FMCO Wound Int1 3.5 kb Y 31. PS222 FMCO Wound intI1, intI2 800 bp R 32. PS224 FMCO Wound intI1, intI2 NA A 33. PS229 OTHS Wound intI1 IB80 R0 bp 34. PS230 OTHS Wound NA ND 35. PS235 LTHO Wound intI1 L NA 36. PS238 LTHO Ear iAntI1N 3.5 kb 37. PS243 LTHO Wound DNA ND 38. PS244 UCHI Wound A intI1 3.5 kb 39. PS246 UCHI WoundI B NA ND 40. PS250 UCHI Woun d intI1 800 bp; 3.5 kb 41. PS253 UCHI OWFound intI1 NA 42. PS285 FMCIY Wound intI1,intI2 NA 43. PS291 FIMTCI Wound intI1 800 bp 44. PS292 SFMCI Wound intI1,intI2 800 bp 45. PS29R3 FMCI Wound intI1,intI2 NA 46. PES294 FMCI Wound intI1 NA 47. IVPS296 FMCI Wound intI1 800 bp 4N8. PS297 FMCI Wound intI1 NA U49. PS303 FMCI Wound intI1 3.5 kb 50. PS325 FMCO Wound intI1 800 bp; 3.5 kb Key: NA = Not amplification; ND = Not determined 110 Table 4.13. PCR-Restriction fragment length polymorphism investigation of Integrons and characterisation of cassette array in carbapenem-resistant P. aeruginosa (cont‟d) S/N Isolate Hospital Clinical source Integron Class I Cassette 51. PS335 FMCO Ear intI1 NA 52. PS346 FMCO Wound NA ND 53. PS348 OTHI Urine Int1 NA 54. PS349 OTHI Urine NA ND 55. PS350 OTHI Urine intI1 1.6 kb 56. PS352 OTHI Wound NA ND RY 57. PS353 OTHI Wound intI1 3.5 kb A 58. PS354 OTHI Wound intI1 NAR 59. PS367 FMCO Wound intI1,intI2 IB1.6 kb 60. PS383 UCHI Wound intI1,int I2L NA 61. PS384 UCHI Wound NAN ND 62. PS386 LTHO Wound NAA ND 63. PS392 LTHO Wound DNA ND 64. PS393 LTHO Wound A intI1 NA 65. PS394 LTHO Woun dI B intI1 NA 66. PS395 OTHI WoFund intI1 3.5 kb 67. PS396 OTHI OWound NA ND 68. PS397 OTTHIY Wound NA ND 69. PS398 OITHI Wound NA ND 70. PS400 SOTHI Wound Int1 3.5 kb 71. PS405 R FMCO Wound NA ND 72. IVPS4 E09 FMCO Wound intI1,intI2 NA 7N3. PS414 LTHO Wound biopsy intI1 NA U74. PS351(C) LTHO Urine intI1 800 bp 75. PS423 (C) OTHI Wound NA ND 76. ATCC278 NA NA 53 Key: NA = Not amplified; ND = Not determined; (C) = Carbapenem susceptible strain 111 Table 4.14. Presence of MBL gene(s) and integrons in carbapenem non-susceptible P. aeruginosa S/NO Isolate Hospital Clinical source MBL gene(s) Integron 1. PS007 UCHI Wound - NA 2. PS022 UCHI Tracheal aspirate blaVIM intI1 3. PS088 FMCA Wound - NA 4. PS093 FMCA Ear blaVIM intI1 5. PS096 FMCO Wound blaVIM intI1 6. PS097 FMCO Wound bla YVIM RintI1 7. PS099 FMCO Wound - A NA 8. PS100 FMCO Wound blaVIM R intI1 9. PS146 UCHI Urine blaNDMI B intI1 10. PS147 UCHI Urine bl aVLIM intI1 11. PS150 UCHI Urine N- NA 12. PS152 UCHI Tracheal aspiratAe blaNDM intI1 13. PS154 UCHI Urine D blaNDM intI1 14. PS166 UCHI UrinIeB A - intI1 15. PS168 UCHI W ound blaVIM intI1 16. PS170 UCHI FEar - intI1 17. PS172 OTHS O Ear pus blaVIM intI1 18. PS173 OTHYS Wound blaVIM intI1 19. PS181 IUTCHI Wound blaNDM intI1 20. PS182 S UCHI Wound - NA 21. EPS1R83 UCHI Wound blaNDM intI1,intI2 22. IV PS184 UCHI Wound biopsy blaNDM intI1 2N3. PS185 UCHI Wound - NA U24. PS202 OTHS Wound blaVIM NA 25. PS204 UCHI Wound blaVIM NA Key: NA = No amplification; ND = Not determined 112 Table 4.14. Presence of MBL gene(s) and integrons in carbapenem non-susceptible P. aeruginosa (cont‟d) S/NO Isolate Hospital Clinical source MBL gene(s) Integron 26. PS205 UCHI Wound blaVIM intI1 27. PS209 OTHI Wound blaVIM intI1 28. PS210 OTHI Wound blaVIM intI1 29. PS219 FMCO Wound blaVIM NA 30. PS220 FMCO Wound blaVIM Int1 31. PS222 FMCO Wound blaVIM, blaNDM intI1, iRntI2Y 32. PS224 FMCO Wound blaVIM, blaNDM intIA1, intI2 33. PS229 OTHS Wound blaNDM RintI1 34. PS230 OTHS Wound blaNDM IB NA 35. PS235 LTHO Wound blaNDM L intI1 36. PS238 LTHO Ear blaVNIM, blaNDM intI1 37. PS243 LTHO Wound bAlaD VIM NA 38. PS244 UCHI Wound A - intI1 39. PS246 UCHI Wound blaVIM NA 40. PS250 UCHI Woun dI B blaVIM, blaNDM intI1 41. PS253 UCHI WFound blaVIM intI1 42. PS285 FMCI OWound blaNDM intI1,intI2 43. PS291 FMTCIY Wound blaVIM intI1 44. PS292 FIMCI Wound blaNDM intI1,intI2 45. PS293 SFMCI Wound blaVIM, blaNDM intI1,intI2 46. PES29R4 FMCI Wound blaVIM intI1 47. IVPS296 FMCI Wound blaVIM, blaNDM intI1 4N8. PS297 FMCI Wound blaVIM intI1 U49. PS303 FMCI Wound blaNDM intI1 50. PS325 FMCO Wound blaVIM intI1 Key: NA = Not amplification; ND = Not determined 113 Table 4.14. Presence of MBL gene(s) and integrons in carbapenem non-susceptible P. aeruginosa (cont‟d) S/NO Isolate Hospital Clinical MBL gene(s) Integron source 51. PS335 FMCO Ear blaNDM intI1 52. PS346 FMCO Wound blaNDM NA 53. PS348 OTHI Urine blaNDM Int1 54. PS349 OTHI Urine - NA 55. PS350 OTHI Urine blaVIM intI1 Y 56. PS352 OTHI Wound blaNDM NA R 57. PS353 OTHI Wound blaVIM RinAtI1 58. PS354 OTHI Wound blaNDM B intI1 59. PS367 FMCO Wound blaNDML I intI1,intI2 60. PS383 UCHI Wound bNlaND M intI1,intI2 61. PS384 UCHI Wound blaNDM NA 62. PS386 LTHO Wound DAblaNDM NA 63. PS392 LTHO Wound A blaNDM NA 64. PS393 LTHO W oIuBnd blaNDM intI1 65. PS394 LTHO FWound blaNDM intI1 66. PS395 OTH IO Wound blaNDM intI1 67. PS396 OYTHI Wound blaVIM, blaNDM NA 68. PS397 ITOTHI Wound blaNDM NA 69. PS398 S OTHI Wound blaNDM NA 70. PS400R OTHI Wound blaVIM, blaNDM Int1 71. PSE405 FMCO Wound blaVIM, blaNDM NA 72. IVPS409 FMCO Wound blaVIM intI1,intI2 7N3. PS414 LTHO Wound biopsy blaNDM intI1 U74. PS351(C) LTHO Urine intI1 75. PS423 (C) OTHI Wound NA 76. ATCC27853 NA Key: NA = Not amplified; ND = Not determined; (C) = Carbapenem susceptible strain 114 L 2 3 4 5 6 7 8 2.0 kb 1.5 kb Y 1.0 kb R 750 bp A R 500 bp LIB 491 bp 250 bp N AD A IB F Plate 4.8a. Agarose gel (1.5%)O showing PCR products for integrons before digestion with Rsa1 enzyme L = 1kYb pl us bp ladder; Lanes 2-5, 7,8 showed positive results for integrons (PS152, PS1I5T4, PS168, PS183, PS210, PS220); Lane 6 (PS204) is negative for integrons 1P85S E R IV UN 115 L 2 3 4 5 6 7 8 750 bp RY 500 bp RA491 bp 250 bp B N LI DA A Plate 4.8b. Agarose gel (1.5%) showingI BPCR products for integrons before digestion with Rsa1 enzyme L = 1 kb plus Fbp ladder; Lanes 2,3,6,7 (PS181, PS184, PS209, PS205) showed positive resu ltOs for integrons; Lanes 4,5,8 (PS185, PS202, PS210) showed negative results foYr integrons. IT S VE R UN I 116 L 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 3.0 kb 2.0 kb 1.5 kb 1.2 kb 1.0 kb 900 bp 800 bp 700 bp 600 bp Y 500 bp R 491 bp 400 bp 300 bp A 334 bp 200 bp 100 bp R 157 bp LI B L 17 18 19 20 21 22 23 24 25A 2N6 27 28 29 30 31 D 3.0 kb 2.0 kb A 1.5 kb 1.2 kb IB 1.0 kb 900 bp 800 bp F 700 bp 600 bp O 500 bp Y 491 bp 400 bp 334 bp 300 bp T 200 bp SI 157 bp 100 bp ER I V PlaNte 4.9a. Agarose gel (1.5%) showing PCR products for integrons L = 100 plus bp ladder; Lanes 2-4, U6-11, 13,14, 16,18-23, 27,28,30,31 shows different carbapenem-resistant P. aeruginosa positive for class integrons after digestion of PCR-products with RsaI enzyme (PS154, PS181, PS205, PS152, PS220, PS209, PS168, PS184, PS210, PS414, PS335, PS297, PS348, PS235, PS354, PS393, PS303, PS395, PS253, PS291, PS173, PS394); Lanes: 5, 12, 15,24-26,29 shows different carbapenem-resistant P. aeruginosa positive for class 1 and 2 integrons after digestion of PCR-products with RsaI enzyme (PS183, PS285, PS222, PS409, PS367, PS383, PS293, PS224) 117 L 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 3.0 kb 2.0 kb 1.5 kb 1.2 kb 1.0 kb 900 bp Y 800 bp 700 bp 600 bp R 500 bp A 491 bp 400 bp R 334 bp 300 bp 200 bp 100 bp LI B 157 bp DA N A IB Plate 4.9b. Agarose gel (1.5%O) shFowing PCR products for integrons. Lane 2,4-16 shows different carbapenem-r esistant P. aeruginosa positive for class 1 integrons after digestion of PCR-productsY with RsaI enzyme (L = 100 plus bp ladder; Lane 2-16 = PS093, PS096, PS146I,T PS100, PS097, PS022, PS244, PS147, PS294, PS325, PS294, PS296, PS250R, PSS172); Lane 3 (PS292) shows different carbapenem-resistant P. aeruginosaE positive for class 1 and 2 integrons after digestion of PCR-products with RsaI enVzyme NIU 118 L 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 4.0 kb 3.5 kb 3.0 kb 2.5 kb 2.0 kb 1.5 kb Y 1.0 kb R 750 bp A 500 bp LIB R 250 bp AN AD Plate 4.10. Agarose gel electrophoresis (B1.5%) of class 1 integron gene cassette L = 1kb ladder; Lane 2-15 (PS097, PS0F93 , IPS220, PS250, PS209, PS244, PS210, PS168, PS291, PS160, PS367, PS222, OPS350, PS303) shows different P. aeruginosa class 1 integrase gene cassette; N Y= N egative control. IT RS VE I UN 119 Table 4.15. Prevalence of Type III effector toxins in carbapenem-resistant P. aeruginosa according to hospital Hospital (no) T3SS toxins exoT exoY exoU exoS exoU and exoS UCHI (23) 23 (100) 23 (100) 14 (60.9) 9 (39.1) - FMCA (2) 2 (100) 2 (100) - 2 (100) - Y OTHS (5) 5 (100) 5 (100) 1 (20) 4 (80) - AR OTH1 (13) 13 (100) 23 (100) 8 (61.5) 4 (30.8) IB1 (7 R.7) LTHO (8) 8 (100) 8 (100) 6 (75.0) 2 (25. 0L) - FMCO (14) 14 (100) 14 (100) 4 (28.6) A10N (71.4) - FMCI (8) 8 (100) 8 (100) 2 (2A5.0D) 5 (62.5) 1 (12.5) Key: IB - = absence of amplified gene OF UCHI = University College H ospital Ibadan FMCA = Federal Medical YCentre Abeokuta; OTHS = Olabisi OnabIaTnjo University Teaching Hospital Sagamu; OTHI = Obafemi ASwolowo University Teaching Hospital Ile-Ife; LTHO = LadoRke Akintola University Teaching Hospital Osogbo; FMCO V= FEederal Medical Centre Owo; FMCII = Federal Medical Centre Ido-Ekiti UN 120 L 2 3 4 5 6 7 8 9 10 11 12 13 14 15 N Table 4.16. Prevalence of Type III effector toxins in carbapenem-resistant P. aeruginosa according to clinical source T3SS toxins Clinical source (no (%)) Wound Tracheal Ear (5) Urine (8) Wound (56) aspirate (2) biopsy (2) exoT 56 (100) 2 (100) 5 (100) 8 (100) 2 (100) exoY 56 (100) 2 (100) 5 (100) 8 (100) 2 (10Y0) exoU 24 (42.9) 1 (50.0) 2 (40.0) 6 (75.0) A2R (100) exoS 30 (53.6) 1 (50.0) 3 (60.0) 2 (B25.0R) - I exoU and exoS 2 (3.6) - - L- - Key: N - = absence of amplified gene DA IB A OF SI TY ER NI V U 121 L 2 3 4 5 6 7 8 3.0 kb 2.0 kb 1.5 kb 1.2 kb 1.0 kb 900 bp Y 800 bp 700 bp R 600 bp A 500 bp R 400 bp B 300 bp LI 289 bp 200 bp N bp A 152 bp 100 bp D 134 bp 118 bp A IB OF Plate 4.11a. Genotyping Yof e xoU, exoS, exoT and exoY in carbapenem-resistant P. aeruginosa (CRPA) isTolates with multiplex PCR. L: 100 bp plus ladder. Lane 2 (PS181) is exoY, eSxoTI and exoU positive strain; Lanes 3-5, 7,8 is exoY, exoT and exoS positive strainsR (PS205, PS170, PS093, PS172, PS022). Lane 6 (PS293) is exoY, exoT, exoU and eExoS positive strain (exoY-289bp; exoT-152bp; exoU-134bp; exoS-118bp) IV UN 122 L 2 3 4 5 6 7 8 3.0 kb 2.0 kb 1.5 kb 1.2 kb 1.0 kb 900 bp Y 800 bp 700 bp R 600 bp A 500 bp R 400 bp B 300 bp LI 289 bp 200 bp bp N 152 bp 100 bp A 134 bp 118 bp AD F I B Plate 4.11b. Genotyping of e xOoU, exoS, exoT and exoY in carbapenem resistant P. aeruginosa (CRPA) isolatYes by multiplex PCR. Lanes 2, 4-8 is exoY, exoT and exoS positive strains (PS00I7T, PS088, PS096, PS099, PS100, PS150); Lane 3 (PS335) exoY, exoT and exoU posSitive strain (exoY-289bp; exoT-152bp; exoU-134bp; exoS-118bp) R E V UN I 123 L 2 3 4 5 6 7 3.0 kb 2.0 kb 1.5 kb 1.2 kb 1.0 kb 900 bp 800 bp Y 700 bp 600 bp R 500 bp A 400 bp LIB R 289 bp 300 bp bp 200 bp 152 bp 134 bp 100 bp AN 118 bp BA D I OF Plate 4.11c. Genotyping Yof e xoU, exoS, exoT and exoY in carbapenem resistant P. aeruginosa (CRPA) isoTlates by multiplex PCR. Lanes 3 and 4 are exoY, exoT and exoS positive strains (PSS29I1, PS292); Lanes 2,5,6 are exoY, exoT and exoU positive strains (PS224, PS097R, PS348); Lane 8 (PS354) is exoY, exoT, exoU and exoS positive strain (exoY-289bEp; exoT-152bp; exoU-134bp; exoS-118bp) V UN I 124 L 2 3 4 5 6 7 8 3.0 kb 2.0 kb 1.5 kb 1.2 kb 1.0 kb 900 bp 800 bp Y 700 bp R 600 bp 500 bp A 400 bp BR 300 bp LI 289 bp bp 200 bp N 152 bp A 134 bp 118 bp 100 bp BA D OF I Plate 4.11d. GenotypIiTng Yof exoU, exoS, exoT and exoY in carbapenem resistant P. aeruginosa (CRPAS) isolates by multiplex PCR. Lanes 2,3,6,7 is exoY, exoT and exoS positive strainRs (PS173, PS202, PS205, PS088); Lanes 4,5,8 (PS183, PS209, PS210) exoY, exoTE and exoU positive strain (exoY-289bp; exoT-152bp; exoU-134bp; exoS- 118bp) V U N I 125 MBL Integron ExoU 100 90 80 70 60 50 AR Y 40 R 30 LI B 20 N 10 DA 0 UCHI FMCA OTHS IBOT AHI LTHO FMCO FMCI Hospital F Figure 4.4. Distribution of MOBL(s), integron and exoU in carbapenem-resistant P. aeruginosa (CRPA) accordYing to hospital Key: IT UCHI = UniveRrsityS College Hospital, Ibadan FMCA = FEederal Medical Centre, Abeokuta OTHS =V Olabisi Onabanjo University Teaching Hospital, Sagamu LTNHOI = Ladoke Akintola University Teaching Hospital, Osogbo UOTHI = Obafemi Awolowo University Teaching Hospital Complex, Ile-Ife FMCO = Federal Medical Centre, Owo FMCI = Federal Medical Centre, Ido-Ekiti 126 Frequency of isolate (%) 100 90 80 70 60 RY 50 RA MBL IB Integron40 L exoU 30 AN 20 AD 10 F I B 0 Wound Tracheal Ear Urine Wound Yasp ir Oate biopsy T Clinical source SI Figure 4.5. DRistribution of MBL(s), integron and exoU in carbapenem-resistant P. aeruginosaE (CRPA) according to clinical source NI V U 127 Frequency of the isolates (%) 4.15 Quantification of efflux pumps expression in carbapenem-resistant P. aeruginosa Plate 4.12 shows the RNA isolated from carbapenem-resistant Pseudomonas aeruginosa on an agarose gel. Upregulation of the four efflux systems (MexAB-OprM, MexCD-OprJ, MexEF-OprN and MexXY-OprM) was determined in 48 randomly selected carbapenem-resistant P. aeruginosa and one susceptible strain by determining the relative transcription levels of mexA, mexB, mexC, mexD, mexE, mexF, mexX and mexY genes by quantitative real-time PCR. The mRNA was said to be overexpressed if the equivalent mRNA level was at least 2-fold (mexA, mexB), 4-fold (mexX aRnd YmexY) or 100-fold (mexC, mexD, mexE and mexF) higher than that for P. aerugAinosa ATCC 27853 (Hocquet et al., 2006). The analysis of gene expression showeRd that 33 (68.8%) of the carbapenem-resistant P. aeruginosa clinical isolates overeIxBpressed one or more efflux pump genes while 15 (31.3%) isolates showed no ov eLrexpression of any of the efflux gene. One isolate (PS293) overexpressed all the foNur pumps. MexXY-OprM was the most overexpressed pump occurring in 28 (58A.3%) of the isolates followed by MexAB-OprM which was overexpressed inA 22D (45.8%) strains, MexCD-OprJ in 5 (10.4%) strains and MexEF-OprN was fouBnd in 3 (6.25%) of the isolates (Table 4.17). I Among the carbapenem-resistant P. ae ruginosa strains, mexX was the most expressed gene as observed in 26 (54.2%) stFrains with fold increase from 4.98 to 996.3 while mexY was overexpressed Yin 1 8 O (37.5%) strains with fold increase from 5.3 to 643.8 compared to P. aerugIinTosa ATCC 27853. In 19 (39.6%) of the carbapenem-resistant strains, fold increSase in mexA from 2.50 to 310.4 was observed while 18 (37.5%) showed increaRsed mexB from 3.1 to 214.5 fold at the transcriptional mRNA levels compared Eto P. aeruginosa ATCC 27853. Increase in mexC was demonstrated in 4 (8.3%I) Vstrains from 133.4 to 1661.6 fold while 5 (10.4%) isolates showed an increase inN mexD from 211.1 to 2550.9 fold levels compared to P. aeruginosa ATCC 27853. UThe mexE gene was overexpressed in 2 (4.2%) strains with fold increase of 108.3 and 279.9 while only one strain (2.1%) overexpressed mexF with a fold increase of 101.3 compared to P. aeruginosa ATCC 27853 (Table 4.19). In 16 (33.3%) isolates, both mexX and mexY genes were overexpressed while mexX and mexY were overexpressed seperately in 10 (20.8%) and 2 (4.2%) isolates, respectively. Both mexA and mexB genes were overexpressed in 14 (29.2%) of isolates while mexA and mexB genes were overexpressed individually in 5 (10.4%) and 4 128 (8.3%) isolates, respectively. Both mexC and mexD genes was overexpressed in 4 (8.3%) of the strains. None of the isolates overexpressed both mexE and mexF genes (Figure 4.6). 4.16 Quantification of ampC overexpression and diminished expression of oprD porin Table 4.18 shows the relative fold expression of ampC and oprD porin. The ampC gene was overexpressed in 13 (27.1%) strains with fold increase of between 50.2 a nd 5086.7 compared to P. aeruginosa ATCC 27853 (Table 4.19). UnderexpYressed outermembrane porin (oprD) was observed in 37 (77.1%) carbapenem-reRsistant P. aeruginosa strains. Among the ten carbapenem-resistant P. aeruginosAa that lacked MBL genes, the only mode of resistance in four isolates wasI aBttri Rbutable to efflux pump overexpression (PS007, PS088, PS099, PS349) while defective oprD alone was responsible for carbapenem resistance in three isolates (LPS150, PS166, PS185). Carbapenem resistance in PS170 and PS244 wasA duNe to efflux pump and ampC overexpression while in strain PS414, caDrbapenem resistance results from overexpression of efflux pump and underexpAression of oprD. In this study, only ten isolates had both reduced oprD mRNA tIraBnscription and ampC overexpression. 4.17 MIC of antibiotics againstF carbapenem-resistant P. aeruginosa isolates in relation to MBL and efOflux pump genes Table 4.20 compares the MYIC of selected antibiotics against CRPA isolates relative to the presence or absenIcTe of MBL efflux pump overexpression. It was observed that isolates that lack MSBL genes but overexpressed efflux pump gene had reduced MIC against imEipeRnem, meropenem, ceftazidime, cefepime and ciprofloxacin when comparVed with isolates having combination of MBL genes and efflux pump. I UN 129 L 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 3 4 5 6 7 8 9 10 11 12 13 `14 15 16 23S 16S Y 17 18 19 20 21 22 23 24 25 26 27 28 29 30 R 31 32 BR A LI23S 16S N AD A B F I Plate 4.12. Agarose gel (1. 0%O) showing total RNA extracted from carbapenem-resistant P. aeruginosa Ywith PureLinkTM Micro-to-Midi Total RNA Extraction System (Invitrogen). TLane 1-32 carbapenem-resistant P. aeruginosa total RNA (PS429, PS099, PSS34I9, PS397, PS170, PS250, PS007, PS230, PS246, PS292, PS022, PS405, PS224R, PS173, PS202, PS222, PS100, PS398, PS394, PS166, PS384, PS210, PS296, PS1E85, PS291, PS414, PS183, PS238, PS243, PS400, PS152, PS088 V UN I 130 Table 4.17. Relative fold expression of efflux pump genes in carbapenem-resistant P. aeruginosa clinical isolates compared to P. aeruginosa ATCC 27853 S/N Isolate Efflux pump genes mexA mexB mexC mexD mexX mexY mexE mexF 1. PS007 2.9 1.8 2.5 0.6 5.3 1.3 1.4 1.0 2. PS022 0.5 0.3 1.8 0.7 4.4 1.7 0.5 0.4 3. PS088 1.7 0.8 1.5 0.3 21.3 1.3 3.1 0.9 4. PS093 4.3 2.5 10.9 7.5 16.2 93.0 17.9 Y4. 4 5. PS097 0.8 0.4 0.4 0.2 1.3 0.1 0.2R 0.1 6. PS099 2.9 3.7 29.0 8.9 4.9 1.2 A1.0 1.1 7. PS100 2.8 25.3 21.4 0.7 17.7 1.6 R 4.5 3.2 8. PS150 0.2 0.5 0.7 0.2 1.1 0I.3B 0.4 0.2 9. PS152 0.6 0.4 0.03 0.3 9.2 L0.6 0.3 0.01 10. PS166 0.9 1.2 0.4 0.5 0.N7 3.4 0.6 0.2 11. PS170 6.8 2.0 59.8 55.0 A166.3 256.9 108.3 41.6 12. PS173 0.5 0.6 0.4 0.2 D 1.0 2.9 0.4 0.1 13. PS181 0.5 0.7 0.8 IB0 A.3 34.9 49.6 0.3 0.01 14. PS182 7.9 67.6 42.1 39.2 176.9 171.7 42.0 101.3 15. PS183 1.6 0.9 1F.1 1.1 78.9 10.4 0.5 0.3 16. PS185 1.7 1.1 O1.1 0.5 2.3 0.2 0.6 1.6 17. PS202 31.5 T6Y.6 425.5 2550.9 740.7 580.5 23.1 1.3 18. PS204 3.8 I 2.0 9.2 14.3 37.2 43.3 12.4 3.0 19. PS209 0.S5 13.7 0.4 0.3 51.5 65.5 2.5 2.3 20. PS2E10 R3.3 5.3 0.3 0.7 2.6 1.5 0.4 0.6 21I. V PS219 310.4 214.5 1661.6 112.0 245.7 332.5 47.8 101.6 N22. PS220 23.59 23.3 48.4 244.2 643.2 3.4 4.2 3.5 U 23. PS222 1.94 1.1 1.2 2.5 0.9 3.5 0.8 0.5 24. PS224 0.85 1.6 1.0 0.0064 2.0 11.4 0.3 0.2 25. PS230 0.67 0.5 0.7 0.5 1.7 0.7 0.2 0.4 131 Table 4.17. Relative fold expression of efflux pump genes in carbapenem-resistant P. aeruginosa clinical isolates compared to P. aeruginosa ATCC 27853 (cont‟d) S/N Isolate Efflux pump genes mexA mexB mexC mexD mexX mexY mexE mexF 26. PS238 0.1 3.3 0.4 0.3 10.6 24.2 0.1 0.01 27. PS243 0.3 0.7 2.6 0.4 34.7 46.1 0.7 0.6 28. PS244 141.6 54.2 323.4 276.3 276.4 113.4 11.8 3 2.4 29. PS246 0.5 0.8 0.5 0.1 0.9 1.6 0.3 0.4 30. PS250 0.9 1.3 0.8 0.3 2.4 1.2 0.R2 Y0.1 31. PS291 1.7 0.1 0.1 0.1 3.4 0.04 A0.1 0.05 32. PS292 3.5 5.3 0.6 0.2 7.0 3.3 R 0.3 0.2 33. PS293 75.8 20.2 133.4 211.1 996.3 I2B430.8 279.9 41.7 34. PS296 1.0 0.2 1.6 0.02 3.3 L 0.2 1.1 1.0 35. PS303 0.1 0.1 0.2 0.1 1N7.1 0.4 0.1 0.01 36. PS349 0.2 2.3 5.5 1.D8 A25.6 0.5 0.03 0.7 37. PS367 2.5 0.2 1.8 0.2 91.3 2.3 0.5 0.1 38. PS384 0.4 0.6 0.I2 B A0.1 0.6 0.6 0.5 0.1 39. PS392 0.6 0.4 1 1.8 8.1 40.4 54.4 8.1 5.3 40. PS393 0.3 0.2 F0.4 0.1 12.1 2.9 0.1 0.01 41. PS394 1.8 2.O3 2.3 0.8 2.9 1.0 0.02 0.5 42. PS395 8.T1 Y0.01 21.5 26.6 11.1 8.5 1.0 0.9 43. PS397 I3.1 1.4 1.5 0.5 1.6 0.7 0.1 0.2 44. PS398R S4.5 2.0 0.7 0.5 1.3 2.3 0.4 0.7 45. PSE400 1.5 0.7 5.9 3.2 24.1 5.3 0.9 1.4 46I. V PS405 2.2 3.1 0.2 0.1 2.7 1.2 0.04 0.3 N47. PS409 0.1 0.8 5.2 0.2 1.2 1.4 0.2 0.2 U 48. PS414 0.4 0.7 0.3 0.3 3.3 31.2 0.01 0.03 49. PS351 0.8 0.4 2.4 0.4 0.2 0.4 0.03 0.3 (nonMDR) 132 Table 4.18. Relative fold expression of ampC and outermembrane porin (oprD) in carbapenem-resistant P. aeruginosa clinical isolates compared to P. aeruginosa ATCC 27853 S/N Isolate oprD ampC 1. PS007 5.4 0.8 2. PS022 0.0002 1.1 3. PS088 4.1 1.9 4. PS093 0.02 405.2 5. PS097 0.001 1.3 RY 6. PS099 2.0 2.2 A 7. PS100 0.5 1.9 R 8. PS150 0.002 2.0 B 9. PS152 0.001 1.6 N LI 10. PS166 0.01 13.1 11. PS170 1.6 510.6 12. PS173 0.2 8.3 DA 13. PS181 0.01 0.7I A 14. PS182 1.8 28 23B.7 15. PS183 0.3 F113.8 16. PS185 0.003 O 3.7 17. PS202 0.Y2 5086.7 18. PS204 IT1.1 217.4 19. PS209S 0.0 0.4 20. EPSR210 0.004 0.3 21I. V PS219 0.2 464.1 N22. PS220 1.1 7.1 U 23. PS222 0.08 0.2 24. PS224 0.03 1.6 25. PS230 0.001 0.6 133 Table 4.18. Relative fold expression of ampC and outermembrane porin (oprD) in carbapenem-resistant P. aeruginosa clinical isolates compared to P. aeruginosa ATCC 27853 (cont‟d) S/N Isolate oprD ampC 26. PS238 0.001 0.2 27. PS243 0.1 118.5 28. PS244 1.5 50.3 29. PS246 0.4 0.8 30. PS250 0.003 137.2 RY 31. PS291 0.02 0.3 A 32. PS292 0.3 0.1 BR33. PS293 0.3 1767.5 34. PS296 0.3 2.4 LI 35. PS303 0.001 0.1 N 36. PS349 1.5 0.6 A 37. PS367 0.1 A8D.0 38. PS384 0.5 IB 1.0 39. PS392 0.0 02 460.3 40. PS393 F0.002 6.0 41. PS394 O0.01 0.4 42. PS395 Y 0.1 9.4 43. PS397I T 1.3 2.0 44. PSS398 1.1 0.9 45. ERPS400 0.02 50.2 46I. V PS405 0.2 0.7 N47. PS409 0.4 0.2 U 48. PS414 0.002 0.3 49. PS351(nonMDR) 0.7 8.4 134 Table 4.19. MDR efflux gene overexpression, ampC overexpression and oprD loss in carbapenem-resistant P. aeruginosa Efflux gene No (%) of isolates with overexpression Range of fold increase mexA 19 (39.6) 2.5 - 310.4 mexB 18 (37.5) 3.1 - 214.5 mexC 4 (8.3) 133.4 - 1661.6 Y mexD 5 (10.4) 211.1 – 2550.A9 R mexE 2 (4.2) 108.3I B- 27 R9.9 mexF 1 (2.1) 1 0L1.3 mexX 26 (54.2) AN5.0 - 996.3 mexY 18 (37.5) AD 5.3 - 2430.8 oprD 37 (77.1) IB 0.001 - 0.5 ampC 13 (27.1) OF 50.2 – 5086.7 ITY RS IV E UN 135 60 54.2 50 39.6 40 37.5 37.5 AR Y 30 R LI B 20 AN 10.4 D 10 8.3 IB A 4.2 F 2.1 0 MexA MexB OY MexC MexD MexE MexF MexX MexYIT Efflux genes Figure 4.6. CompSarison of proportion of two efflux genes overexpressed in each of the four effluxE pumRp system in carbapenem-resistant P. aeruginosa IV U N 136 Percentage of isolates with overexpression RY Table 4.20. MIC of antibiotics against carbapenem-resistant P. aeruginosa isolates in relation to MDR efflux gene overeAxpressed and MBL S/N Isolate Sample Ipm Mem Fep Caz Cip Effux gene Efflux pump R MBL 1. PS007 UCHI 16 16 >128 >256 128 A MexAB-OprM - 2. PS088 FMCA 32 16 4 128 16 X MexXY-OprM IB - 3. PS093 FMCA 2 8 16 128 >128 A, B, X, Y MexAB-OprM, MLexXY-OprM + 4. PS099 FMCO 8 8 16 256 4 A, B, X MexAB-ONprM, MexXY-OprM - 5. PS100 FMCO 4 8 16 256 8 A, B, X MexAB-OprM, MexXY-OprM + 6. PS152 UCHI 32 >64 >128 >256 >128 X MexXAY-OprM + 7. PS170 UCHI 8 16 8 >256 32 A, X, Y, E DMexAB-OprM, MexXY-OprM, MexEF-OprN - 8. PS182 UCHI 4 8 4 4 16 A, B, X, Y, FA MexAB-OprM, MexXY-OprM, MexEF-OprN - 9. PS183 UCHI 16 64 >128 >256 >128 XY B MexXY-OprM + 10. PS202 OTHS 8 16 16 >256 32 A, B,I C, D, X, Y MexAB-OprM, MexCD-OprJ, + 11. PS204 UCHI >64 >64 >128 >256 >128 A, B, X, Y MexAB-OprM, MexXY-OprM + 12. PS209 OTHI 8 >64 32 128 4 FB, X, Y MexAB-OprM, MexXY-OprM + 13. PS210 OTHI 16 >64 32 256 O8 A, B MexAB-OprM + 14. PS219 FMCO >64 64 64 >256 >128 A, B, C, D, X, Y MexAB-OprM, MexCD-OprJ, MexXY-OprM + 15. PS220 FMCO 16 64 >128 3Y2 >128 A, B, D, X MexAB-OprM, MexCD-OprJ, MexXY-OprM + 16. PS224 FMCO 32 32 64I T64 32 Y MexXY-OprM + 17. PS238 LTHO >64 >64 >128 >256 >128 B, X, Y MexAB-OprM, MexXY-OprM + Key: Ipm = Imipenem, Mem = MeroRpenemS, Fep = Cefepime, Caz = Ceftazidime, Cro = Ceftriaxone, Cip = Ciprofloxacin, + = MBL present, - = MBL absent, A = mexA, B = meExB, C = mexC, D = mexD, E = mexE, F = mexF, X = mexX, Y = mexY IV 137 UN Y R Table 4.20. MIC of selected antibiotics against carbapenem-resistant P. aeruginosa isolates in relation to MDR effluxA gene overexpressed and MBL (cont‟d) R S/N Isolate Sample Ipm Mem Fep Caz Cip Efflux gene Efflux pump MBL 18. PS243 LTHO 64 >64 >128 >256 >128 X, Y MexXY-OprM IB + 19. PS244 UCHI >64 >64 >128 >256 >128 A, B, C, D, X, Y MexAB-OprM, MexCD-OprJ,L MexXY-OprM - 20. PS292 FMCI 2 8 >128 8 64 A, B, X MexAB-OprM, MexXY-Op rM + 21. PS293 FMCI >64 >64 >128 >256 >128 A, B, C, D, X, Y, E MexAB-OprM, MexCND-OprJ, MexXY-OprM, MexEF-OprN + 22. PS303 FMCI >64 >64 >128 >256 >128 X MexXY-OprM + 23. PS349 OTHI 4 >64 >128 >256 >128 B, X MexAB-DOprM,A MexXY-OprM - 24. PS367 FMCO >64 >64 >128 >256 >128 A, X MexAB-OprM, MexXY-OprM + 25. PS392 LTHO >64 >64 128 256 >128 X, Y MeAxXY-OprM + 26. PS393 LTHO >64 >64 >128 >256 >128 X MexXY-OprM + 27. PS394 LTHO >64 >64 16 >256 >128 B BMexAB-OprM + 28. PS395 OTHI >64 >64 >128 >256 >128 A, X, Y F I MexAB-OprM, MexXY-OprM + 29. PS397 OTHI >64 >64 >128 >256 >128 A MexAB-OprM + 30. PS398 OTHI >64 >64 >128 >256 >128 A, B MexAB-OprM + 31. PS400 OTHI >64 >64 >128 >256 >128 XO, Y MexXY-OprM + 32. PS405 FMCO >64 >64 >128 >256 >128 A, B MexAB-OprM + 33. PS414 LTHO >64 >64 >128 >256 Y>128 Y MexXY-OprM + Key: Ipm = Imipenem, Mem = Meropenem, Fep =S CefepIim Te, Caz = Ceftazidime, Cro = Ceftriaxone, Cip = Ciprofloxacin, + = MBL present, - = MBL absent, A = mexA, B = mexB, C = mexC, D = mexD, E = mexE, F = mexF, X = mexX, Y = mexY R VE 138 I UN 90 80 70 60 50 AR Y 40 BR30 LI 20 N 10 A 0 AD MBL Efflux pu mIpB oprD ampC OReFsistance mechanisms Figure 4.7. Occurrence ofY various carbapenem resistance mechanisms in carbapenem- resistant P. aeruginosIa T S VE R NIU 139 Percentage prevalence Table 4.21. Singles and combinations of resistance mechanisms in CRPA Resistance mechanisms Number of isolates MBL, Efflux pump, oprD, ampC 8 MBL, Efflux pump, oprD 13 MBL, Efflux pump, ampC 1 MBL, oprD 8 RY MBL, Efflux pump 3 RA Efflux pump, ampC 2 LIB Efflux pump, oprD 1 N MBL, oprD, ampC 1 DA MBL B4A Efflux pump F I 4 oprD O 3 ITY ER S NI V U 140 4.18 Molecular typing of carbapenem-resistant Pseudomonas aeruginosa The three rep-PCR methods used for typing revealed a lot of genetic diversity among the isolates. Amplification of genomic DNA of seventy-three carbapenem-resistant and two carbapenem-sensitive strains of P. aeruginosa with REP1R and REP2 primers produced 48 fingerprinting patterns with 27 common types and 21 single types from the dendrogram. ERIC-PCR primers (ERIC1R and ERIC2) gave 50 fingerprinting patterns containing 25 common types and 25 unique patterns were found. Fifty-two (52) fingerprinting profiles were obtained with BOXAIR primer, 22 of which w ere common types while the remaining 30 were unique types (Table 4.22). RREPY-PCR produced 2 to 14 bands with molecular weight ranging from 200bp to 40A00bp. ERIC- PCR gave 4 to 15 bands with band sizes ranging from 100bp to 350R0bp while BOX- PCR produced 7 to 15 bands with molecular weight of betweIenB 250bp and 3500bp (Plates 4.13 – 4.15). L Cluster analysis of ERIC-PCR type at 0.15 distances reNvealed twenty two (22) types containing sixteen (16) clusters and six (6) singleD isoAlates with discriminatory index of 0.934. The prevalent cluster was type C consAisting of eleven (11) isolates followed by type B with ten (10) isolates. BOX -PICBR type produced twenty three (23) types containing twelve (12) clusters with 11 single isolates with discriminatory index of 0.896 at 0.15 distances. Type B waFs the most common cluster consisting of eighteen isolates followed by typYes E O and I with thirteen (13) and eight (8) isolates, respectively. REP-PCIRT type gave twelve (12) types consisting of eight (8) clusters with four (4) singSle isolates having discriminatory index of 0.854. Type A was the prevalent consRisting of eighteen (18) isolates followed by types F and B with fifteen (15) and foEurteen (14) isolates, respectively (Figures 4.8 - 4.10). REP-PICVN R type could not differentiate between pairs of seven set of isolates (PS303 = UPS395; PS235 = PS184; PS296 = PS351; PS205 = PS088; PS100 = PS022; PS166 = PS350; PS168 = PS209). PS303 and PS395 strains obtained from different hospitals (FMCI and OTHI, respectively) could not be distinguished with REP-PCR. The two isolates harboured exoU and blaNDM and also belong to the same cluster C and E on ERIC- and BOX- PCR. PS235 and PS184 were obtained from LTHO and UCHI, respectively, however, these strains could not be differentiated with REP-PCR but belong to the same cluster F on ERIC-PCR and cluster I on BOX- PCR dendrogram. Both isolates harboured blaNDM and have the exoU gene. REP-profile was not able to 141 distinguish between PS296 and PS351 isolates that PS296 was resistant to all the three carbapenem (imipenem, meropenem and doripenem) as well as harboured blaVIM and blaNDM while PS351 was susceptible to the three carbapenems and do not possess neither blaVIM nor blaNDM although they both have exoS gene. PS296 and PS351 belong to different clusters A and I on ERIC-profile as well as clusters A and L on BOX-profile dendrograms, respectively. PS205 and PS088 obtained from different hospitals (UCHI and FMCA respectively) are not distinguishable on REP-PCR dendrogram but belong to different clusters C and D on ERIC-profile as well as clusters E and F on BOX-profile, respectively. PS205 contain bla whicRh wYVIM as not present in PS088 but both have exoS toxin. PS100 and PS022 was anothAer pair of two different strains obtained from different hospitals (FMCI and UCHI, Rrespectively) that could not be distinguished with BOX-profile. The two isolates bIotBh belong to cluster B by ERIC-profile but cluster differs (clusters B and G) by LBOX-profile. They both contain blaVIM contain exoS gene. PS166 and PS350 weNre also obtained from different hospitals (UCHI and OTHI, respectively) but were inAdistinguishable by REP-profiling. PS350 belongs to cluster P while PS166 doAes nDot cluster with any isolate on ERIC-profile while on BOX-profile, PS166 Band PS350 belong to clusters G and H, respectively. PS350 contain blaVIM w hIich was not present in PS166 but they both contain exoU gene. PS168 and PS2F09 was another pair of epidemiologically unrelated isolates obtained from differe ntO hospitals (UCHI and OTHI, respectively) but was not distinguishable by REP-prYofiling. PS168 and PS350 belong to different clusters B and P by ERIC-profile anIdT clusters J and G by BOX-profile, respectively. Both contain blaVIM and exoU toSxin (Table 4.23). BOX-profiElingR was unable to differentiate between PS297 and PS285 isolates from the same IhVospital (FMCI). Both isolates have different type 3 effector toxin and also coNntain different MBL gene (blaVIM:exoS and blaNDM:exoU, respectively). Another set Uof isolates that could not be differentiated with BOX-PCR was PS202, PS383 and PS250. PS202 isolate was obtained from OTHS and was carrying blaVIM and also contain exoS gene; for PS383 and PS250 obtained from UCHI, PS383 harboured blaNDM only while PS250 harboured both blaVIM and blaNDM even though both isolates have exoU type III effector toxin. Likewise, BOX-profile could not distinguish between PS303 and PS354 obtained from FMCI and OTHI, respectively. These isolates belong to the same cluster (B) on ERIC- and REP- PCR dendrograms, both 142 harboured blaNDM and contain exoU effector toxin but PS354 also possessed exoS alongside exoU toxin. Similarly, PS238 and PS152 obtained from LTHO and UCHI, respectively could not be differentiated with BOX-PCR and also belong to the same clusters C and I on REP- and ERIC- PCR. Although both PS238 and PS152 contain exoU gene, PS152 harboured blaNDM only while PS238 harboured both blaVIM and blaNDM. Only BOX-PCR was able to group the two carbapenem susceptible isolates into the same cluster L (Table 4.24). Table 4.25 gave the genotypic characteristic of four clones confirmed by two orY more rep-PCR methods. Two strains in clone 01 both possessed blaVIM, clasAs 1R integrons with approximately 3.5 kb cassette, exoS type III effector gene, ampRC overexpression absent, underexpressed oprD, while PS100 overexpressed mexAB, mexB and mexX, PS022 do not overexpress any of the efflux pump genes. N LI DA A F I B O SI TY VE R I UN 143 RY L 2 3 4 5 6 7 8 9 10 11 12 13 14 15 L 17 18 19 20 21 22 23 24 25 R26 27A 28 29 30 LIB 3.5 kb 3.0 kb 2.5 kb 2.0 kb DA N 1.5 kb 1.0 kb 750 bp BA 500 bp I 250 bp OF Plate 4.13. Representative image of fingerprYinting patterns of carbapenem-resistant P. aeruginosa strains by gel electrophoresis after REP-PCR; Lanes L = 1 kb plus ladder, ILaTnes 2-15, 17-30 = PS383, PS202, PS297, PS335, PS147, PS394, PS367, PS250, PS293, PS405, PS007, PS400, PS285, PS398, PSS235, PS184, PS352, PS244, PS170, PS181, PS230, PS397, PS253, PS409, 348, PS243, PS220 and PS246. ER IV 144 UN R-TYPE SPECIMEN HOSP. MBL INTEG. T3SS S R1 WOUND FMCO VIM INT1 S R1 WOUND OTHS NDM - S R2 WOUND OTHI VIM INT1 S R3 WOUND FMCI NDM INT1,2 U R3 WOUND OTHI NDM - S R4 WOUND UCHI - INT1 S R5 WOUND UCHI VIM INT1 S R5 WOUND LTHO VIM - U R6 WOUND FMCO VIM INT1 S A R7 WOUND LTHO NDM INT1 S R7 WOUND FMCI VIM INT1 S R8 WOUND UCHI - - S R8 URINE OTHI - - S R9 WOUND FMCO N DM - S R9 EAR FMCA V IM INT1 S R10 WOUND OTHS VIM - U R10 WOUND LTHO NDM - S R11 WOUND FMCO VIM INT1,2 S R12 WOUND OTHI NDM - YU R13 WOUND UCHI VIM, NDM RINT1 UR13 WOUND OTHI NDM INT1 S R14 WOUND FMCI NDM INT1 U R14 WOUND OTHI NDMA INT1 U R15 WOUND LTHOR NDM INT1 U R16 WOUND OTHI VIM, NDM INT1 U R16 WOUND BFMCI NDM INT1,2 U B R17 WOUNDI LTHO NDM INT1 U R17 WOUND B UCHI NDM INT1 U R18 WLOUND FMCO NDM INT1,2 U R19 URINE UCHI VIM INT1 U R20 URINE OTHI NDM INT1 U R21 EAR FMCO NDM INT1 U NR21 WOUND FMCI VIM, NDM INT1,2 UR22 WOUND UCHI - - S A R22 WOUND OTHI - - U R23 WOUND FMCO VIM, NDM - S R23 EAR LTHO VIM, NDM INT1 D S C R24 URINE UCHI NDM INT1 U R24 T. ASP UCHI NDM INT1 U A R25 WOUND LTHO NDM - U R26 WOUND UCHI - - U B R26 WOUND FMCO VIM - U I R27 WOUND FMCO VIM, NDM INT1,2 S R28 WOUND OTHI NDM - S D R28 WOUND OTHS NDM INT1 S F R29 WOUND UCHI VIM - U R30 EAR UCHI - INT1 U R30 WOUND UCHI VIM - S E O R31 WOUND FMCI VIM, NDM INT1 S R31 WOUND LTHO - INT1 S R32 WOUND UCHI VIM INT1 S Y R32 WOUND FMCA - - S R33 WOUND FMCO VIM INT1 S R34 WOUND FMCI VIM INT1 T U I R34 WOUND OTHI VIM, NDM - S R35 WOUND FMCO VIM INT1 U R35 URINE UCHI NDM INT1 F S S R36 WOUND B LTHO NDM INT1 U R37 WOUND FMCO VIM INT1 U R R37 T. ASP UCHI VIM INT1 S R38 URINE OTHI VIM INT1 S R38 URINE UCHI - INT1 E U R39 URINE UCHI - - U R40 WOUND UCHI NDM INT1,2 S V R40 WOUND UCHI NDM - U I R41 WOUND FMCO - - G S R42 WOUND UCHI VIM INT1 S N R42 WOUND OTHI VIM INT1 U R43 WOUND OTHI VIM INT1 U R44 WOUND FMCO VIM, NDM INT1,2 U H U R44 EAR OTHS VIM INT1 S R45 WOUND OTHS VIM INT1 S R46 WOUND FMCI VIM INT1 R 4F7 igWuOUrNeD 4.U8CH.I S D NDeMn droIgNTr1a m S R48 WOUND UCHI NDM INT1,2 U showing cluster analysis of carbapenem-resistant P. aerug ino sa stra ins by REP-PCUR using phylotree software and Unweighted Pair Group Me thod with Arithmetic Mean (UPGMA). Key: U= exoU; S= exoS 145 RY A L 2 3 4 5 6 7 8 9 10 11 12 13 14 L 16 17 18 19 20 21 22 23 24 R25 26 27 28 29 B I 3.5kb L 3.0 kb 2.5 kb 2.0 kb N 1.5 kb A 1.0 kb 750 bp D A 500 bp B 250 bp I O F Y Plate 4.14. Representative image of fingIeTrprinting patterns of carbapenem-resistant P. aeruginosa strains by gel electrophoresis following ERIC-PCR; Lanes L = 1 kb plus laSdder, Lanes 2-14, 16-29 = PS296, PS325, PS97, PS100, PS185, PS182, PS154, PS168, PS173, PS204, PS210, PS224, PS414, PS383R, PS202, PS297, PS335, PS147, PS394, PS367, PS250, PS293, PS405, PS007, PS400, PS285 and PS398. E IV 146 UN E-TYPE SPECIMEN HOSPITAL MBL INTEG T3SS E1 WOUND FMCI VIM, NDM INT1 S E1 WOUND FMCO VIM INT1 S E2 A WOUND FMCO VIM, NDM INT1,2 S E3 WOUND OTHI VIM INT1 U E4 WOUND FMCO VIM INT1 S E4 WOUND OTHI VIM INT1 U E5 WOUND FFMCO - - S E6 WOUND UCHI - - U E7 WOUND UCHI VIM INT1 U E7 WOUND OTHS VIM INT1 S B E8 URINE UCHI NDM INT1 U E8 T. ASP. UCHI VIM INT1 S E9 WOUND UCHI VIM - U E10 WOUND FMCI VIM INT1 S E11 WOUND OTHI VIM INT1 S E12 WOUND OTHI NDM INT1 US E12 WOUND LTHO NDM INT1 U E13 WOUND FMCO NDM - S E14 WOUND FMCI NDM INT1 U E14 WOUND OTHI NDM INT1 U C E15 WOUND FMCO VIM INYT1 S E15 WOUND OTHI - - S E16 EAR FMCA VIM INT1 S E17 WOUND UCHI VAIM RINT1 S E17 EAR OTHS VIM INT1 S E18 URINE UCHI NDM INT1 U E18 WOUND FMCA - - S E19 WOUND UCHI NDM - S D E19 WOUND OTHRI VIM, NDM - U E20 WOUND OTHS NDM INT1 U E21 WOUND FMCI VIM INT1 S E E21 BWOUND FMCO VIM - S E22 LIWOUND LTHO NDM INT1 U E22 WOUND B. UCHI, NDM INT1 U E23 NURINE OTHI NDM INT1 U F E24 WOUND UCHI NDM INT1 U E25 WOUND OTHI NDM - U AE26 WOUND UCHI - INT1 S E26 WOUND FMCO VIM INT1 U E27 D WOUND UCHI VIM INT1 S E28 WOUND OTHI NDM - S G E28 WOUND LTHO VIM - S A E29 WOUND FMCO VIM INT1,2 S E29 WOUND UCHI VIM - S B E30 WOUND OTHS NDM - S I E31 WOUND UCHI NDM INT1,2 U E31 EAR FMCO NDM INT1 U H E32 WOUND OTHI NDM - S E33 WOUND LTHO NDM INT1 S F E34 T. ASP. UCHI NDM INT1 U I E34 EAR LTHO VIM, NDM INT1 U INT1 S O E35 WOUND LTHO - E36 WOUND FMCO NDM INT1,2 U J E36 WOUND FMCI VIM, NDM INT1,2 US E37 WOUND B. LTHO NDM INT1 U Y E37 URINE OTHI - - S E38 WOUND UCHI NDM INT1,2 U K T E39 WOUND UCHI - - U I E39 WOUND FMCI VIM, NDM - S E40 WOUND FMCO VIM, NDM INT1,2 S E41 WOUND OTHI VIM, NDM INT1 U S L E41 WOUND FMCI NDM INT1,2 U E42 WOUND OTHS VIM - S M R E42 WOUND FMCI NDM INT1,2 S E43 WOUND FMCO VIM, NDM - S N E E43 WOUND FMCI VIM INT1 S E44 URINE UCHI VIM INT1 U O E44 WOUND UCHI VIM, NDM INT1 U V E45 WOUND LTHO NDM - U I E46 EAR UCHI - INT1 S E47 WOUND FMCO VIM INT1 U P N E47 URINE OTHI VIM INT1 U E48 URINE UCHI - INT1 U E49 URINE UCHI - - U U E 50 WOUND UCHI - - U Figure 4.9. Dendrogram showing cluster analysis of carbapenem-resistant P. aeruginosa strains by ERIC-PCR using Phylotree software and Unweighted Pair Group Method with Arithmetic Mean (UPGMA). Key: U= exoU; S= exoS 147 Y R L 2 3 4 5 6 7 8 9 10 11 12 13 14 15 L 17 18 19 20 21 22 23 24 25 2R6 27A 28 29 30 3.5kb LI B 3.0 kb 2.5 kb 2.0 kb N 1.5 kb 1.0 kb AD A 750 bp 500 bp B 250 bp OF I TY Plate 4.15. Representative image ofI fingerprinting patterns of carbapenem-resistant P. aeruginosa strains by gel electrophoresis following BOX-PCR; Lanes L = 1S kb plus ladder, Lanes 2-15, 17-30 = PS386, PS353, PS392, PS346, PS354, PS393, PS222, PS291, PS303, PS395, PS096, PS349R, PS093, PS292, PS235, PS184, PS352, PS244, PS170, PS181, PS230, PS397, PS253, PS409, 348, PS243, PS220 and PS246. E IV 148 UN B-TYPE SPECIMEN HOSPITAL MBL INTEG. T3SS B1 WOUND FMCI VIM, NDM INT1 S B1 WOUND FMCO A VIM, NDM - S B2 WOUND FMCO VIM INT1 S B2 WOUND FMCO VIM INT1 S B3 WOUND OTHI VIM INT1 S B4 WOUND OTHS VIM INT1 S B5 WOUND FMCI VIM INT1 S B5 WOUND FMCI NDM INT1,2 U B6 WOUND FMCI VIM INT1 S B7 URINE UCHI VIM INT1 U B7 WOUND FMCI VIM, NDM INT1,2 US B8 WOUND UCHI - - S B B9 WOUND OTHI NDM - S B10 WOUND UCHI NDM INT1,2 U B10 WOUND OTHS VIM - S B10 Y WOUND UCHI VIM, NDM INT1 U B11 WOUND FMCO NDM INT1,2 U B12 WOUND OTHI VIM, NDM INT1 U B13 WOUND FMCI NDM INT1,2 S B14 EAR FMCO NDM RINT1 U B15 WOUND FMCO VIM INT1 S B16 URINE UCHI NDM INT1 U C B16 WOUND UCHI - A INT1 S B17 WOUND FMCRO VIM, NDM INT1,2 S B17 D WOUND B. LTHO NDM INT1 U B18 WOUNDI BLTHO NDM - U B19 WOUND OTHI NDM INT1 US B19 WOLUND FMCI NDM INT1 U B20 WOUND OTHS NDM INT1 U B21 EAR UCHI - INT1 S B22 WOUND OTHI NDM INT1 U NB22 WOUND UCHI NDM - S E B23 WOUND FMCI VIM INT1 S AB24 URINE OTHI - - S B24 WOUND UCHI VIM INT1 S B24 EAR OTHS VIM INT1 S D B25 WOUND OTHI VIM, NDM - U B26 EAR FMCA VIM INT1 S A B27 WOUND FMCO NDM - S B27 F WOUND FMCA - - S B28 WOUND OTHI B VIM INT1 U B29 WOUND LTHO I NDM - U B30 T. ASP. UCHI VIM INT1 S B30 URINE OTHI VIM INT1 U B31 WOUND FMCO G - - S F B32 URINE UCHI - - U B32 URINE UCHI NDM INT1 U B33 WOUND OTHS NDM - S O B34 WOUND FMCO VIM INT1 U B34 WOUND FMCO VIM INT1 U H B35 WOUND LTHO VIM - S Y B35 URINE UCHI - INT1 U B36 WOUND LTHO NDM INT1 U T B36 WOUND UCHI NDM INT1 U I B37 WOUND B. UCHI NDM INT1 U B37 WOUND OTHI NDM - S B38 I URINE OTHI NDM INT1 U S B39 WOUND FMCO VIM INT1,2 S B40 WOUND UCHI VIM INT1 S R B40 WOUND UCHI VIM - S B41 WOUND FMCO VIM, NDM INT1,2 S B42 WOUND OTHI NDM - U E B43 WOUND UCHI - - U J B44 WOUND UCHI VIM INT1 U V B44 WOUND OTHI VIM INT1 U I B45 WOUND LTHO NDM INT1 U B46 WOUND FMCO VIM - S B47 WOUND UCHI - - U N B48 WOUND UCHI VIM - U B49 T. ASP. UCHI NDM INT1 U K U B49 EAR LTHO V IM, NDM INT1 U B50 WOUND LTHO - INT1 S L B50 WOUND OTHI - - S B51 WOUND UCHI NDM I NT1 S B52 WOUND UCHI NDM INT1,2 U Figure 4.10. Dendrogram showing cluster analysis of carbapenem resistant P. aeruginosa by BOX-PCR using Phylotree software and Unweighted Pair Group Method with Arithmetic Mean (UPGMA). Key: U = exoU, S = exoS 149 Table 4.22. Comparison between REP-, ERIC- and BOX- PCR Characteristic REP-PCR ERIC-PCR BOX-PCR Fingerprint pattern 48 (27 CT, 21 ST) 50 (25CT, 25 ST) 52 (22 CT, 30 ST) No of bands 2 – 14 4 – 15 7 – 15 Molecular weight 200 – 4000 bp 100 – 3500 bp 250 – 3500 bp Discriminatory 0.854 0.934 0.896 index (D) Y Typeable strains 61/75 strains 75/75 strains 66/75 straiRns Clusters 8 16 12 A Single isolates 4 6 11 R Prevalent cluster A (18 isolates) C (11 isolates) LIBB (18 isolates) Key: CT = Common type N ST = Single type DA BA I O F Y IT ER S NI V U 150 Table 4.23. Indistinguishable isolates with REP-PCR Clone Isolate Hospital T3SS MBL ERIC-PCR BOX-PCR cluster cluster 01 PS303 FMCI exoU blaNDM C E PS395 OTHI exoU blaNDM C E 02 PS235 LTHO exoU blaNDM F I PS184 UCHI exoU blaNDM F I 03 PS296 FMCI exoS blaVIM, blaNDM A YA PS351 LTHO exoS - I R L 04 PS205 UCHI exoS blaVIM C A E PS088 FMCA exoS - D R F 05 PS100 FMCI exoS blaVIM IB B PS022 UCHI exoS blaVIM L B G 06 PS166 UCHI exoU - N P G PS350 OTHI exoU blaVAIM - H 07 PS168 UCHI exoU AbDlaVIM B J PS209 OTHI eIxBoU blaVIM P G O F ITY RS E NI V U 151 Table 4.24. Indistinguishable isolates with BOX-PCR Clone Isolate Hospital T3SS MBL ERIC-PCR REP-PCR cluster cluster 01 PS297 FMCI exoS blaVIM E F PS285 FMCI exoU blaNDM L B 02 PS202 OTHS exoS blaVIM M A PS383 UCHI exoU blaNDM H G PS250 UCHI exoU blaVIM, blaNDM O BY 03 PS303 FMCI exoU blaNDM C RB PS354 OTHI exoU, exoS blaNDM C A B 04 PS238 LTHO exoU blaVIM, blaNDM I R C PS152 UCHI exoU blaNDM LII B C DA N A F I B O ITY ER S V UN I 152 RY Table 4.25. Genotypic characteristic of four clones confirmed by two or more methods A Clone Isolate Clinical source MBL(s) Integron Class 1 T3SS Efflux geneR(s) ampC oprD integrase B 01 PS022 Tracheal aspirate blaVIM intI1 3.5 kb exoS - I - + PS100 Wound blaVIM intI1 3.5 kb exoS mLN exA, mexB, mexX - + 02 PS303 Wound blaNDM intI1 3.5 kb exoU mexX - + PS395 Wound blaNDM intI1 3.5 kb exoU mexA, mexX, mexY - + PS354 Wound blaNDM intI1 NA ex AoU, exoS ND ND ND 03 PS235 Wound blaNDM intI1 NAA DexoU ND ND ND PS184 Wound biopsy blaNDM intI1 B3.5 kb exoU ND ND ND 04 PS238 Ear blaVIM, blaNDM intI1 3.5 kb exoU mexB, mexX, mexY - + PS152 Tracheal aspirate blaNDM intFI1 I NA exoU mexX - + Key: + = present; - = absent; ND = Not determined; NOA = Not amplified SI TY ER NI V U 153 CHAPTER FIVE DISCUSSION 5.1 Distribution of isolates in clinical samples Pseudomonas aeruginosa is frequently recovered from wound infection. WYou nd isolates were predominant (69.3%) in this study followed by isolates from eaRr (11.4%) which was in agreement with the report by Oladipo et al. (2015) anAd Brown and Izundu, (2004). A study from Iran also documented highest pRrevalence of P. aeruginosa in wound samples with the percentage (69.9%) thaIt Bwas similar to what was obtained from this study (Khosravi et al. 2017). Stud ieLs from Southeastern and Central Nigeria have also documented highest prevaNlence of P. aeruginosa from wound sample (Eyo et al., 2015; Zubair and IregbuA, 2018). However, a report from Southwestern Nigeria with highest prevalenAce Dof P. aeruginosa from urine sample does not agree with findings from this studBy (Odumosu et al., 2012). 5.2 Antibiotic susceptibility prFofil eI of clinical isolates of P. aeruginosa Pseudomonas aeruginosa is a reOcognised Gram-negative bacterium which is inherently resistant to many antibioticYs d ue to collection of resistance mechanisms such as efflux pumps and antibioticI Thydrolyzing enzymes that are disseminated through mobile genetic elements. ISn this present study, all the isolates were resistant to more than four antibiotics. ToRtal resistance to ampicillin, cephalothin and cefuroxime was observed in this presenEt study while a study in Egypt reported 95.8% resistance to cephalothin and cefuroIxVime (Afifi et al., 2013). Fluoroquinolones such as ciprofloxacin and levNofloxacin are known to be potent extensive-spectrum antibiotics with efficacy on an Uextensive collection of bacteria including P. aeruginosa. This study reported higher resistant rates of 39.3% and 38.1% to ciprofloxacin and levofloxacin, respectively when compared to the report of the study from Southwestern Nigeria with resistance rate of 27.8% and 37.0%, respectively (Odumosu et al., 2012), while higher resistance rate of 42.9% and 47.6% to ciprofloxacin and ofloxacin was reported by another study from southwestern Nigeria (Igbalajobi et al., 2016). Resistance to ofloxacin observed in this study (43.3%) was lower than the report by Olayinka et al. (2009) with 154 resistance rate of 82.6%. Gentamicin is considered as the first drug to select when it comes to treatment of P. aeruginosa infections (Oduyebo et al., 1997), nonetheless, this study reports that 43.3% of P. aeruginosa isolates were not sensitive to gentamicin. This frequency was in range with earlier reports in which 40.7% resistant strains were observed (Odumosu et al., 2012) but higher than the report from Egypt with 21.1% resistant strains (Afifi et al., 2013). This study showed high resistance rate of 32.6% and 27.4% to tobramycin and amikacin correspondingly. Amikacin is considered as a replacement in the treatment of Gram-negative bacterial infectioYns t hat are resistant to tobramycin, but has now been observed to have almost the sRame level of resistance with tobramycin (Brown and Izundu, 2004). A This present study reports higher resistance rate to third and fRourth generation cephalosporins with percentage resistance of 45.8, 38.1 anIdB 54.4 to cefepime, ceftazidime and ceftriaxone, respectively in contrast with th eL report of Odumosu et al. (2012) from the same region who recorded resistance raNte of 9.3%, 14.8% and 27.8% to cefepime, ceftazidime and ceftriaxone, respectiveAly. Higher resistance to cefepime was noticed- in this study when linked with priDor information from the same region (Oladipo et al., 2015) where 6.0% BweAre resistant. Increasing resistance to cephalosporins most especially the foIurth generation cephalosporins necessitates proper guidelines for antibiotic presFcriptions. The rate of resistance of P. aeruginosa to aztreonam (16.7%) found in thOis study was less than the percentage reported from Nigeria (33.3%), Egypt (4Y6.1%) and Turkey (48.0%) (Odumosu et al., 2012; Afifi et al., 2013; GuvensenI Tet al., 2017). In this study, only 16 (21.9%) CRPA were susceptible to aztrSeonam which is the only beta lactam with efficacy against MBL- producing EGraRm-negative bacteria but cannot resist hydrolysis by ESBLs, suggesting the presVence of ESBLs among carbapenem resistance P. aeruginosa. I CaNrbapenems are potent and are used as last-line antibiotics in the therapy of Gram-Unegative bacterial infections especially the ESBLs. However, resistance to carbapenems is on the rise (Nabarro et al., 2017). Carbapenem resistance as recorded in this study was defined as non-susceptibility to at least one carbapenem (imipenem, meropenem and doripenem). Overall prevalence of 18.8% CRPA described in this study was higher than the report of 15.2% from Nigeria in 2015 and lower than 19.6% from India in 2008 (Javiya et al., 2008; Eyo et al., 2015). In the systematic evaluation on carbapenemase-producing Gram-negative bacteria in clinical settings in Africa by 155 Manenzhe et al. (2015), the overall prevalence ranged from as low as 2.3% and 9.0% to as high as 60.0% and 67.7% in sub-Saharan Africa and North Africa, respectively. Among the carbapenems, reduced resistance to imipenem (15.8%) was observed compared with meropenem (18.8%) and doripenem (17.8%) in this study. In a previous work on multidrug resistant P. aeruginosa from southwest Nigeria, Odumosu et al. (2012) reported the prevalence of 1.9% resistance to imipenem, much lower than the 15.8% reported in this study. This suggests a massive increase in the emergence of imipenem resistance. This study observed higher resistance to meropenem (18.8 %) compared to imipenem (15.8%) as reported in a study from Calabar, NigerRia (EYyo et al., 2015) but contrary to the report from Egypt with higher resistance ratAe of 41.4% to imipenem than 18.0% resistance to meropenem (Afifi et al., 2013).R The explanation for the high variation in resistance observed among carbapeneImBs could probably be due to differences in prescription among hospitals since i mLipenem has been linked with progression of resistance in the course of treatmenNt (Carmeli et al., 1999). High resistance to doripenem was observed in this studAy. This is because doripenem is structurally related to imipenem leading to crossD resistance to this antibiotic. This also calls for concern as this antibiotic is not inB cliAnical use in Nigeria. I The result of this study revealed that p olymyxin B and colistin sulphate were the only antibiotics that were active ag aOinst Fmajority of carbapenem-resistant P. aeruginosa in this study. Majority of carbapenem resistant isolates (68.5%) were extensively drug resistant, showing resIiTstanYce to virtually all the antipseudomonal antibiotics except colistin sulphate and polymyxin B. This is in conformity with the report by Palavutotai et al. (2018), RwheSre all the extensively drug resistant strains of P. aeruginosa were susceptibleE to colistin. Carbapenem resistant Gram-negative bacteria showing resistaInVce to all antibiotics apart from colistin have been reported. In carbapenem- seNnsitive strains, lower resistance rates of 6.2% and 4.9% were recorded to colistin and Upolymyxin B, respectively as experimental in last studies (Pitout et al., 2005; Odumosu et al., 2012; Guvensen et al., 2017). This emerging resistance to polypeptides also calls for serious intervention because it is the only antibiotic that was active on greater number of MBL-producing Gram-negative bacteria. 156 5.3 Evaluation of antibiotic susceptibility profile of phenotypically detected MBL-producing and non-MBL-producing strains Statistical analysis revealed significant difference in activity of seventeen (17) antibiotics against MBL-positive and MBL-negative P. aeruginosa isolates (p < 0.05) while in six antibiotics statistical analysis showed no significant difference in sensitivity profile of MBL-positive and MBL-negative P. aeruginosa isolates (p < 0.05). Varaiya et al. (2008) have also reported statistically sisgnificant difference in sensitivity profile of cefepime, ceftazidime ceftriaxone, and ciprofloxacin as observ ed in this study. Both MBL-positive and MBL-negative strains were 100% unaRffecYted by ampicillin, cephalothin and cefuroxime while only MBL-positive strains sAhowed 100% resistance to ticarcillin, ceftazidime, ceftriaxone, doripenem and meRropenem. Across the seven hospitals, only colistin sulphate and polymyxin B exhIiBbited highest activity against MBL-positive strains when compared with MBL-ne gaLtive strains. This implies that MBL-genes do not confer resistance to colistinN sulphate and polymyxin B. However, MBL-positive strains remained totally reAsistant to imipenem, meropenem and doripenem across six hospitals (UCHI, FAMCDA, OTHS, LTHO, OTHI and FMCI) while 7.1% sensitivity to imipenem was Bobserved among MBL-positive strains from FMCO as compared to 100% sensitiv itIy exhibited by MBL-negative strains. This is also an indication that MBL pFroduction significantly affects the efficacy of carbapenems (imipenem, mero pOenem likewise doripenem). 5.4 Molecular detIeTctioYn of carbapenemases Genes encoding bSlaSME, blaNMC-A, blaGES and blaBIC-1 class A carbapenemases were investigated inR this study, none of the isolates was positive for blaSME, blaNMC-A, blaGES and blaBIC-E1 enzymes. However, only blaGES has been reported in P. aeruginosa from severaIl Vcountries like Brazil, China, Korea, Poland and Saudi Arabia (Wang et al., 20N06; Ahmed and Asghar, 2017). blaKPC is also a class A carbapenemase commonly Ureported in Enterobacteriaceae especially in K. pneumoniae but has also been described in P. aeruginosa (Mushi et al., 2014; Galetti et al., 2016). CRPA isolates were not screened for blaKPC in this study. Among the class D carbapenemases, blaOXA-48 and blaOXA-58 were sought but not detected in this study. There has heretofore been only one report of blaOXA-48 in Enterobacteriaceae from Nigeria (Jesumirhewe et al., 2017). Although there are 157 several reports of blaOXA-48 in Gram-negative bacteria including P. aeruginosa from African countries, blaOXA-58 was not common (Manenzhe et al., 2015). The genes encoding class B carbapenemases (MBLs) were investigated by means of primers specific for blaIMP, blaVIM, blaSIM, blaSPM, blaSIM, blaNDM, blaGIM, blaAIM and blaDIM. Only blaVIM and blaNDM were amplified. The prevalence of blaVIM and blaNDM in carbapenem-resistant P. aeruginosa in this study was 86.3%, a rate which was extremely higher than that documented by Mohanam and Menon (2017) who repor ted that blaVIM and blaNDM were prevalent in 55% carbapenem-resistant P. aeruYginosa from India. The whole incidence of MBL-genes in P. aeruginosa from Rsouthwest Nigeria was found to be 14.7%, a rate which was higher than thatR of Athe study by Zubair and Iregbu (2018) from Central Nigeria with the prevalIeBnce rate of 2.5% but almost agrees with the report from Thailand with overall LMBL prevalence rate of 17.3% in clinical isolates of P. aeruginosa (Piyakul etN al., 2012). The prevalence of blaVIM and blaNDM in this study were 8.1% and A8.6% respectively, while the co-existence of blaVIM with blaNDM was 2.1%. D Sequence analysis of nine randomly seleBcteAd amplicons for blaVIM revealed 99.46 – 99.73% identity with blaVIM-5 while sIequence analysis of nine randomly selected blaNDM amplicons revealed that theF isolates were 100% identical to blaNDM-1. Several studies from Africa have als oO reported the presence of blaVIM-2 in P. aeruginosa (Manenzhe et al., 2015). IYn Africa, blaVIM-2 has been confined to P. aeruginosa with the exception of the reIpTort of blaVIM-2 producing E. coli from Tunisia (Manenzhe et al., 2015). Zubair andS Iregbu (2018) have also reported the presence of blaVIM-2 in clinical isolates of EP. aReruginosa from Nigeria, there is no single report of blaVIM-5 (i. e. variant numberV 5) in clinical isolates of P. aeruginosa from Africa. However, a recent study byN AdIelowo et al. (2018) has reported the occurrence of blaVIM-5 in environmental Ustrains of Pseudomonas putida from Nigeria. This study is the first report of blaVIM-5 (i. e. variant number 5) in clinical isolates of P. aeruginosa from Nigeria. In Africa, blaVIM-4 was first reported in Pseudomonas aeruginosa Northeastern Algeria (Mellouk et al., 2016), followed by a report from Egypt (Hashem et al., 2017). Another report from Algeria has recently documented the presence of blaVIM-4 in Pseudomonas aeruginosa (Merradi et al., 2019). 158 The blaNDM-1 gene was first discovered in Enterobacteriaceae from Swedish patient who received treatment in a hospital in New Delhi, India (Pitout et al., 2005). Since then, majority of reported cases of blaNDM-1 have been linked with international travel or hospitalisation in India or Pakistan (Govind et al., 2013). In South Africa, blaNDM-1 was discovered in Enterobacter cloacae isolated from urine of a patient in South Africa. The patient had been on admission in India prior to this discovery (Govind et al., 2013). In Africa, blaNDM-1 was first noticed in K. pneumoniae isolated from urine sample of a patient in Kenya (Poirel et al., 2011). Several studies have a lso documented cases of blaNDM-1 in P. aeruginosa (Jovcic et al., 2011; Liew etR al.,Y 2018; Mohamed et al., 2018). Nevertheless, there is limited data with respect toA the incidence of blaNDM-1 in Gram-negative bacteria from Nigeria. Abdullahi etR al. (2017) have documented the presence of blaNDM-1 in EnterobacteriaceaeI Bfrom Northwestern Nigeria. Emergence of blaNDM-1 in Enterobacteriaceae from pLoultry in Nigeria has also been documented (Ogunleye et al., 2016). Zubair and IrNegbu (2018) have reported the presence of blaVIM-2 and the absence of blaNDM-1 in Aclinical isolates of P. aeruginosa. To the best of my understanding, this represeAntsD the first report of blaNDM-1 in clinical isolates of P. aeruginosa from Nigeria. TBhis is also the first report of co-existence of blaNDM-1 with blaVIM-5 and the first det eIction of blaVIM-5 (variant 5) in clinical isolates of P. aeruginosa from Africa. MohFanam and Menon (2017) have also documented the co-existence of blaVIM with b laONDM in clinical isolates of P. aeruginosa from India. There have been several reYports of blaNDM-1 co-existing with other carbapenemases in Gram-negative bacterIiaT. The concurrence of blaNDM-1 with blaOXA-23 in Acinetobacter baumannii from InSdia and Nepal has been reported (Karthikeyan et al., 2010; Joshi et al., 2017). EAlsRo, the coexistence of blaNDM-1 with blaOXA-48 in Enterobacteriaceae from India anVd K. pneumoniae from China has been reported (Karthikeyan et al., 2010; Xie et Nal.,I 2017). Likewise, coexistence of blaNDM-1 with blaKPC in Escherichia coli from USouthern Vietnam was recently documented (Hoang et al., 2019) The presence of MBL genes on plasmids is an evidence of parallel transmission of these resistance genes among bacterial species (Karthikeyan et al., 2010). Amplification of transformant plasmid DNA with primers specific for blaVIM and blaNDM gave amplicon sizes that correspond to the expected products. This confirms that both blaVIM and blaNDM are carried on resistance plasmids which are capable of 159 parallel transmission of resistance genetic material between bacterial genera and species. This indicates a serious threat to our health care delivery system. Modified Hodge Test (MHT) was not successful at detecting carbapenemase production in clinical isolates of P. aeruginosa in this study. This is in conformity with the report that MHT was previously recommended for screening carbapenem-resistant Enterobacteriaceae and not Gram-negative non-fermenters of which P. aeruginosa is among (Gniadek et al., 2016) but is no longer recognized as screening method for carbapenemase production because of ambiguity about the isolates real suscepYtibility to carbapenems (Humphries et al., 2019). Combined disc test (CDT) pAhenRotypically detected 66 (15.4%) isolates as MBL-producers while MBL-genes were truly present in 58 of the 66 isolates; that is, CDT wrongly identified 8 isolates aRs MBL-producers with PCR as the „gold standard‟. CDT also failed to detect MLBIL B in 5 isolates that in reality carried blaVIM and blaNDM. The sensitivity and specificity of CDT was 92.1% and 33.3%, respectively while positive predictive valAuesN (PPV) and negative predictive values (NPV) of 87.9% and 55.6%, respectivelDy were obtained. This is also in line with the report that EDTA-based assay prodAuced false-positive results as observed in this study (Gniadek et al., 2016). The pIrBevalence of MBL in P. aeruginosa by CDT test (phenotypic) 15.4% versus genFoty pic 14.7% shows that CDT may be also used to detect MBL in P. aeruginosa w Ohere molecular screening methods are not available. Among the MBLs, blaIMYP, blaVIM and blaNDM, are the most commonly reported enzymes from AfricIa Twhile blaDIM was reported in a study from Sierra Leone (Manenzhe et al., S2015). Though blaIMP was among the most commonly reported MBL from AfricEa, Ronly a study from Nigeria has documented blaIMP-1 in E. coli from AbattoiVr (Chika et al., 2017). This study did not detect blaIMP, blaSIM, blaSPM, blaGIM, blNaAIMI and blaDIM. There have been no reported cases of blaSIM, blaSPM, blaGIM, blaAIM Uand blaDIM MBL-encoding genes from Nigeria. The blaAIM has not been reported again since its discovery in P. aeruginosa (Yong et al., 2006). IMP is more prevalent in Europe and Asian countries (Nordmann et al., 2011). All the isolates that possessed MBL(s) gene were resistant to the three antipseudomonad carbapenems with the exclusion of a strain, PS219, which was resistant to only meropenem and doripenem. Among the ten carbapenem-resistant clinical isolates that were negative for all the MBL genes tested, eight were resistant to 160 all the three carbapenems used in this study while one isolate (PS088) was sensitive to imipenem but resistant to meropenem and doripenem and the remaining one of the ten isolate was sensitive to imipenem and doripenem but showed resistance to only meropenem. The fact that 13.7% of carbapenem resistant P. aeruginosa do not possess any MBL- encoding gene as observed in this present study confirms that carbapenem resistance is not attributable to the existence of known carbapenemases alone. Other resistan ce mechanisms such as efflux pump overexpression, ampC cephalosporinasYe and deficient outer membrane porin also perform a contributory role in the Rresistance observed towards carbapenems and other antibiotics in this study. TRhereA may also be previously undescribed mechanisms. IB 5.5 PCR-RFLP analysis of integrons in carbapenem-r eLsistant P. aeruginosa Integrons are heritable elements that are capable of acquNiring gene cassettes that carry antibiotic resistance genes that can be transferred Afrom one bacterial species to the other (Poirel et al., 2001). The role of inAtegrDons in the distribution of antibiotic resistance genetic material in Gram-negaBtive bacteria is well documented (Mohanam and Menon, 2017). Dissemination o fI class 1 integron-carrying blaVIM has been described (Touati et al., 2013). QuiFte a lot of studies have described the prevalence of class 1 and 2 integrons in Gra mO-negative bacteria (Odumosu et al., 2013; Alabi et al., 2017; Izadi et al., 2017Y). Class 1 integrons alone was detected in 57.5% of carbapenem-resistant IPT. aeruginosa in this study. This correlates with the work of Odumosu et al. (S2013) who stated the prevalence of 57% in clinical isolate of P. aeruginosaE froRm the same region but less than the report of Hassuna et al. (2015) who detectedV class 1 integrons in 71% P. aeruginosa from wound patients in Egypt. HoNweIver, Odumosu and his co-workers did not detect class 2 integrons while this Ustudy found co-existence of class 1 and 2 integrons in 12.3% CRPA. To the best of my understanding, this is the first documentation of class 2 integrons and co-existence of class 1 and 2 integrons in P. aeruginosa from Nigeria, though class 2 integrons have been documented in Enterobacteriaceae from Southwest Nigeria (Alabi et al., 2017; Odetoyin et al., 2017). Khosravi et al. (2017) have documented the co-existence of class 1 and 2 integrons in P. aeruginosa from burn wound in Iran. 161 In this study, two strains (PS396 and PS405) with co-existence of blaVIM-5 and blaNDM-1 genes did not harbour integrons. This result corresponds with the report by Zubair and Iregbu (2018) that all the blaVIM-2 containing isolates did not harbour integrons. These reports imply that blaVIM and blaNDM containing P. aeruginosa isolates may not contain class 1 integrons as well as other integron classes. Although statistical analysis revealed positive association between integrons and MBL genes (p = 0.0064). This study was able to amplify cassette region that was not greater than 3.6 kb in s ize but gene cassette could not be amplified in 33.3% of the integron-containing isYolates. This report is compatible with the study of Odetoyin et al. (2017), which weRre able to amplify integrase of not greater than 4.0 kb with some unamplifiableR intAegrase 1 gene cassette in E. coli from Nigeria. IB 5.6 MDR efflux pump overexpression, ampC ov eLrexpression and oprD underexpression in carbapenem-resistant P. aNeruginosa Natural expression of efflux pumps in unmutated Astrains of P. aeruginosa plays a worthwhile part in moderately reduced susceAptibDility to antibiotics but overexpression of these efflux pumps in mutants leads to raised level of resistance to antibiotics (Lister, 2002). Efflux pumps have capa cIitBy to eject numerous categories of antibiotics from the cytoplasm and the periplasFmic space, subsequently leading to cross-resistance to other classes of antibiotics bOecause inhibitory concentration of antibiotic could not be achieved at the target sYite of action (Lister, 2002). Out of the 12 RND type efflux pumps so far identifieIdT, four are of clinical importance. These include MexAB-OprM, MexCD-OprJ, MeSxEF-OprN and MexXY-OprM (Mesaros et al., 2007). MexXY- OprM expoErts Rfluoroquinolones, aminoglycosides, and cefepime while MexAB-OprM exportsV beta-lactams, macrolides, fluoroquinolones, novobiocin, sulfonamides, triNmetIhoprim, tetracycline, chloramphenicol and meropenem (Koo, 2015). In this Ustudy, expression of two efflux genes was determined in each of the four efflux systems. Efflux pump overexpression was detected in 68.8% of CRPA in this study. MexXY-OprM was the most expressed efflux pump in this study with 58.3% prevalence. This is in support of the work of Xavier et al. (2010) who also reported highest prevalence of MexXY-OprM pump with prevalence of 50.8% in P. aeruginosa from Brazil. MexAB-OprM was overexpressed in 47.9% of isolates in this study. This was lower than the report by Pourakbari et al. (2016) which showed that 62% of P. aeruginosa overexpressed MexAB-OprM. Both MexAB-OprM and MexXY-OprM 162 were overexpressed simultaneously in 18 (37.5%) isolates. Co-expression of two or more efflux pumps in clinical strains of P. aeruginosa has been documented. For example, concomitant overexpression of MexAB-OprM, and MexXY-OprM showing resistance to multiple antibiotics in P. aeruginosa was reported from France (Llanes et al., 2004). MexAB-OprM and MexXY-OprM were naturally expressed in „wild type‟ cell. MexCD-OprJ and MexEF-OprN were not active in unmutated cell but overexpressed in mutants (Koo, 2015). Since all these carbapenem-resistant strains were also impervious to many other classes of antibiotics together w ith fluoroquinolones, aminoglycosides, beta-lactams and monobactam, wRhicYh are substrates for these efflux pumps, this underscores the task of efflAux pumps in antibiotic resistance. BR Overexpression of efflux pump genes has been associated wiLth Iantibiotic resistance in P. aeruginosa. However, high level of resistance may no t result from efflux pump expression alone. In this study, majority of the isolaAtes Nwith >64 µg/mL to imipenem and meropenem had combination of MBDL-encoding genes, efflux pump overexpression and /or oprD down-regulatioAn. Resistance to carbapenems could be linked to efflux pump overexpressionI Balone in four isolates. Overexpression of MexAB-OprM was the resistant meFcha nism identified in one stran (PS007), which had MIC of 16 µg/mL agains t Oimipenem and meropenem while MexXY-OprM overexpression was the onYly resistance mechanism identified in strain PS088 with MIC of 32 and 16 µg/mL aIgTainst imipenem and meropenem, respectively. Consistent with the previous reporSt by Hocquet et al. (2006) simultaneous overexpression of MexAB- OprM and MeRxXY-OprM was found in two strains (PS182 and PS349). These isolates had intact EoprD, lacked ampC overexpression and MBL genes with MIC of 4 and 8 µg/mLI Vagainst imipenem with MIC of 8 and >64 µg/mL against meropenem, reNspectively. This study also observed that isolates with efflux pump alone had reduced UMIC against imipenem, meropenem, ceftazidime, cefepime and ciprofloxacin when compared with isolates showing combination of MBL resistance genes and efflux pump. This study also observed that 77.1% of CRPA had oprD down-regulation. OprD porin is the only porin through which carbapenems enter the bacterial cell. Mutation or alteration in genes that code for OprD protein results in resistance to carbapenems and 163 reduced expression of oprD. Deficient OprD porin is mainly seen in carbapenem- resistant P. aeruginosa (Livermore, 1992). Deficient OprD as a result of mutations has been reported to confer high resistance to imipenem than meropenem and doripenem (Livermore, 1992; Quale et al., 2006). Carbapenem resistance in three strains was due to loss of OprD porin alone. MBL-encoding genes, efflux pump and ampC overexpression were not present in these isolates. This shows that deficient oprD porin alone could contribute to carbapenem resistance. Increased ampC expression was found in 13 (27.1%) isolates. This outcome wYas in concordance with the report of Cabot et al. (2011) who stated that 24.2R% of the isolates overexpressed ampC but higher than the report of Xavier eRt alA. (2010) who documented that 11.9% of P. aeruginosa isolates from Brazil overexpressed ampC. The overexpression of ampC as a resistance mechanism couIldB not be linked with carbapenem resistance in this study, because all the isolates Lwhich overexpressed this gene also have other carbapenem resistant mechanismsN. This was backed-up with the report that ampC overexpression alone in DP. Aaeruginosa could only lead to insignificant reduction in the activity of Acarbapenems but combination of other mechanisms of resistance could hav e IaB great impact in lowering the activity of carbapenems (Quale et al., 2006).F Therefore, ampC overexpression alone may not necessarily be responsible for cOarbapenem resistance. This study for the first time in Nigeria has quantified gYene expression (efflux pumps, oprD and ampC) in P. aeruginosa. This studIy Talso identified the entire four clinically important efflux pumps in P. aeruginosa. RS5.7 AssEociation of carbapenem resistance with increased expression of ampC, Vefflux pump and oprD underexpression ConcoImitant presence of MexAB-OprM, MexXY-OprM and MexEF-OprN was found UinN two carbapenem-resistant isolates (PS170 and PS182) that have increased expression of ampC, lack carbapenemase genes and have intact oprD porin suggesting the impact of efflux pumps and AmpC in antibiotic resistance of these isolates. PS351 which was susceptible to all the three carbapenems tested in this study and other antibiotics did not overexpress any of the efflux genes, ampC and also had intact oprD porin. Co-existence of MexAB-OprM, MexXY-OprM, MexCD-OprJ, ampC and oprD deficient was also observed in two strains (PS202 and PS219). This shows that the part involvement of OprD loss, AmpC and efflux pumps overexpression in antibiotic 164 resistance could not be neglected and there is interconnection among various mechanisms of carbapenem resistance with a significant effect on resistance to carbapenem. Gomaa et al. (2013) documented the presence of all the four efflux pumps (MexAB-OprM, MexXY-OprM, MexCD-OprJ and MexEF-OprN) in 15% of P. aeruginosa, however, one (2.1%) isolate was found to overexpress all the four efflux pumps in this study. This isolate had deficient oprD, increased expression of ampC found to possess all the four efflux pumps suggesting multiple mechanisms of resistance. This isolate was also resistant to all of the antipseudomonal antibioticYs us ed in this study except colistin and Polymyxin B. R 5.8 Combination of resistance mechanisms in CRPA A Eight different combinations of resistance mechanisms IwBere R observed with combinations of MBL, Efflux pumps overexpression and oLprD loss being the most prevalent. The combinations of all the four resistance mNechanisms (MBL, efflux pump overexpression, oprD deficient and ampC overexpresAsion) were exhibited in 8 (16.7%) strains while MBL and oprD deficient alone wDas found also in 8 (16.7%) strains. Combination of efflux pumps and ampC wasA detected in 3 (6.3%) strains while MBL, oprD and ampC was present simulte nIeoBusly in 1 (2.1%) strain only (Table 4.20). Among the ten isolates that lack FMBL-resistance genes, four demonstrated efflux pump overexpression alone, th rOee isolates showed oprD loss alone while the remaining three isolates had the comYbination of efflux pump and ampC overexpression. In this study, ampC overexIprTession alone was not detected as the only mechanism of resistance to carbaSpenems as opposed to the report by Castanheira et al. (2014) where overexpressionR of ampC alone was observed as the sole resistance mechanism in twenty-oneE isolates followed by combination of ampC overexpression and oprD deficieInVt. In this present study, combination of ampC overexpression and oprD deNficient as carbapenem resistance mechanism was not observed. U 5.9 Prevalence of type III effector toxins in CRPA It is believed that type III effector toxins help P. aeruginosa to escape host immune response, stimulate proliferation of the organism at the infection site and also inhibit host DNA synthesis resulting into death of the host cell (Jabalameli et al., 2012). According to Feltman et al. (2001), most clinical isolates of P. aeruginosa encode genes for type III effector toxins, but not all clinical strains are capable of secreting 165 effector toxins. Data on the prevalence of type III effector toxins in P. aeruginosa from West Africa are lacking. This study therefore, has made a novel report of the prevalence of the four type III effector toxins in P. aeruginosa from Nigeria. All the carbapenem-resistant P. aeruginosa isolates screened in this study have both exoT and exoY. This result justifies the report by Roy-Burman et al. (2001) that most strains of P. aeruginosa have exoT and exoY toxins. While several studies have reported the presence of exoT in all the isolates of P. aeruginosa, exoY was not always found in all of the isolates (Jabalameli et al., 2012; Adwan, 2017). This study also described the occurrence of exoU in strains that lack exoS and vice versa except in two sRtrainYs that have both exoU and exoS. This conforms to the report of other studieAs which also document the concurrent presence of exoU and exoS in a few numberR of P. aeruginosa strains (Feltman et al., 2001; Jabalameli et al., 2012; GawishI Bet al., 2013). In this study, prevalence of exoS (49.3%) in strains not having exo UL was a little higher than the prevalence of exoU (48.0%) in strains not having exNoS. This agrees with the work of Adwan (2017) who also reported high prevalence Aof exoS but absence of exoU in all of the clinical isolates thereby disagreeing AwithD the report by Gawish et al. (2013) where 64.7% and 38.2% of the isolates pBroduced exoU and exoS, respectively. There was statistical significant difference in Ithe dissemination of exoU and exoS genes in CRPA isolates (p<0.0001). There Fwas no statistically significant difference in the distribution of exoU and exoS amOong wound and non-wound isolates (p = 0.399). It has been establishedI TthatY exoS, heightens the capability of P. aeruginosa to pass over the epithelial obstaScle of the host cell (Soong et al., 2008). The exoU toxin is the most destructive of Rthe type III effector proteins with phospholipase A2 activity which is only expresEsed by a few number of hospital isolates (Shaver and Hauser, 2004). In this presenIt Vstudy, 48.0% of CRPA strains have exoU toxin, showing 48.0% of CRPA in thNis study to be highly cytotoxic. The health implication of being infected with Ucytotoxic strains that have acquired resistance to virtually all beta-lactam antibiotics is deadly. Therefore, proper implementation of appropriate strategies to curtail the transmission of these antibiotic resistant strains is a matter of urgency and of utmost importance. 166 5.10 Molecular typing of CRPA with three repetitive sequence-based PCR methods The three repetitive sequence based PCR methods; REP-, ERIC- and BOX- PCR have been successfully used by various researchers for typing P. aeruginosa (Syrmis et al., 2004; Wolska et al., 2012; Al-Haik et al., 2016). Syrmis et al. (2004) reported both ERIC- and BOX- PCR as potent tool for typing clinical isolates of P. aeruginosa. However, this study reports ERIC-PCR as the most reliable typing method among the three repetitive sequence-based PCR methods because ERIC-PCR type was able to discriminate all the unrelated 75 isolates and also have higher discriminatorRy inYdex D (0.934) than BOX- and REP- PCR method with discriminatory index Aof 0.896 and 0.854, respectively. R Out of the twelve clusters identified on BOX-profile, four clustIeBrs comprised at least two or more strains displaying 100% BOX-profile (B5, BN10 , B L19, B49), three of which (B5, B19, B49) contained two indistinguishable straAins each while the remaining one contained three indistinguishable strains. The remDaining eight groups clustered two or more isolates that were related to each other Awith identical strains. Out of the four sets of indistinguishable isolates with BOX I-PBCR, two pair of strains (PS238:PS152 and PS303:PS354) were recognised as Fclonal complex based on REP-PCR profile, while ERIC-profile identified PS2 3O8:PS152 and PS303:PS354 as a clone and clonal complex, respectively. Y Out of eight clusters IidTentified on REP-profile, seven sets of indistinguishable strains (R14, R17, R3R1, RS32, R37, R38, R42), found on three clusters comprised at least two or more stErains, displaying 100% REP-profile. These groups clustered two or more isolateIsV that were related to each other with identical strains. Out of the seven pair of inNdistinguishable isolates on REP-profile, PS303:PS395 and PS235:PS184 was Uidentified as clone and clonal complex on ERIC- and BOX-profile, respectively while the third pair of indistinguishable strains, PS100:PS022 was identified as clonal complex on ERIC-profile but exist on distant clusters on BOX-profile. Among the three PCR-based typing methods used in this study, ERIC-PCR was able to distinguish all the seventy-five unique strains of P. aeruginosa unlike REP- and BOX- PCR that could not distinguish between some set of closely related isolates from different hospitals. However, some isolates which clustered together with ERIC-profile 167 were also found in the same cluster on REP- and BOX- profile. For instance, PS147 and PS250 which belong to group O of ERIC-profile was found as part of clonal complex in group B of REP- and BOX- profile. They both contained exoU and class 1 integron with integrase gene of 3.5 kb but PS147 strain carried blaVIM while PS250 co-harboured blaVIM and blaNDM. PS147 was isolated earlier than PS250 from the same hospital; it appeared PS250 strain had later acquired blaNDM through horizontal transfer of plasmid containing blaNDM. Consequently, confirming cross-transmission within hospital. Also, PS235 (LTHO), PS184 (UCHI) and PS348 (OTHI) strains were part of clonal complex. These strains belong to group F on ERIC-profile and were aRlsoY found together in the same cluster B and I of REP- and BOX- profile, respectivAely. The three strains possess blaNDM, class 1 integron and exoU. Integrase 1 geneR found on PS184 was 3.5 kb while integrase 1 gene on PS235 and PS348 coIuBld not be amplified. Likewise, PS152 (UCHI) and PS238 (LTHO) were clonal an dL belong to group I and K of ERIC- and BOX- profile, respectively. These strainNs were part of clonal complex present on cluster C of REP-profile. PS152 harbouAred blaD NDM only while PS238 co- harboured blaVIM and blaNDM but they both cAontain class 1 integron and exoU. PS238 integrase gene was 3.5 kb while PS152 integrase 1 gene could not be amplified. In the same way, PS250 (UCHI), PS335 (FM CIOB), PS293 (FMCI), PS400 (OTHI) were four strains from different hospitals thaFt were part of clonal complex. These strains were grouped together on REP- and BOOX- profile. Three of these strains (PS293, PS335 and PS400) co-harboured blaVYIM and blaNDM; two strains (PS250 and PS400) harboured integrase 1 gene of 3I.5T kb while integrase gene of the remaining two strains (PS335 and PS293) could Snot be amplified. R 5.11 RelEationship between Type Three Secretion System (T3SS) and Repetitive IVElement Sequence-Based PCR (rep-PCR) SiNxteen out of eighteen strains that clustered into group A of REP-PCR type contain UexoS type III effector gene while all the fourteen strains that clustered into group B contain exoU type III effector gene. The only two strains (PS293, PS354) out of seventy-three CRPA strains which produced both exoS and exoU also clustered in this group. On ERIC-PCR dendrogram, all strains in clusters A, E, M, N which comprised 3, 2, 2, 2 isolates, respectively contained exoS gene while all strains in clusters F, L, O, P which contained 4, 2, 2, 3 strains, respectively contained exoU. On BOX-PCR dendrogram, all the strains in clusters A, F, L which comprised 2, 2, 2 strains, 168 respectively contain exoS gene while all strains in clusters J, L which contained 3, 2 strains, respectively contained exoU toxin. It appears there is association between exoU type III effector gene and clonality. This is because majority of the strains found in clones and part of clonal complex contained exoU toxin. This correlates with the report by Wiehlmann et al. (2007) that exoU containing strains dominated clonal complex. However, there was no statistically significant difference in the presence of exoU and exoS genes among clonal and unrelated isolates (p = 0.351). 5.12 Limitations of this study Y 1. Whole genome sequence (WGS) and multilocus sequence typingR (MLST) which are highly discriminatory sequence-based methods as weAll as pulsed- field gel electrophoresis (PFGE) which is the „gold standaRrd‟ for typing P. aeruginosa genome were not used in this study. ThereforIeB, global spread could not be ascertained because clones obtained in thiNs stu d Ly could not be compared with international clones 2. Some isolates may be left out as it waDs nAot certain that all P. aeruginosa isolated during the period of sample coAllection were given. 3. This study employed Miniprep ImBethod which is not suitable for isolation of large low copy number plasmids 4. T3SS effector toxins w eOre no Ft sought in carbapenem-sensitive strains 5.13 RecommendationsY 1) High level of rIeTsistance demonstrated to most of antibiotics tested in this study could be liSnked with unnecessary use of expanded spectrum antibiotics. Public enliEghtRenment of populace on the consequences of inappropriate use of Vantibiotics and implementation of proper surveillance strategy to encourage NIantibiotic stewardship is required in order to curb the menace of antibiotic U resistance. 2) Imipenem and meropenem are already in clinical use in Nigeria because they are very stable agents against the ESBLs produced by Gram-negative bacteria. However, bacteria are already developing resistance to these carbapenems. Phenotypic detection of MBL-producing strains should be included in the routine screening of bacteria in clinical laboratories 169 3) Since imipenem spontaneously selects mutants during treatment, susceptibility testing of antibiotics before carbapenems prescription is suggested for proper management of antibiotic therapy within healthcare settings 4) The presence of different combinations of efflux pumps, MBLs, deficient oprD, and ampC resistance mechanisms is a cause calls for concern. The use of antibiotics as growth promoters in animals and livestocks should be totally eradicated. 5.14 Contributions to knowledge Y This study has made contributions to the body of knowledge in the area oAf baRcterial drug resistance as follows: i. Prevalence of resistance to carbapenems and other antibioticsR in P. aeruginosa from southwest Nigeria was reported. IB ii. Novel report has been made on the presence of blaV LIM-5 and blaNDM-1 and also on the coexistence of blaNDM-1 with blaVIM-5 in clNinical isolates of P. aeruginosa from southwest Nigeria. DA iii. Class 2 integrons and coexistence wiAth class 1 integrons in clinical isolates of P. aeruginosa from southwest NIiBgeria was reported for the first time in this study. iv. This study for the f iOrst Ftime in Nigeria has quantified resistance gene expression (efflux pYumps, oprD and ampC) in clinical isolates of P. aeruginosa from southwesIt TNigeria as well as reporting the entire four clinically important efflux pumps (MexAB-OprM, MexXY-OprM, MexCD-OprJ and MexEF- OprN) Rin PS. aeruginosa from southwest Nigeria. v. ThiEs study has bridged a knowledge gap on type III effector toxins by IVproviding a novel report of the prevalence of the four type III effector toxins in N P. aeruginosa from southwest Nigeria. Uvi. This study has reported for the first time, typing of P. aeruginosa from southwest Nigeria using repetitive element sequence-based PCR methods (REP-, ERIC- and BOX- PCR). 5.15 Conclusion This study identified highly resistant Pseudomonas aeruginosa strains with 50 out of 63 MBL-containing isolates being extensively drug resistant. Colistin sulphate and polymyxin B were the only antibiotics that were active against most of these resistant 170 strains. This study reports high incidence of transmissible MBL genes, presence of co- occurring blaVIM-5 and blaNDM-1 genes, and co-existence of class 1 and 2 integrons in carbapenem-resistant P. aeruginosa isolates from Nigeria. Also, noticeable is the high expression levels of all the four clinically relevant multidrug efflux pumps, MexAB- OprM, MexCD-OprJ, MexEF-OprN and MexXY-OprM as well as oprD and ampC at mRNA transcription level in P. aeruginosa from Nigeria. This study also observed that there was co-occurrence among these resistance mechanisms. Transformation experiment also indicated that these resistant plasmids carrying YMBL- encoding genes are on potentially mobile elements. The introduction of Ra plasmid carrying MBL-encoding genes into susceptible strains might result in the Aemergence of highly resistant strains. This study has discovered the emergBencRe of extensively resistant P. aeruginosa isolates exhibiting resistance to majoLritIy of antibiotics tested with high prevalence of transmissible MBL-resistance g enes among the isolates thereby posing threat to health care settings. The ouAtcoNme of this research is of great health implication, and as such proper strategy Dhas to be put in place to reduce the spread of antibiotic resistance. A Repetitive element sequence-based PC IRB revealed a great level of genetic diversity among carbapenem-resistant PseuFdomonas aeruginosa strains. ERIC-PCR showed good discriminatory ability in Ocomparison with other two methods (REP- and BOX- PCR) that could not dTiffYerentiate some highly similar but unrelated strains. Since ERIC-PCR is not expIensive and the methodology involved is not difficult, this method may be employed Sin monitoring outbreak of P. aeruginosa infections. No outbreak of carbapenemE-reRsistant P. aeruginosa carrying MBL genes was apparent but there was evidenIcVe of cross-transmission within the hospital settings. This is a serious hazard to heNalth care system and underscores the need for regular surveillance and monitoring Uresistance to carbapenems and associated resistance genes. Moreover, presence of integrons may heighten the ability of these strains to acquire multidrug resistance genes, which has enabled their persistence under the selective pressure of antibiotic and may spread in due course within the hospital setting. Therefore, implementation of active measures to control the spread of these strains in our hospitals is highly imperative. 171 References Abdullahi, S.A., Arzai, A.H., Yusuf, I., Adamu, S.M., Adamu, S., Koki, Y.A., Rabi‟u, M.A. and Abbas, A.M. 2017. Molecular detection of New Delhi metallo beta lactamase 1 (NDM-1) producing bacterial isolates in Kano- Northwestern Nigeria. Annual Research and Review in Biology 14.4: 1-6 Abubakar, S. 2007. Implementing infection control programme in Kano, Northern Nigeria. Presented at the: 8th Congress of the International Federation of Infection Control, 18–27 October 2007, Budapest, Hungary. Adeleke, O.E. and Odelola, H.A. 2000. Beta-lactamase in antibiotic resistaRnceY of clinical strains of Staphylococcus aureus. African Journal of MAedical and Pharmaceutical Scieences 3: 35-39. R Adelowo, O.O., Vollmers, J., Mausezahl, I., Kaster, A-K. and IMBuller, J.A. 2018. Detection of the carbapenemase gene blaVIM-5 Lin members of the Pseudomonas putida group isolated from poNlluted Nigerian wetlands. Scientific Reports 8.1: 15116. A Adwan, G. 2017. Detection of type IIIA seDcretion toxins encoding genes of Pseudomonas aeruginosa isolateBs in the West Bank-Palestine. Journal of Advances in Biology and Biotec hInology 11.3: 1-10. Afifi, M.M., Suelam, I.I.A., SoFliman, M.T.A. and El-Gohary, M.G.S. 2013. Prevalence and antimic rOobial susceptibility pattern of Pseudomonas aeruginosa isolated from envirYonmental and clinical samples in upper Egypt. International Journal of BioIloTgical and Chemical Sciences 7.2: 47-57 Ahmed, O.B. and SAsghar, A.H. 2017. Antibiotic susceptibility pattern of Pseudomonas aerEuginRosa expressing blaGES and blaPER genes in two different hospitals. VAfrican Journal of Biotechnology 16.21: 1197-1202. AjNayi,I T., Allmond, L.R., Sawa, T. and Wiener-Kronish, J.P. 2003. Single-Nucleotide-U Polymorphism mapping of the Pseudomonas aeruginosa type III secretion toxins for development of a diagnostic multiplex PCR system. Journal of Clinical Microbiology 41.8: 3526–3531 Alabi, O.S., Mendonca, N., Adeleke, O.E. and da Silvag, G.J. 2017. Molecular screening of antibiotic-resistant determinants among multidrug-resistant clinical isolates of Proteus mirabilis from SouthWest Nigeria. African Health Science 17.2: 356–365. 172 Al-Haik, W.M., Al-Mahbash, A.A, Al-Mahdi, A.Y., Mohamed, M.M.E. and Al- Haddad, A.M. 2016. Genotypic characteristics of clinical and non-clinical isolates of Pseudomonas aeruginosa: Distribution of different antibiogram profiles and molecular typing. Jordan Journal of Biological Sciences 9.3: 185- 194 Armand-Lefèvre, L., Angebault, C., Barbier, F., Hamelet, E., Defrance, G., Ruppé, E., Bronchard, R., Lepeule, R., Lucet, J.C., El Mniai, A., Wolff, M., Montravers, P., Plésiat, P. and Andremont, A. 2013. Emergence of imipene m- resistant Gram-negative bacilli in intestinal flora of intensive careR paYtients. Antimicrobial Agents Chemotherapy 57: 1488-1495 A Atif, M.L., Bezzaoucha, A., Mesbah, S., Djellato, S., Boubechou, N.R and Bellouni, R. 2006. Evolution of nosocomial infection prevalenIcBe in an Algeria university hospital (2001 to 2005). Médecine et Ma lLadies Infectieuses 36: 423-428 N Bambeke, V.F., Balzi, E. and Tulkens, P.M. A2000. Antibiotic efflux pumps. Biochemical Pharmacology. 60: 457–470D. Barguigua, A., El-Otmani F., Lakbakbi EBl YAaagoubi, F., Talmi, M., Zerouli, K. and Timinouni, M. 2012. First rIeport of a Klebsiella pneumoniae strain coproducing NDM-1, VIMF-1 and OXA-48 carbapenemases isolated in Morocco. All Purpose MOedical Information System 121: 675–677. Bassetti, M., Vena, A., YCroxatto, A., Righi, E. and Guery, B. 2018. How to manage PseudIoTmonas aeruginosa infections. Drugs Context 7: 212-527. Bennett, P.M. 200S4. Genome plasticity: insertion sequence elements, transposons and inteEgroRns, and DNA rearrangement. Methods in Molecular Biology 266: 71–113. BeNnneItt V, P.M. 2008. Plasmid encoded antibiotic resistance: acquisition and transfer of U antibiotic resistance genes in bacteria. British Journal of Pharmacology 153: S347–357. Bielecki, P., Glik, J., Kawecki, M., dos Santos, M. 2008. Towards understanding Pseudomonas aeruginosa burn wound infections by profiling gene expression. Biotechnology Letters 30: 777–790. Bonomo, R.A. and Szabo, D. 2006. Mechanisms of multidrug resistance in Acinetobacter species and Pseudomonas aeruginosa. Clinical and Infectious Diseases 43.1: S49-56 173 Boucher, Y., Labbate, M., Koenig, J.E. and Stokes, H.W. 2007. Integrons: mobilisable platforms that promote genetic diversity in bacteria. Trends in Microbiology 15: 301–309. Brown, P.D. and Izundu, A. 2004. Antibiotic resistance in clinical isolates of Pseudomonas aeruginosa in Jamaica. Revista Panamericana de Salud Publica 16.2: 125–130. Brusch, J.L. 2017. Infective Endocarditis. M.S. Bronze, Ed. “Retrieved August 21, 2018, from https://emedicine.medscape.com” Burstein, D., Satanower, S., Simovitch, M., Belnik, Y., Zehavi, M., YerusRhalmYi, G., Ben-Aroya, S., Pupko, T. and Banin, E. 2015. Novel type AIII effectors in Pseudomonas aeruginosa. mBio 6.2: e00161-15. R Bush, K. and Jacoby, G.A. 2010. Updated functional classificIaBtion of β-lactamases (Short Survey). Antimicrobial Agents Chemotherapy 5L4.3: 969-976 Bush, K., Jacoby, G.A. and Medeiros, A.A. 1995. A fuNnctional classification scheme for β-lactamases and its correlation with mAolecular structure. Antimicrobial Agents Chemotherapy 39: 1211-1233. D Cabot, G., Ocampo-Sosa, A.A., Tubau, BF., AMacia, M.D., Rodriguez, C., Moya, B., Zamorano, L., Suarez, C., Pena , IC., Martinez-Martinez, L., Oliver, A., and the Spanish Network for ResearFch in Infectious Diseases. 2011. Overexpression of AmpC and efflux pOumps in Pseudomonas aeruginosa isolates from bloodstream infectYions: Prevalence and impact on resistance in a Spanish multicenter stuIdTy. Antimicrobial Agents Chemotherapy 55: 1906-1911 Cambray, G., GuSerout, A.M. and Mazel, D. 2010. Integrons. Annual Review of GenEeticRs 44: 141–166. V Carmeli, Y., Troillet, N., Eliopoulos, G.M. and Samore, M.H. 1999. Emergence of NIAntibiotic-resistant Pseudomonas aeruginosa: Comparison of risks associated U with different antipseudomonal agents. Antimicrobial Agents Chemotherapy 43.6: 1379–1382. Castanheira, M., Mills, J.C., Farrell, D.J. and Jones, R.N. 2014. Mutation-driven β- lactam resistance mechanisms among contemporary ceftazidime- nonsusceptible Pseudomonas aeruginosa isolates from U.S. hospitals. Antimicrobial Agents Chemotherapy 58.11: 6844-6850. Centre for Disease Prevention and Control (CDC), 2012. Surgical Site Infection (SSI) “Retrieved August 31, 2018, from https://www.cdc.gov/hai/ssi/ssi.html” 174 Centre for Disease Prevention and Control (CDC), 2014. Pseudomonas aeruginosa in healthcare settings. “Retrieved on March 27, 2018, from https://www.cdc.gov/hai/organisms/pseudomonas.html Centre for Disease Prevention and Control (CDC), 2017. Catheter-associated Urinary Tract Infections (CAUTI). “Retrieved August 31, 2018, from https://www.cdc.gov/hai/ca_uti/uti.html” Chang, A.Y., Chau, V.W.Y., Landas, J.A. and Pang, Y. 2017. Preparation of calcium competent Escherichia coli and heat-shock transformation. The Journal of Experimental Microbiology and Immunology 1: 22-25 RY Chastre, J. and Fagon, J-Y. 2002. Ventilator-associated pneumonia. AmeArican Journal of Respiratory and Critical Care Medicine 165: 867–903 R Chika, E., Charles, E., Ifeanyichukwu, I., Okeh, I.D., Malachy, IUB., Chika, E., Carissa, D. and Michael, A. 2017. Molecular identification o fL MBL genes blaIMP-1 and blaVIM-1 in Escherichia coli strains isolated froNm abattoir by multiplex PCR technique. Research Journal of MicrobiologyA 12: 266-273. Cholley, P., Ka, R., Guyeux, C., ThouverezA, MD., Guessennd, N., Ghebremedhin, B., Frank, T., Bertrand, X. and Hocquet, D. 2014. Population structure of clinical Pseudomonas aeruginosa from wIeBst and central African countries. PLoS ONE 9.9: e107008. F Clinical and Laboratory Stand aOrds Institute (CLSI), 2012. Performance standards for antimicrobial susceYptibility testing; twenty-second informational supplement. CLSI documenItT M100-S22. Clinical and LaboSratory Standards Institute (CLSI), 2015. Performance standards for antiEmicRrobial susceptibility testing; twenty-fifth informational supplement. VCLSI document M100-S25. CoNburIn, B., Sekirov, I. and Finlay, B.B. 2007. Type III Secretion Systems and Disease. U Clinical Microbiology Review 20.4: 535–549. Codjoe, S. and Donkor, S. 2018. Carbapenem Resistance: A Review. Medical Sciences 6.1: 1-28 Curran, B., Jonas, D., Grundmann, H., Pitt, T and Dowson, C.G. 2004. Development of a multilocus sequence typing scheme for the opportunistic pathogen Pseudomonas aeruginosa. Journal of Clinical Microbiology 42.12: 5644-5649 Davies, J.C. and Bilton, D. 2009. Bugs, biofilms, and resistance in cystic fibrosis. Respiratory Care 54: 628–640 175 De Vos, D., Lim, A. J., De Vos, P., Sarniguet, A., Kersters, K. and Cornelis, P. 1992. Detection of the outer membrane lipoprotein I and its gene in fluorescent and non-fluorescent pseudomonads: implications for taxonomy and diagnosis. Journal of General Microbiology 139: 2215–2223. Dean, R.A., Cox, J.H., Bellac, C.L., Doucet, A., Starr, A.E. and Overall, C.M. 2008. Macrophage-specific metalloelastase (MMP-12) truncates and inactivates ELR CXC chemokines and generates CCL2, -7, -8, and -13 antagonists: potent ial role of the macrophage in terminating polymorphonuclear leukocyte influx. Blood 112: 3455–3464. RY Deplano, A., Denis, O. and Poirel, L. 2005. Molecular characterisation oAf an epidemic clone of Pan-antibiotic-resistant Pseudomonas aerugRinosa. Clinical Microbiology 43: 1198-1204. IB Deshpande, P., Rodrigues, C., Shetty, A., Kapadia, F., Hedg eL, A. and Soman, R. 2010. New Delhi Metallo-beta lactamase (NDM-1) inN Enterobacteriaceae: treatment options with carbapenems compromised. AJournal of the Association of Physians in India 58: 147-149. D Dia, N.M., Ka R, Dieng C, Diagne R, DiaB MLA, Fortes L, Diop, B.M., Sow, A.I. and Sow, P.S. 2008. Prevalence of n oIsocomial infections in a university hospital (Dakar, Senegal). Médecine Fet Maladies Infectieuses 38: 270-274 Domingues, S., da Silva, G .OJ. and Nielsen, K.M. 2012. Integrons: Vehicles and pathways for horizYontal dissemination in bacteria. Mobile Genetic Elements 2.5: 211–223. IT Donkor, E.S., NewSman, M.J. and Yeboah-Manu, D. 2012. Epidemiological aspects of nEon-Rhuman antibiotic usage and resistance: Implications for the control of Vantibiotic resistance in Ghana. Journal of Tropical Medicine and NI International Health 17: 462–468. UDouard, G, Praud, K., Cloeckaert, A. and Doublet, B. 2010. The Salmonella genomic island 1 is specifically mobilized in trans by the IncA/C multidrug resistance plasmid family. PLoS One 5: e15302. Doublet, B., Praud, K., Weill, F.X. and Cloeckaert, A. 2009. Association of IS26- composite transposons and complex In4-type integrons generates novel multidrug resistance loci in Salmonella genomic island. Journal of Antimicrobial Chemotherapy 63: 282–289. 176 Dumas, L., van Delden, C., Perron, K. and Köhler, T. 2006. Analysis of antibiotic resistance gene expression in Pseudomonas aeruginosa by quantitative real- time-PCR. FEMS Microbiology Letter 254: 217–225. El Amin, N., Giske, C. G., Jalal, S., Keijser, B., Kronvall, G. and Wretlind, B. 2005. Carbapenem resistance mechanisms in Pseudomonas aeruginosa: alterations of porin OprD and efflux proteins do not fully explain resistance patterns observed in clinical isolates. All Purpose Medical Information System 113: 187-196 Elkhatib, W.F., Haynes, V.L. and Noreddin, A.M. 2008. Unexpected indRuctYion of resistance Pseudomonas aeruginosa biofilm to fluoroquinolones Aby diltiazem: a new perspective of microbiological drug-drug interacRtion. Journal of Infection and Public Health 1.2: 105-112. IB Ellington, M. J., Kistler, J., Livermore, D. M. and Woodford ,L N. 2007. Multiplex PCR for rapid detection of genes encoding acquNired metallo-beta-lactamases. Journal of Antimicrobial Chemotherapy 59: 3A21–322. Erdem, I., Hakan, T., Ceran, N., Metin, F., AkDcay, S.S., Kucukercan. M., Berkman, M.Z. and Goktas, P. 2008. ClinicaBl feAatures, laboratory data, management and the risk factors that affect thIe mortality in patients with postoperative meningitis. Neurology IndiaF 56.4: 433-437. Etebu, E. and Arikekpar, I. 20 1O6. Antibiotics: Classification and mechanisms of action with emphasis on Ymolecular perspectives. International Journal of Applied Microbiology IanTd Biotechnological Research 4: 90-101 Eyo, A-O., IbeneSme, E.O., Thumamo, B.D.P. and Asuquo, A.E. 2015. Antibiotic resiEstanRce profiles of clinical and environmental isolates of Pseudomonas Vaeruginosa in Calabar, Nigeria. Journal of Pharmaceutical and Biological NISciences 10: 9-15. UFalkinham, J.O., Hilborn, E.D., Arduino, M.J., Pruden, A. and Edwards, M.A. 2015. Epidemiology and ecology of opportunistic premise plumbing pathogens: Legionella pneumophila, Mycobacterium avium, and Pseudomonas aeruginosa. Environmental Health Perspectives 123.8: 749-758. Faridi, F. and Javidpour, S. 2015. REP-PCR Typing, Antibiogram Pattern and Distribution of clinical isolates of Pseudomonas aeruginosa in a teaching hospital in south of Iran. Journal of Molecular Medicine 1.1: 39-47 177 Fazeli, H., Akbari, R., Moghim, S., Narimani, T., Arabestani M.R. and Ghoddousi, A.R. 2012. Pseudomonas aeruginosa infections in patients, hospital means, and personnel's specimens. Journal of Research in Medical Sciences 17.4: 332–337 Feltman, H., Schulert, G., Khan, S., Juan, M., Peterson, L. and Hauser, A.R. 2001. Prevalence of type III secretion genes in clinical and environmental isolates of Pseudomonas aeruginosa. Microbiology 147.10: 2659-2669. Freschi, L., Jeukens, J., Kukavica-Ibrulj, I., Boyle, B., Dupont, M.J., Laroche, J. et al. 2015. Clinical utilization of genomics data produced by the international Pseudomonas aeruginosa consortium. FrontiersR Y in Microbiology 6: 1036. A Galetti, B., Andrade, L.N., Chandler, M., de Mello Varani, A. and CRosta Darini, A.C. 2016. New small plasmid harboring blaKPC-2 in PseuIdoBmonas aeruginosa. Antimicrobial Agents Chemotherapy 60.5: 3211–321 4L. Galle, M., Carpentier, I. and Beyaert, R. 2012. StructurNe and function of the type III secretion system of Pseudomonas aeruginoAsa. Current Protein and Peptide Science 13.8: 831-842. D Garcia, M.M. 2013. Carbapenemases: A reaAl threat. Alliance for the Prudent Use of Antibiotics Newsletter 31: 4–6. IB Gawish, A.A., Mohammed, N.A., EFl-Shennawy, G.A. and Mohammed, H.A. 2013. An investigation of type O3 secretion toxins encoding-genes of Pseudomonas aeruginosa isolatesY in a University Hospital in Egypt. Journal of Microbiology and InfectiousI DTiseases 3.3: 116-122. Gaynes, R. and ESdwards, J.R. 2005. National Nosocomial Infections Surveillance systEemR: Overview of nosocomial Infections caused by Gram-negative bacilli. VClinical and Infectious Diseases. 41: 848-854 GnNiadIek, T.J., Carroll, K.C. and Simner, P.J. 2016. Carbapenem-resistant non-glucose-U fermenting Gram-negative bacilli: the missing piece to the puzzle. Jounal of Clinical Microbiology 54: 1700–1710 Goldberg, J.B. 2010. Why is Pseudomonas aeruginosa a pathogen? F1000 Biology Reports 2: 29 Gomaa, N.I.M., Metwally, L., Attallah, M.F., Younis, S. and Kamel, N. 2013. Measurement of efflux pumps expression by relative quantitative real-time polymerase chain reaction in multi-drug resistant Pseudomonas aeruginosa 178 isolated from Suez Canal University Hospital. Journal of Molecular Diagnosis and Vaccines 11: 1-8 Gosling, R., Mbatia, R., Savage, A., Mulligan, J. A. and Reburn, H. 2003. Prevalence of hospital-acquired infections in a tertiary referral hospital in northern Tanzania. Annual Tropical Medicine and Parasitology 97: 69-73 Govan, J.R. and Nelson, J.W. 1992. Microbiology of lung infection in cystic fibrosis. British Medical Bulletin 48.4: 912-930. Govind, C.N., Moodley, K., Peer, A.K. Pillay, N., Maske, C., Wallis, C., Viana, R. a nd Chetty, A. 2013. NDM-1 imported from India-first reported caseR in YSouth Africa. South African Medical Journal 103: 476-478. A Grundmann, H., Schneider, C., Hartung, D., Daschner, F.D., RPitt, T.L. 1995. Discriminatory power of three DNA-based typing techniIquBes for Pseudomonas aeruginosa. Journal of Clinical Microbiology 33: 52 8-L534. Gupta, V. 2008. Metallo beta lactamases in PsNeudomonas aeruginosa and Acinetobacter species. Expert Opinion on InvAestigational Drugs 17.2: 131-143 Gutiérrez, O., Juan, C., Cercenado, E., NavarAro, DF., Bouza, E., Coll, P., Pérez, J.L. and Oliver, A. 2007. Molecular epidBemiology and mechanisms of carbapenem resistance in Pseudomonas aeIruginosa isolates from Spanish hospitals. Antimicrobial Agents ChemoFtherapy 51.12: 4329-4335. Güvensen, N.C., Keskin, D. a nOd Sankur, F. 2017. Detection and antibiotic resistance profile of PseudomYonas aeruginosa and prevalence of multiple antibiotics resistant (MAIRT) in different clinical ısolates. European Journal of BiotechnolSogy and Bioscience 5.3: 74-78 Hall, R.M.E andR Collis, C.M. 1995. Mobile gene cassettes and integrons: capture and Vspread of genes by site-specific recombination. Molecular Microbiology 15: NI593–600. UHandzlik, J., Matys, A. and Kieć-Kononowicz, K. 2013. Recent advances in multi-drug resistance (MDR) efflux pump inhibitors of Gram-positive bacteria Staphylococcus aureus Antibiotics 2: 28-45 Hanson, N.D., and Sanders, C.C. 1999. Regulation of inducible AmpC β-lactamase expression among Enterobacteriaceae. Current Pharmaceutical Design 5: 881- 894. Hashem, H., Hanora, A., Abdalla, S., Shaeky, A. and Saad, A. 2017. Dissemination of metallo‐ β‐ lactamase in Pseudomonas aeruginosa isolates in Egypt: mutation 179 in blaVIM‐ 4. Acta Pathologica, Microbiologica, et Immunologica Scandinavica 125: 499-505. Hassuna, N.A., Mohamed, A.H.I., AboEleuoon, S.M., and Rizk, A.H.A. 2015. High Prevalence of Multidrug Resistant Pseudomonas aeruginosa Recovered from Infected Burn Wounds in Children. Archives of Clinical Microbiology 6.4: 1-7 Hauser, A.R. 2009. The type III secretion system of Pseudomonas aeruginosa: infection by injection. Nature Reviews Microbiology 7: 654–665. Hauser, A.R., Cobb, E., Bodi, M., Mariscal, D., Valles, J., Engel, J.N. and ReYllo, J. 2002. Type III protein secretion is associated with poor clinical ouRtcomes in patients with ventilator-associated pneumonia caused by APseudomonas aeruginosa. Critical Care Medicine 30: 521-528 R Havelaar, A.H., Bosman, M. and Borst, J. 1983. Otitis exteIrBna by Pseudomonas aeruginosa associated with whirlpools. Journal of H ygLiene 90: 489-498 Hiett, K.L. and Seal, B.S. 2009. Use of repetitive elemeNnt palindromic PCR (rep-PCR) for the epidemiologic discrimination of foAodborne pathogens. Methods in Molecular Biology 551: 49-58 D Hikida, M., Kawashima, K., Yoshida M. BandA Mitsuhashi S. 1992. Inactivation of new carbapenem antibiotics by deh ydIropeptidase-I from porcine and human renal cortex. Journal of AntimicroFbial Chemotherapy 30: 129–134 Hishinuma, T., Tada, T., Uch idOa, H., Shimojima, M. and Kirikae, T. 2019. A novel VIM-type metallo Ybeta-lactamase variant, VIM-60, with increased hydrolyzing activity againsIt Tfourth generation cephalosporins in Pseudomonas aeruginosa clinical isoSlates in Japan. Antimicrobial Agents and Chemotherapy. 63.6: 1-4. Hoang, C.EQ., RNguyen, H.D., Vu, H.Q., Nguyen, A.T., Pham, B.T., Tran, T.L., VNguyen, H.T.H., Dao, Y.M., Nguyen, T.S.M., Nguyen, D.A., Tran, H.T.T. NIand Phan, L.T. 2019. Emergence of New Delhi metallo-beta-lactamase (NDM) U and Klebsiella pneumoniae carbapenemase (KPC) Production by Escherichia coli and Klebsiella pneumoniae in Southern Vietnam and appropriate methods of detection: A cross-sectional study. BioMed Research International 2019: 1-9 Hocquet, D., Nordmann, P., El Garch, F., Cabanne, L. and Plesial, P. 2006. Involvement of the MexXY-OprM efflux system in emergence of cefepime resistance in clinical strains of Pseudomonas aeruginosa. Antimicrobial Agents Chemotherapy 50: 1347-1351. 180 Hong, S.S., Kim, K., Huh, J.Y., Jung, B., Kang, M.S. and Hong, S.G. 2012. Multiplex PCR for Rapid Detection of Genes Encoding Class A Carbapenemases. Annals of Laboratory Medicine 32: 359-361 Hu, Z. and Zhao, W.H. 2009. Identification of plasmid- and integron-borne blaIMP-1 and blaIMP-10 in clinical isolates of Serratia marcescens. Journal of Medical Microbiology 58: 217–221. Huang, C.R., Lu, C.H., Chuang, Y.C., Tsai, N.W., Chang, C.C., Chen, S.F., Wang, H.C., Chien, C.C. and Chang, W.N. 2007. Adult Pseudomonas aerugino sa meningitis: High incidence of underlying medical and/or postneuRrosuYrgical conditions and high mortality rate. Japan Journal of InfectiousA Diseases 60: 397-399. R Huang, H., Jeanteur, D., Pattus, F. and Hancock, R.E. 1995.I MBembrane topology and site-specific mutagenesis of Pseudomonas aLeruginosa porin OprD. Molecular Microbiology 16: 931–941. N Humphries, R.M., Abott, A.N. and Hindler, AJ.A. 2019. Understanding and addressing CLSI breakpoint revisionsD: a primer for clinical laboratories. Clinical Microbiology 57.6: 1-15 A Igbalajobi, O.A., Oluyege, A.O., Olad eIjBi, A.C. and Babalola, J.A. 2016. Antibiotic resistance pattern of PseudoFmonas aeruginosa isolated from clinical samples in Ekiti State University OTeaching Hospital, Ado-Ekiti, Ekiti state of Nigeria. British MicrobiologYy Research Journal 12.4: 1-6. Ige, O.K., AdesanmiI, TA.A. and Asuzu, M.C. 2011. Hospital-acquired infections in a Nigerian tSertiary health facility: An audit of surveillance reports. Nigerian MedEicaRl Journal 52.4: 239–243. Izadi, BV., Souzani, R., Farahani, A., Mehrabian, S. and Mohajeri, P. 2017. NIFrequency of class 2 integrons in multidrug-resistant Acinetobacter U baumannii isolated from patients in West of Iran. Annals of Tropical Medicine and Public Health 10.1: 104-108 Jabalameli, F., Mirsalehian, A., Khoramian, B., Aligholi, M., Khoramrooz, S.S., Asadollahi, P., Taherikalani, M. and Emaneini, M.. 2012. Evaluation of biofilm production and characterisation of genes encoding type III secretion system among Pseudomonas aeruginosa isolated from burn patients. Burns 38.8: 1192-1197. 181 Jacobs, C., Frère, J.M. and Normark, S. 1997. Cytosolic intermediates for cell wall biosynthesis and degradation control inducible β-lactam resistance in Gram- negative bacteria. Cell 88: 823-832. Jacoby G.A. 2009. AmpC beta-lactamases. Clinical Microbiology Review. 22: 161- 182. Javiya, V.A., Ghatak, S.B., Patel, K.R. and Patel, J.A. 2008. Antibiotic susceptibility patterns of Pseudomonas aeruginosa at a tertiary care hospital in Gujar at, India. Indian Journal Pharmacology 40.5: 230-234. Jesumirhewe, C., Springer, B., Lepuschitz, S., Allerberger, F. and Ruppitsch,R W.Y 2017. Carbapenemase-producing Enterobacteriaceae-isolates from Edo AState, Nigeria Antimicrobial Agents Chemotherapy 61.8: 1-3 R Jones, R.N. 2001. Resistance pattern among nosocomial pathoIgBens: trends over the past few years. Chest 119: S397-404. L Joshi, P.R., Acharya, M., Kakshapati, T., LeungtongkNam, U., Thummeepak, R. and Sitthisak, S. 2017. Co-existence Aof blaOXA-23 and bla NDM-1 genes of Acinetobacter baumannii isolated fAromD Nepal: antimicrobial resistance and clinical significance AntimicrobialB Resistance and Infection Control 6: 21-27 Jovcic, B., Lepsanovic, Z., Suljagic, V .I, Rackov, G., Begovic, J., Topisirovic, L. and Kojic, M. 2011. Emergence Fof NDM-1 metallo-beta-lactamase in Pseudomonas aeruginosa clinical Oisolates from Serbia. Antimicrobial Agents Chemotherapy 55: Y3929–3931. Juan, C., Moyá, B., IPTérez, J.L. and Oliver, A. 2006. Stepwise upregulation of the PseudomonSas aeruginosa chromosomal cephalosporinase conferring high-level β-laEctamR resistance involves three AmpD homologues. Antimicrobial Agents Chemotherapy 50: 1780-1787. JuNhi, TI V., Bibhabati, M., Archana, T., Poonam, L. and Vinita, D. 2009. Pseudomonas U aeruginosa meningitis in post neurosurgical patients. Neurology Asia. 14.2: 95-100. Karthikeyan, K., Thirunarayan, M.A. and Krishnan, P. 2010. Coexistence of blaOXA-23 with blaNDM-1 and armA in clinical isolates of Acinetobacter baumannii from India. Journal of Antimicrobial. Chemotherapy 65.10: 2253-2254. Khan, H.A., Baig, F.K. and Mehboob, R. 2017. Nosocomial infections: Epidemiology, prevention, control and surveillance. Asian Pacific Journal of Tropical Biomedicine 7.5: 478-482 182 Khosravi, A.D., Motahar, M. and Abbasi Montazeri, E. 2017. The frequency of class1 and 2 integrons in Pseudomonas aeruginosa strains isolated from burn patients in a burn center of Ahvaz, Iran. PLoS ONE 12.8: e0183061. Kidd, T.J., Grimwood, K., Ramsay, K.A., Rainey, P.B. and Bell, S.C. 2011. Comparison of three molecular techniques for typing Pseudomonas aeruginosa isolates in sputum samples from patients with cystic fibrosis. Journal of Clinical Microbiology 49.1: 263–268. Koo, S.H. 2015. Overexpression of efflux pump in multiresistant Pseudomon as aeruginosa: How you will discover and treat it? Journal of InfeRctioYn and Chemotherapy 47.2: 142–144. A Ktari, S., Arlet, G., Mnif, B., Gautier, V., Mahjoubi, F., Jmeaa, MR.B., Bouaziz, M., Hammami, A. 2006. Emergence of multidrug-resistant KIlBebsiella pneumoniae isolates producing VIM-4 metallo-beta-lactamas e,L CTX-M-15 extended- spectrum beta-lactamase, and CMY-4 AmpCN b-lactamase in a Tunisian university hospital. Antimicrobial Agents CheAmotherapy 50: 4198–4201. Kwon, D-H., El-Zaatari, F.A.K., Woo, J-S., PAernDg, C-L., Graham, D.Y. and Goo, M.F. 1998. Rep-PCR fragments as biomarkers for differentiating gastroduodenal disease-specific Helicobacter p ylIoBri. Digestive Diseases and Sciences 43: 980- 987. F Lamas-Ferreiro, J.L., Álvarez OOtero, J., González González, L., Novoa Lamazares, L., Arca Blanco, AY., Bermúdez Sanjurjo, J.R., Rodríguez Conde, I., Fernández Soneira, M. aInTd de la Fuente Aguado, J. 2017. Pseudomonas aeruginosa urinary traSct infections in hospitalized patients: Mortality and prognostic factEors.R PLoS One 12.5: e0178178. Lari, AV.R., Honar, H.B. and Alaghehbandan, R. 1998. Pseudomonas infections in NITohid Burn Center, Iran. Burns 24.7: 637-641 ULauretti, L., Riccio, M.L., Mazzariol, A., Cornaglia, G., Amicosante, G., Fontana, R. and Rossolini, G.M. 1999. Cloning and characterisation of blaVIM, a new integron-borne metallo-β–lactamase gene from a Pseudomonas aeruginosa clinical isolate. Antimicrobial Agents Chemotherapy 43: 1584-1590 Lee, N.L., Yuen, K.Y. and Kumana, C.R. 2001. Beta-lactam antibiotic and β-lactamase inhibitor combinations. The Journal of American Medical Association 285.4: 386-388. 183 Liew, S.M., Rajasekaram, G., Puthucheary, S.D. and Chua, K.H. 2018. Detection of VIM-2-, IMP-1- and NDM-1-producing multidrug-resistant Pseudomonas aeruginosa in Malaysia. Journal of Global Antimicrobial Resistance 13: 271- 273. Lim, A. Jr., De Vos, D., Brauns, M., Mossialos, D., Gaballa, A., Qing, D. and Cornelis, P. 1997. Molecular and immunological characterization of OprL, the 18kDa outer membrane peptidoglycan-associated (PAL) lipoprotein of Pseudomonas aeruginosa. Microbiology 143.5: 1709-1716 Lin, T-I., Huang, Y-F., Liu, P-Y., Chou, C-A., Chen, Y-S., Chen, Y-Y., HRsiehY, K-S. and Chen, Y-S. 2016. Pseudomonas aeruginosa infective enAdocarditis in patients who do not use intravenous drugs: Analysis of Rrisk factors and treatment outcomes. Journal of Microbiology Immunology and Infection 49.4: 516-512 LI B Lister, P.D. 2002. Chromosomally-Encoded ResistanceN Mechanisms of Pseudomonas aeruginosa. American Journal of PharmacogAenomics 2.4: 235–243 Lister, P.D., Gardner, V.M. and Sanders, C.CA. 19D99. Clavulanate induces expression of the Pseudomonas aeruginosa ABmpC cephalosporinase at physiologically relevant concentrations and ant aIgonizes the antibacterial activity of ticarcillin. Antimicrobial Agents ChemoFtherapy 43: 882-889. Liu, P.Y-F. and Wu, W-L. 1 9O97. Use of different PCR-based DNA fingerprinting techniques and pulYsed-field gel electrophoresis to investigate the epidemiology of AcinetobactIeTr calcoacetics-Acinetobacter baumannii complex. Diagnostic MicrobioloSgy and Infectious Disease 28: 19-28. Livermore,E DR.M. and Brown. D.F.J. 2001. Detection of β-lactamase-mediated Vresistance. Journal of Antimicrobial Chemotherapy 48: S59-64. LiNvermIore, D.M. 1992. Interplay of impermeability and chromosomal β-lactamase U activity in imipenem-resistant Pseudomonas aeruginosa. Antimicrobial Agents Chemotherapy 36: 2046-2048. Llanes, D., Hocquet, C., Vogne, D., Benali-Baitich, C. and Neuwirth, P. 2004. Clinical strains of Pseudomonas aeruginosa overproducing MexAB–OprM and MexXY efflux pumps simultaneously. Antimicrobial. Agents Chemotherapy 48.5: 1797- 1802 Lolans, K., Queenan, A.M., Bush, K., Sahud, A. and Quinn, J.P. 2005. First Nosocomial Outbreak of Pseudomonas aeruginosa producing an integron- 184 borne metallo-ß-lactamase (VIM-2) in the United States. Antimicrobial Agents Chemotherapy 49.8: 3538-3540. Lucena, A., Dalla Costa, L.M., Nogueira, K.S., Matos, A.P., Gales, A.C., Paganini, M.C., Castro, M.E. and Raboni, S.M. 2014. Nosocomial infections with metallo-beta-lactamase-producing Pseudomonas aeruginosa: Molecular epidemiology, risk factors, clinical features and outcomes. Journal of Hospital Infection 87: 234-240. Mackie, T.J. and MacCartney, J.E. 1996. Practical Medical Microbiology. J.G. Coll ee, J.P. Duguid, A.G. Fraser and B.P. Marmion, Eds. Churchill LivingstoRne. Y Magiorakos, A.-P., Srinivasan, A., Carey, R.B., Carmeli, Y., Falagas, MA.E., Giske, C.G., Harbarth, S., Hindler, J.F., Kahlmeter, G., OlssoRn-Liljequist, B., Paterson, D.L., Rice, L.B., Stelling, J., Struelens, M.J., VIaBtopoulos, A., Weber, J.T. and Monnet, D.L. 2012. Multidrug-resistant, ext eLnsively drug-resistant and pandrug-resistant bacteria: an international experNt proposal for interim standard definitions for acquired resistance. ClinicalA Microbiology and Infection 18: 268–281 D Maiden, M.C. 2006. Multilocus sequenceA typing of bacteria. Annual Review of Microbiology 60: 561–588. IB Manenzhe, R., Zar, H.J., NicolFl, M.P. and Kaba, M. 2015. The spread of carbapenemase-produ ciOng bacteria in Africa: a systematic review. Journal of Antimicrobial CheYmotherapy 70: 23-40 Marshall, J.C., ChriIsTtou, N.V. and Meakins, J.L. 1993. The gastrointestinal tract. The “undraSined abscess” of multiple organ failure. Annals of Surgery 218: 111- 119 R MayhalVl, CE.G. 1996. Nosocomial burn wound infection. Hospital epidemiology and NI infection control. G.C. Mayhall Ed. William and Wilkins Co.: Baltimore, MD, U USA. 225-236. Mellouk, F.Z., Bakour, S., Meradji, S., Al-Bayssari, C., Bentakouk, M.C., Zouyed, F., Djahoudi, A., Boutefnouchet, N. and Rolain, J.M. 2016. First detection of VIM-4-producing Pseudomonas aeruginosa and OXA-48-producing Klebsiella pneumoniae in Northeastern (Annaba, Skikda). Microbial Drug Resistance 23: 335-344. Meradji, S., Barguigua, A., Zerouali, K., Mazouz, D., Chettibi, H., Elmdaghri, N. and Timinouni, M. 2015. Epidemiology of carbapenem non-susceptible 185 Pseudomonas aeruginosa isolates in Eastern Algeria. Antimicrobial Resistance and Infection Control 4: 27-34. Merradi, M., Kassah-Laouar, A., Ayachi, A., Heleili, N., Menasria, T., Hocquet, D., Cholley, P. and Sauget, M. 2019. Occurrence of VIM-4 metallo-β-lactamase- producing Pseudomonas aeruginosa in an Algerian hospital. Journal of Infection in Developing Countries 13: 284-290. Mesaros, N., Nordmann, P., Plésiat, P., Roussel‐ Delvallez, M., Van Eldere, J., Glupczynski, Y., Van Laethem, Y., Jacobs, F., Lebecque, P., Malfroot, A., Tulkens, P.M. and Van Bambeke, F. 2007. Pseudomonas aeRrugYinosa: resistance and therapeutic options at the turn of the new millennAium. Clinical Microbiology and Infection 13.6: 560-578 R Micek, S.T., Lloyd, A.E., Ritchie, D.J., Reichley, R.M., Fraser,I VB.J. and Kollef, M.H. 2005. Pseudomonas aeruginosa Bloodstream In fLection: Importance of Appropriate Initial Antimicrobial TreatmNent. Antimicrobial Agents Chemotherapy 49.4: 1306–1311. A Mohamed, N. and Raafat, D. 2011. PhenotypicD and genotypic detection of metallo- beta-lactamases in imipenem-resisBtantA Acinetobacter baumannii isolated from a tertiary hospital in Alexandria , IEgypt. Research Journal of Microbiology 6: 750-760. F Mohamed, S.E.R., Alobied, AO., Saeed, M.I. and Hussien, W.M. 2018. New Delhi Metallo-β-lactamaYse (NDM)-mediated Carbapenem-resistant Pseudomonas aeruginosa cliInTical isolates in Sudan. South Asian Journal of Research MicrobioloSgy 1.3: 1-5 MohammeEd, YR., Zailani, S.B. and Onipede, A.O. 2015. Characterisation of KPC, VNDM and VIM type carbapenem resistance Enterobacteriaceae from NINortheastern, Nigeria. Journal of Biosciences and Medicine 3: 100-107. UMohanam, L. and Menon, T. 2017. Coexistence of metallo-beta-lactamase-encoding genes in Pseudomonas aeruginosa. Indian Journal of Medical Research 146: S46-52 Moraes, T.F., Spreter, T. and Strynadka, N.C. 2008. Piecing together the type III injectisome of bacterial pathogens. Current Opinion in Structural Biology 18: 258–266. 186 Morgulis, A., Coulouris, G., Raytselis, Y., Madden, T.L., Agarwala, R. and Schäffer, A.A. 2008. Database indexing for production MegaBLAST searches. Bioinformatics 24: 1757-1764. Moya, B., Juan, C., Alberti, S., Perez, J.L. and Oliver, A. 2008. Benefit of Having Multiple ampD genes for acquiring -lactam resistance without losing fitness and virulence in Pseudomonas aeruginosa. Antimicrobial Agents Chemotherapy 52.10: 3694-3700 Mulu, W., Kibru G., Beyene, G. and Damtie, M. 2012. Postoperative Nosocom ial Infections and Antimicrobial Resistance Pattern of Bacteria IsolatRes aYmong Patients Admitted at Felege Hiwot Referral Hospital, BahirdAar, Ethiopia. Ethiopian Journal Health Science 22.1: 7–18. R Murphy, T.A., Simm, A.M., Toleman, M.A., Jones, R.N. anIdB Walsh, T.R. 2003. Biochemical characterization of the acquired metallo -Lβ-lactamase SPM-1 from Pseudomonas aeruginosa. Antimicrobial Agents NChemotherapy 47: 582-587 Mushi, M.F., Mshana, S.E., Imirzalioglu, C. and BAwanga, F. 2014. Carbapenemase genes among multidrug resistant GramA-neDgative clinical isolates from a tertiary hospital in Mwanza, Tanzania. BBiomedical Research International 2014: 303104. I Mushtaq, S., Ge, Y. and LivFermore, D.M. 2004. Comparative activities of doripenem versus Oisolates, mutants, and transconjugants of EnterobacteriacYeae and Acinetobacter species with characterised beta- lactamases. AnItiTmicrobial Agents Chemotherapy 48: 1313-1319. Nabarro, L.E., ShaSnkar, C., Pragasam, A., Mathew, G., Jeyaseelan, V., Veeraraghavan, B. aEndR Verghese, V. 2017. Clinical and bacterial risk factors for mortality in Vchildren with carbapenem-resistant Enterobacteriaceae bloodstream infections NI in India. The Paediatric Infectious Disease Journal 36.6: e161 – e166. UNejad, S.B., Allegranzi, B., Syed, S.B., Ellis, B. and Pittet, D. 2011. Health-care-associated infection in Africa: a systematic review. Bullettin of World Health Organisation 89: 757-765 Nicas, T.I., Bradley, J., Lochner, J.E. and Iglewski, B.H. 1985. The role of exoenzyme S in infections with Pseudomonas aeruginosa. Journal of Infectious Diseases 15.2: 716-721. 187 Nicasio, A.M., Kuti, J.L. and Nicolau, D.P. 2008. The current state of multidrug- resistant Gram-negative bacilli in North America. Pharmacotherapy 28: 235– 249 Nijssen, S., Florijn, A., Top, J., Willems, R., Fluit, A. and Bonten, M. 2005. Unnoticed spread of integron-carrying Enterobacteriaceae in Intensive Care Units. Clinical and Infectious Diseases 41.1: 1-9 Nordmann, P., Naas, T. and Poirel, L. 2011. Global spread of carbapenemase- producing Enterobacteriaceae. Emerging Infectious Diseases 17: 1791–1798. Obaseiki-Ebor, E.E. 1988. Trimethoprim/sulphaethoxazole resistance in EsRcheYrichia. coli and Klebsiella species urinary isolates. African Journal of MeAdical Science 17: 175-180 R Odetoyin, B.W., Labar, A.S., Lamikanra, A., Aboderin, A.O. aInBd Okeke, I.N. 2018. Correction: Classes 1 and 2 integrons in faecal Esch erLichia coli strains isolated from mother-child pairs in Nigeria. PLoS ONE 13N.5: e0197202. Odumosu, B.T., Adeniyi, B.A., Dada-Adegbola, H. Aand Chandra, R. 2012. Multidrug resistant pseudomonas aeruginosa frDom southwest Nigeria hospitals. International Journal of PharmacBeutiAcal Sciences Review and Research 15.2: 11-15 I Odumosu, B.T., Adeniyi, B.A. andF Chandra, R. 2013. Analysis of integrons and associated gene caOssettes in clinical isolates of multidrug resistant PseudomoYnas aeruginosa from Southwest Nigeria. Annals of Clinical Microbiology IanTd Antimicrobials 12: 29-35 Oduyebo, O., OguSnsola, F.T. and Odugbemi, T. 1997. Prevalence of multi-resistant straEins Rof Pseudomonas aeruginosa isolated at the Lagos University Teaching VHospital from 1994 – 1996. Nigerian Quarterly Journal of Medicine 7: 373– NI376. UOgunleye, A.O., Ajuwape, A.T.P. and Carlson, S.A. 2016. Emergence of New Delhi metalo-β- lactamase (NDM-1) – producing multidrug resistant Gram negative bacteria from poultry in Nigeria. Bulletin of Animal Health Production in Africa 64.1: 139-148 Oladipo, E.K., Oloke, J.K., Omomowo, I.O., Oyeniran, A.O., Awoyelu, E.H. and Ogundele, S.O. 2015. Incidence of cephalosporin resistance among clinical isolates of Pseudomonas aeruginosa in Ibadan, SouthWestern Nigeria. International Journal of Medicine and Biomedical Research 43: 135-141 188 Olayinka, A.T., Olayinka, B.O. and Onile, B.A. 2009. Antibiotic susceptibility and plasmid pattern of Pseudomonas aeruginosa from the surgical unit of a University Teaching Hospital in north central Nigeria. International Journal of Medicine 1.3: 79-83 Olive, D.M. and Bean, P. 1999. Principles and Applications of Methods for DNA- Based Typing of Microbial Organisms. Journal of Clinical Microbiology 37.6: 1661-1669. Palavutitotai, N., Jitmuang, A., Tongsai, S., Kiratisin, P. and Angkasekwinai, N. 20 18. Epidemiology and risk factors of extensively drug-resistant PseRudoYmonas aeruginosa infections. PLoS ONE 13.2: e0193431. A Papp-Wallace, K.M., Endimiani, A., Taracila, M.A. and BonoRmo, R.A. 2011. Carbapenems: Past, Present, and Future. Antimicrobial IABgents Chemotherapy 55.11: 4943–4960. L Peleg, A.Y., Franklin, C., Walters, L.J., Bell, J.M. and SNpelman, D.W. 2006. OXA-58 and IMP-4 carbapenem-hydrolyzing beta-Alactamases in an Acinetobacter jejunii blood culture isolate froAm DAustralia. Antimicrobial. Agents Chemotherapy 50.1: 399-400 Petrova, M., Gorlenko, Z. and Mindlin , ISB. 2011. Tn5045, a novel integron-containing antibiotic and chromate reFsistance transposon isolated from a permafrost bacterium. Research in MOicrobiology 162: 337–345. Pinna, A., Usai, D., SechYi, L.A., Zanetti, S., Jesudasan, N.C.A., Thomas, P.A. and Kaliamurthy, IJ.T 2009. An outbreak of post-cataract surgery endophthalmitis caused by PSseudomonas aeruginosa. Ophthalmology 116.12: 2321-2326 Pitout, J.D.E, GRregson, D.B., Poirel, L., McClure, J.A., Le, P. and Church, D.L. 2005. VDetection of Pseudomonas aeruginosa producing metallo-beta-lactamases in a NI large centralized laboratory. Journal of Clinical Microbiology 43: 3129–3135. UPiyakul, C., Tiyawitsutsri, R. and Boonbumrung, K. 2012. Emergence of metallo-β-lactamase IMP-14 and VIM-2 in Pseudomonas aeruginosa clinical isolates from a tertiary-level hospital in Thailand. Epidemiology and Infection 140.3: 539-541. Poirel L, Revathi G, Bernabeu S and Nordmann, P. 2011. Detection of NDM-1- producing Klebsiella pneumoniae in Kenya. Antimicrobial Agents Chemotherapy 55: 934–936. 189 Poirel, L, Patron, A. and Nordmann, P. 2012. OXA-48-like carbapenemases: the phantom menace. Journal of Antimicrobial Chemotherapy 67: 1597–1606 Poirel, L. and Nordmann, P. 2006. Carbapenem resistance in Acinetobacter baumannii: mechanisms and epidemiology. Clinical Microbiology and Infection 12: 826– 837 Poirel, L., Carrër, A., Pitout, J.D. and Nordmann, P. 2009. Integron mobilisation unit as a source of mobility of antibiotic resistance genes. Antimicrobial Agents Chemotherapy 53: 2492–2498. Poirel, L., Naas, T., Nicolas, D., Collet, L., Bellais, S., Cavallo, J.D. and NorRdmaYnn, P. 2000. Characterisation of VIM-2, a carbapenem hydrolyzinAg metallo-β- lactamase and its plasmid- and integron-borne gene fromR a Pseudomonas aeruginosa clinical isolate in France. Antimicrobial AgeInBts Chemotherapy 44: 891-897. L Poirel, L., Weldhagen, G.F., De Champs, C. anNd Nordmann, P. 2002. A nosocomial outbreak of Pseudomonas aerAuginosa isolates expressing the extended-spectrum β-lactamase GAES-D2 in South Africa. Journal of Antimicrobial Chemotherapy 49: 561-565. Poirel, L., Weldhagen, G.F., Naa sI, BT., De Champs, C., Dove, M.G. and Nordmann, P. 2001. GESF-2, a class A β-lactamase from Pseudomonas aeruginosa with incr eaOsed hydrolysis of imipenem. Antimicrobial Agents Chemotherapy 45: Y2598-2603. Pollini, S., Maradei, SI.T, Pecile, P., Olivo, G., Luzzaro, F. and Docquierr, J-D. 2013. FIM-1, a nSew acquired metallo-β-lactamase from a Pseudomonas aeruginosa clinEicalR isolate from Italy. Antimicrobial Agents Chemotherapy 57: 410-416 PoulakoVu, G., Bassetti, M., Righi, E. and Dimopoulos, G. 2014. Current and future NI treatment options for infections caused by multidrug-resistant Gram-negative U pathogens. Future Microbiology 9.9: 1053-1069. Pourakbari, B., Yaslianifard, S., Yaslianifard, S., Mahmoudi, S., Keshavarz-valian, S. and Mamishi, S. 2016. Evaluation of efflux pumps gene expression in resistant Pseudomonas aeruginosa isolates in an Iranian referral hospital. Iran Journal of Microbiology 8.4: 249-256 Quale, J., Bratu, S., Gupta, J. and Landman, D. 2006. Interplay of efflux system, ampC, and oprD expression in carbapenem resistance of Pseudomonas 190 aeruginosa clinical isolates. Antimicrobial Agents Chemotherapy 50: 1633– 1641. Queenan, A.M. and Bush, K. 2007. Carbapenemases: The versatile β-lactamases. Clinical Microbiology Review 20: 440-458. Reinhart, A.A. and Oglesby-Sherrouse, A.G. 2016. Regulation of Pseudomonas aeruginosa virulence by distinct iron sources. Genes 7.12: E126. Richards, M.J., Edwards, J.R., Culver, D.H. and Gaynes, R.P. 1999. Nosocomial infections in medical intensive care units in the United States. Natio nal Nosocomial Infections Surveillance System. Critical Care Medicine R27.5Y: 887- 892 A Rodriguez-Martinez, J.M., Poirel, L. and Nordmann, P. 2009. ExRtended-spectrum cephalosporinases in Pseudomonas aeruginosa. AInBtimicrobial Agents Chemotherapy 53.5: 1766–1771 L Rodríguez-Pardo, D., Pigrau, C., Lora-Tamayo, J., SNoriano, A., del Toro, M.D., Cobo, J., et al. 2014. Gram-negative prosthAetic joint infection: outcome of a debridement, antibiotics and implantA retDention approach. A large multicentre study. Clinical Microbiology and IBnfection 20: O911–999. Roy-Burman, A., Savel, R.H., Racine, SI., Swanson, B.L., Revadigar, N.S., Fujimoto, J., Sawa, T., Frank, D.W. Fand Wiener-Kronish, J.P. 2001. Type III protein secretion is assoc iaOted with death in lower respiratory and systemic PseudomoYnas aeruginosa infections. Journal of Infectious Diseases 183: I1T767-1774. Russell, A.D. 200S4. Pharmaceutical Microbiology. Types of antibiotics and synthetic antiEmicRrobial agents. S.P. Denyer, N.A. Hodges and S.P. German. Eds. Blackwell Science, UK. 152-186. SaNdikoIt V, R.T., Blackwell, T.S., Christman, J.W. and Prince, A.S. 2005. Pathogen–host U interactions in Pseudomonas aeruginosa pneumonia. American Journal of Respiratory and Critical Care Medicine 171.11: 1209-1223. Sato, H., Frank, D.W., Hillard, C.J., Feix, J.B., Pankhaniya, R.R., Moriyama, K., Finck-Barbancon, V., Buchaklian, A., Lei, M., Long, R.M., Wiener-Kronish, J. and Sawa, T. 2003. The mechanism of action of the Pseudomonas aeruginosa- encoded type III cytotoxin, ExoU. European Molecular Biology Organisation Journal 22: 2959-2969. 191 Schmidtke, A. J. and Hanson, N.D. 2006. Model system to evaluate the effect of ampD mutations on AmpC-mediated β-lactam resistance. Antimicrobial Agents Chemotherapy 50: 2030-2037. Sekiguchi, J., Morita K., Kitao, T., Watanabe, N., Okazaki, M., M., Miyoshi-Akiyama, T., Kanamori, M. and Kirikae, T. 2008. KHM-1, a novel plasmid-mediated metallo-β-lactamase from a Citrobacter freundii clinical isolate. Antimicrobial Agents Chemotherapy 52.11: 4194-4197. Senda, K., Arakawa, Y., Nakashima, K., Ito, H., Ichiyama, S. Shimokata, K., Kato, N. and Ohta, M. 1996. Multifocal outbreaks of metallo-β-lactamase-RprodYucing Pseudomonas aeruginosa resistant to broad-spectrum beta-lactaAms, including carbapenems. Antimicrobial Agents Chemotherapy 40.2: 349-3R53. Sharples, G.J. and Lloyd, R.G. 1990. A novel repeated DNA seIqBuence located in the intergenic regions of bacterial chromosomes. Nucleic LAcids Research 18: 6503- 6508. N Shaver, C.M. and Hauser, A.R. 2004. RelativAe contributions of Pseudomonas aeruginosa ExoU, ExoS, and ExoT Ato Dvirulence in the lung. Infection and Immunity 72: 6969-6977. Soong, G., Parker, D., Magargee, M .I Band Prince, A. 2008. The type III toxins of Pseudomonas aeruginosFa disrupt epithelial barrier function. Journal of Bacteriology 190.8: 28 1O4–2821. Spencer. W.H. 1953. PsYeudomonas aeruginosa infections of the eye. California Medicine 79.6I: 4T38–443. Spencker, F.B., HSaupt, S., Claros, M.C., Walter, S., Lietz, T., Schille, R. and Rodloff, A.CE. 2R000. Epidemiologic characterisation of Pseudomonas aeruginosa in Vpatients with cystic fibrosis. Clinical Microbiology and Infection 6.11: 600– NI607. USyrmis, M.W., O‟Caroll, M.R., Sloots, T.P., Coulter, C., Wainwright, C.E., Bell, S.C. and Nissen, M.D. 2004. Rapid genotyping of Pseudomonas aeruginosa isolates harboured by adult and paediatric patients with cystic fibrosis using repetitive- element-based PCR assays. Journal of Medical Microbiology 53.11: 1089- 1096. Taneja, N., Chatterjee, S.S., Singh, M., Singh, S., and Sharma, M. 2010. Paediatric urinary tract infections in a tertiary care center from north India. Indian Journal of Medical Science 131: 101-105. 192 Timsit, J.F. and Laupland, K.B. 2012. Update on bloodstream infections in ICUs. Current Opinion in Critical Care 18.5: 479-486. Toleman, M.A., Biedenbach, D., Bennett, D.M.C., Jones, R.N. and Walsh, T.R. 2005. Italian metallo-?-lactamases: a national problem? Report from the SENTRY Antimicrobial Surveillance Programme. Journal of Antimicrobial Chemotherapy 55: 61-70. Tomás, M., Doumith, M., Warner, M., Turton, J.F., Beceiro, A., Bou, G., Livermore, D.M. and Woodford, N. 2010. Efflux pumps, OprD Porin, AmpC β-LactaYma se, and multiresistance in Pseudomonas aeruginosa isolates from cystiRc fibrosis patients. Journal of Antimicrobial Chemotherapy 54.5: 2219–2224A Touati, M., Diene, S.M., Dekhil, M., Djahoudi, A., Racherache, A. Rand Rolain, J.M. 2013. Dissemination of a class I integron carrying VIMIB-2 carbapenemase in Pseudomonas aeruginosa clinical isolates from a ho spLital intensive care unit in Annaba, Algeria. Antimicrobial Agents ChemothNerapy 57: 2426-2427. Urwin, R. and Maiden, M.C. 2003. Multilocus seAquence typing: a tool for global epidemiology. Trends in MicrobiologyA 11D: 479–487. Varaiya, A., Kulkarni, N., Kulkarni, M., Bhalekar, P. and Dogra, J. 2008. Incidence of metallo beta lactamase produc inIgB Pseudomonas aeruginosa in ICU patients. Indian Journal of Medical RFesearch 127: 398-402. Versalovic, J., Koeuth, T. a ndO Lupski, R.J. 1991. Distribution of repetitive DNA sequences in eubacYteria and application to fingerprinting of bacterial genomes. Nucleic Acids IRTesearch 9: 6823-6831 Versalovic, J., ScShneider, M., de Bruijn, F.J. and Lupski, J.R. 1994. Genomic fingEerpRrinting of bacteria with repetitive sequence-based polymerase chain reaction. Methods in Cell Biology 5: 25-40. ViNllegIa Vs, M.V., Lolans, K., Correa, A., Kattan, J.N., Lopez, J.A. and Quinn, J.P. 2007. U First identification of Pseudomonas aeruginosa isolates producing a KPC-type carbapenem-hydrolyzing beta-lactamase. Antimicrobial Agents Chemotherapy 51: 1553–1555 Vincent, J., Bihari, D.J., Suter, P.M., Bruining, H.A., White, J., Nicolas-Chanoin, M., Wolff, M., Spencer, R.C. and Hemmer, M. 1995. The prevalence of nosocomial infection in intensive care units in Europe. Journal of American Medical Association 274.8: 639-644 193 Viscol, C. 2016. Bloodstream Infections: The peak of the iceberg. Virulence 7.3: 248– 251. Vitkauskiene, A., Skrodeniene, E., Dambrauskiene, A., Baksyte, G., Macas, A. and Sakalauskas, R. 2011. Characteristics of carbapenem-resistant Pseudomonas aeruginosa strains in patients with ventilator-associated pneumonia in intensive care units. Medicina (Kaunas) 47.12: 652-656 Walsh, C. 2000. Molecular mechanisms that confer antibacterial drug resistance. Nature 406: 775–781. Walsh, T.R., Toleman, M.A., Poirel, L. and Nordmann, P. 2005. Metallo-β-laRctamYases: the quiet before the storm? Clinical Microbiology Review 18.2: 30A6–325. Walsh, T.R., Weeks, J., Livermore, D.M. and Toleman, M.A. 2011. RDissemination of NDM-1 positive bacteria in the New Delhi environment aInBd its implications for human health: an environmental point prevalence Lstudy. Lancet Infectious Disease 11: 355–362. N Wang, C., Cai, P., Chang, D. and Mi, Z. 2006. AA Pseudomonas aeruginosa isolate producing the GES-5 extendedA-speDctrum beta-lactamase. Journal of Antimicrobial Chemotherapy 57: 1261–1262. Watanabe, M., Iyobe, S., Inoue, M. an dI BMitsuhashi, S. 1991. Transferable imipenem resistance in Pseudomonas Faeruginosa. Antimicrobial Agents Chemotherapy 35: 147-151. O White, P.A., McIver C. J., YRawlinson W. D. 2001. Integrons and gene cassettes in the EnterobacteriaIcTeae. Antimicrobial Agents Chemotherapy 45: 2658–2661. White, P.A., McIvSer, C.J., Deng, Y. and Rawlinson, W.D. 2000. Characterisation of twoE nRew gene cassettes, aadA5 and dfrA17. Federation of European VMicrobiology Societies Microbiology Letters 182: 265-269. WNiehlImann, L., Wagner, G., Cramer, N., Siebert, B., Gudowius, P., Morales, G., U Köhler, T., van Delden, C., Weinel, C., Slickers, P. and Tümmler, B. 2007. Population structure of Pseudomonas aeruginosa. Proceedings of the National Academy of Sciences 104.19: 8101-8106 Williamson, R., Collatz, E., and Gutmann, L. 1986. Mechanisms of action of beta- lactam antibiotics and mechanisms of non-enzymatic resistance. La Presse Medicale 15.46: 2282-2289. 194 Wolska, K., Kot, B. and Jakubczak, A. 2012. Phenotypic and genotypic diversity of Pseudomonas aeruginosa strains isolated from hospitals in siedlce (Poland). Brazilian Journal of Microbiology 43.1: 734-738 World Health Organization, 2002. Prevention of hospital-acquired infections: a practical guide. 2nd edition. Retrieved September 6, 2018 from https://www.who.int/csr/resources/publications/whocdscsreph200212.pdf 1-72 Xavier, D.E., Picão, R.C., Girardello, R., Fehlberg, L.C.C. and Gales, A.C. 2010. Efflux pumps expression and its association with porin down-regulation and β- lactamase production among Pseudomonas aeruginosa causing bloRodsYtream infections in Brazil. BMC Microbiology 10: 217-223 A Xie, L., Dou, Y., Zhou, K., Chen, Y., Han, L., Guo, X. and Sun, J. 2R017. Coexistence of blaOXA-48 and truncated blaNDM-1 on different plasmIBids in a Klebsiella pneumoniae isolate in China. Frontiers in Microbiolo gLy 2.8: 133. Yong, D., Choi, Y.S., Roh, K.H., Kim, C.K., Park,N Y.H. and Yum, J.H. 2006. Increasing prevalence and diversity Aof metallo-beta-lactamases in Pseudomonas spp., Acinetobacter spAp. Dand Enterobacteriaceae from Korea. Antimicrobial Agents ChemotherapBy 50: 1884-1886. Yong, D., Toleman, M.A., Bell, J., R itcIhie, B., Pratt, R., Rylee, H. and Walsh, T.R. 2012. Genetic and biochemFical characterisation of an acquired subgroup B3 metallo-beta-lactamase Ogene, blaAIM-1, and its unique genetic context in Pseudomonas aeruYginosa from Australia. Antimicrobial Agents Chemotherapy 56.12: 6154-61I5T9. Yong, D., TolemaSn, M.A., Giske, C.G., Cho, H.S., Sundman, K., Lee, K. and Walsh, T.RE. 20R09. Characterisation of a new metallo-β-lactamase gene, blaNDM-1, and a VNovel erythromycin esterase gene carried on a unique genetic structure NI in Klebsiella pneumoniae sequence type 14 from India. Antimicrobial Agents U Chemotherapy 53.12: 5048-5054. Zhanel, G.G., Wiebe, R., Dilay, L., Thomson, K., Rubinstein, E., Hoban, D.J., Noreddin, A.M. and Karlowsky, J.A. 2007. Comparative review of the carbapenems. Drugs 67: 1027–1052. Zhang, Z., Schwartz, S., Wagner, L. and Miller, W. 2000. A greedy algorithm for aligning DNA sequences. Journal of Computer Biology 7.1: 203-214. 195 Zubair, K.O. and Iregbu, K.C. 2018. Resistance pattern and detection of metallo-beta- lactamase genes in clinical isolates of Pseudomonas aeruginosa in a central Nigeria tertiary hospital. Nigerian Journal of Clinical Practice 21.2: 176-182 Y R A R IB L AN AD IB F O Y IT RS E NI V U 196 LIST OF AWARDS The World Academy of Science (TWAS) Postgraduate Fellowship 2016 University of Ibadan Postgraduate College Teaching and Research Assistantship 2016 RY RA LI B AN D A F I B O ITY S VE R UN I 197 Appendix I Media and antibiotics Pseudomonas cetrimide agar (Oxoid) Weigh 45.3g of powder and dispense in 1000 ml of purified water with addition of 10 O ml glycerol. Steam to soften the contents of the medium totally and sterilize at 121 C for 15 minutes. Formulation gram/litre Magnesium chloride 1.4 Y Gelatin peptone 20.0 R Cetrimide 0.3 RA Potassium sulphate 10.0 IB Agar 13.6 L pH 7.2 ± 0.2 AN Tryptone soy broth (Oxoid) D Weigh 30g of powder and dispense in 10B00 Aml of purified water. Steam to soften the Ocontents of the medium totally and ster ilIize at 121 C for 15 minutes. Formulation F gram/litre Potassium hydrogen phosphat e O 2.5 Pancreatic digest of casTeinI Y 17.0 Sodium chloride 5.0 Glucose R S 2.5 Enzymatic Edigest of soya bean 3.0 pH 7.3I V± 0.2 UNuNtrient agar (Lab M) Weigh 28g of powder and dispense in 1000 ml of purified water. Steam to soften the O contents of the medium totally and sterilize at 121 C for 15 minutes. Formulation gram/litre Beef extract 3.0 Peptone 5.0 Agar No 2 12.0 198 NaCl 8.0 pH 7.3 ± 0.2 Nutrient broth (Lab M) Weigh 13g of powder and dispense in 1000 ml of purified water, stir to mix the O medium completely and sterilize at 121 C for 15 minutes. Formulation gram/litre Peptone 5.0 Beef extract 1.0 Y NaCl 5.0 R Yeast 2.0 A pH 7.4 ± 0.2 RIB Mueller Hinton agar (Oxoid) L Weigh 38g of powder and dispense in 1000 ml of puNrified water. Swir to mix and O steam to soften the contents of the medium tDotalAly and sterilize at 121 C for 15 minutes. Formulation BgAram/litre Casein hydrolysate F I 17.5 Agar 17.0 Beef dehydrated infusion frOom 300.0 Starch TY 1.5 Final pH: 7.3 ± 0.1 SI Luria BertaniR broth (Lab M) Weigh V25gE of the powder and dissolve in 1000 ml of purified water and distribute into test tuI obes and sterilize at 121 C for 15 minutes. FoNrmulation gram/litre UYeast extract 5.0 Tryptone 10.0 Sodium chloride 10.0 Agar technical (Agar no. 3) Oxoid Weigh 12g of the powder and dissolve in 1000 ml of purified water. Steam to soften O the contents of the medium totally and sterilize at 121 C for 15 minutes. 199 Sulphide indole motility (SIM) medium (Oxoid) Weigh 50g of powder and dispense in 1000 ml of purified water. Stir to mix and steam O to soften the contents of the medium totally and sterilize at 121 C for 15 minutes. Formulation gram/litre Ferrous ammonium sulphate 0.2 Tryptone 20.0 Sodium thiosulphate 0.2 Peptone 6.1 Agar 3.5 RY Final pH: 7.3 ± 0.2 A TM Motility indole urea (MIU) base medium (Biomark ) BR Suspend 18g in 950ml purified water. Steam to soften th e LcoIntents of the medium Ototally and dispense in 95ml amount into flasks. Sterilize at 121 C for 15 minutes. o Cool to 50-55 C and aseptically add 5ml sterile 40% AureNa solution per 95 ml medium. Formulation gramD/litre Dextrose 1A.0 Sodium chloride IB5.0 Casein enzymic lysate F 10.0 Agar O 2.0 Phenol red 0.01 Final pH: 6.8 ± 0.2 ITY Motility indolRe orSnithine (MIO) fluid medium (Scharlau Chemie) Weigh 31.5Eg of powder and dispense in 1000 ml of purified water. Swirl to mix and OsteamI tVo soften the contents of the medium totally and sterilize at 121 C for 15 mNinutes. UFormulation gram/litre Gelatin peptone 10.0 Dextrose 1.0 L-ornithine hydrogen chloride 5.0 Yeast extract 3.0 Bromocresol purple 0.02 Agar 2.5 Final pH: 6.6 ± 0.2 200 Koser’s citrate medium (Oxoid CM60) Weigh 5.2g of powder and dispense in 1 litre of distilled water. Swirl to mix and boil O to dissolve the medium completely, then sterilize by autoclaving at 121 C for 15 minutes. Formulation gram/litre Bromothymol blue 0.016 Potassium dihydrogen phosphate 1.0 Sodium ammonium phosphate 1.5 Magnesium sulphate 0.2 RY Sodium citrate 2.5 A pH: 6.8 (approx.) BR Preparation of media for oxidative utilization of sugar testL I One gram of glucose, sucrose or mannitol was added into 100ml of peptone water, followed by the addition of 0.25g of NaCl and 0.025Ag oNf phenol red as indicator. The mixtures was dispensed in 5ml portions intoD test tubes, durham tube was then introduced into each of the tube tubes to trapA the gas which could be produced during the process of oxidation. IB Preparation of 0.5 M EDTA F A 0.5 M EDTA solution was mOade ready by introducing 18.6 g of disodium EDTA into 100 mL of purifiedT wYater and pH was attuned to 8.0 with sodium hydroxide and sterilized. SI Preparation oRf stock of antibiotics suspension 1. IVImi Epenem (Bacqure) NManufacturer: Sun Pharmaceutical Ind. Ltd. Dewas- 455 001, India U Indicated weight = 500 mg imipenem and 500 mg cilastatin Weight of vial + powder = {a}g = 23.706 g Weight of empty vial = {b}g = 22.641 g Weight of powder = {c}g = {a-b}g = 1.065 g Weight of excipient = {c}g - indicated weight = 1.065 g -1.0 g = 0.065 g or 65 mg 201 Therefore, 500 mg of imipenem is contained in 1.065 g of the powder 6.4 mg of imipenem is contained in 1065 x 6.4 = 13.63 mg of powder 500 13.63 mg of powder in 10ml of sterile distilled water = 640 μg/mL stock 2. Ceftazidime (Betazim) NAFDAC No.: A4-1494 Manufacturer: Strides Arcolab Limited, India. For Strides Vital Nigeria Limited Indicated weight = 1000 mg = 1.0 g Weight of vial + powder = {a}g = 14.32 g Weight of empty vial = {b} = 13. 04 g Y Weight of powder = {c} = {a-b} = 1.28 g AR Weight of excipient = {c}g - indicated weight R = 1.28 – 1.00 g IB =0.28 g or 280 mg L Therefore, 1000 mg of ceftazidime is contained in 1.N28 g of the powder = 1.28 g – 1.0 g A = 0.28 g or 280 mg D 1000 mg of ceftazidime is contained in 1280 mAg of the powder 51.2 mg of ceftazidime is contain in 12 8I0 Bx 51.2 = 65.54 mg F 1000 65.54 mg of powder in 10 ml s tOerile distilled water = 5120 μg/mL 3. Ceftriaxone (HafloYne) NAFDAC No.: B4-0331 Manufacturer: NCSPC IH Tebei Huamin Pharmaceutical Co., Ltd. Shijiazhuang, India Indicated weight = 1000 mg = 1.0 g Weight of vEialR + powder = {a}g = 15.85 g WeighItV of empty vial = {b} = 14. 71 g WNeight of powder = {c} = {a-b} = 1.14 g UWeight of excipient = {c}g - indicated weight = 1.14 – 1.00 g =0.14 g or 140 mg Therefore, 1000mg of ceftriaxone is contained in 1.14 g of the powder = 1.14 g – 1.0 g = 0.14 g or 140 mg 1000 mg of ceftriaxone is contained in 1140 mg of the powder 202 5.12 mg of ceftriaxone is contain in 1140 x 5.12 = 5.83 mg 1000 58.3mg of powder in 10 ml sterile distilled water = 5120 μg/mL 4. Cefepime (Fortsporine) NAFDAC No.: A4-6768 Manufacturer: Intracin Pharmaceutical Pvt. Ltd. Gujarat, India Indicated weight = 1000 mg = 1.0 g Weight of vial + powder = {a}g = 31.88 g Weight of empty vial = {b} = 29.85 g Y Weight of powder = {c} = {a-b} = 2.03 g R Weight of excipient = {c}g - indicated weight A = 2.03 – 1.00 g R = 1.03 g or 1030 mg IB Therefore, 1000mg of cefepime is contained in 2.03 g of the pLowder = 2.03 g – 1.0 g N = 1.03 g or 1030 mg A 1000 mg of cefepime is contained in 2030 mgA of Dthe powder 5.12 mg of cefepime is contain in 2030 xI 1B.28 = 2.60 mg 1 000 26.0 mg of powder in 10 ml sterOile dFistilled water = 1280 μg/mL 5. Penicillin G (PhiloYpen ) NAFDAC NO: A4-2115 Manufacturer: ShijazIhaTng Pharma Group, China. For Nwaeze pharmaceutical co. Limited. S Indicated weigRht = 600 mg Weight of vEial + powder = 16.28 g WeighItV of empty vial = 15.68 g WNeight of powder = 16.28 g – 15.68 g U = 0.60 g 203 Appendix II Chemicals, Enzymes, Master mix and DNA isolation kits Molecular biology chemicals Tris base (Phyto Technology Laboratories) Glacial acetic acid (Scharlau) Disodium EDTA (BDH Laboratory Supplies, Poole England) 2-mercaptoethanol Ethanol Y Sodium dodecyl sulphate Lysozyme RA R Calcium chloride IB Enzymes L DNase 1 (Thermo Scientific) N RNase A (Thermo Scientific) A Rsa 1 (Thermo Scientific) AD Master mix IB WizPure™ PCR 2X Master (WizbioFso lutions, Korea South) 2X MyTaq Red Mix (Bioline, LOondon) SYBR Green /ROX qPCRY ma ster mix (Thermo Scientific). DNase/RNase-Free DIisTtilled Water (Invitrogen) ®Ethidium bromideS (CARL ROTH ) Agarose gel (HRydraGene) 100 bp VladdEer, 100 bp plus ladder and 1 kb plus ladder (Thermo Scientific) KNits IUWizPrep™ genomic DNA purification kit WizPrep™ Plasmid DNA Miniprep purification Kit (Wizbiosolutions) PureLink™ Micro-to-Midi Total RNA Purification System (Invitrogen) WizScript™ cDNA Synthesis Kit (Wizbiosolutions, Korea) 10x Tris acetate EDTA Buffer Tris base (48.4 g) Glacial acetic acid (11.4 ml) 204 Disodium EDTA (3.7 g) Tris base, glacial acetic acid and EDTA were dissolved in 800 ml of sterile distilled water and make up to 1 litre AR Y LIB R AN B A D I O F ITY RS E IV N U 205 Appendix III Statistical analysis Fisher‟s exact test for association of integron with MBLs Integron MBL Integron-positive Integron- Total p-value negative MBL-positive 48 15 63 MBL- 3 7 10 0.0064 negative Y Total 51 22 73 AR https://www.socscistatistics.com/tests/fisher/Default2.aspx R Fisher‟s exact test for association between exoU and eLxoSI B T3SS T3SS exoU-positive exoU-negAatiNve Total p-value exoS-positive 2 35 D 37 A exoS-negative 36 0 36 < .0001 Total 38 B35 73 https://www.socscistatistics.comF/te sIts/fisher/Default2.aspx Fisher‟s exact test for a sOsociation of clinical source and T3SS Clinical source T3SS IWToYund Non-wound Total p-value exoU S 28 9 37 exoS R 32 6 38 0.399 TotalE 60 15 75 hIttVps://www.socscistatistics.com/tests/fisher/Default2.aspx UN Fisher‟s exact test for association of T3SS and clonality CLONALITY T3SS Total p-value exoU exoS Clonal 17 13 30 0.351 Non-clonal 20 25 45 Total 37 38 75 www.graphpad.com/quickcalcs/contingency1.cfm 206 One-Way Analysis of variance (ANOVA) to determine the association between antibiotic sensitivity profile of MBL+ve and MBL-ve strains ANOVA Formulas One-Way ANOVA Table Degrees of Sum of Mean Square Source F-Stat P-Value Freedom DF Squares SS MS Between Right tail of k – 1 SSB MSB = SSB / (k − 1) F = MSB / MSW Y Groups RF(k-1,N-k) A Within N – k SSW MSW = SSW / (N − k) Groups BR I SST = L Toal: N – 1 SSB+SSW N Between Groups Degrees of Freedom: D, wADF = k − 1 here k is the number of groups Within Groups Degrees of Freedom: DF = N A– k, where N is the total number of subjects IB Total Degrees of Freedom: DF = N F− 1 k 2 Sum of Squares Between Grou pOs: SSB = S i=1ni (xi − x) , where ni is the number of subjects in the i-th group k 2 Sum of Squares WithiInT GrYoups: SSW = S i=1(ni − 1) Si , where Si is the standard deviation of the i-th group Total Sum of SRquaSres: SST = SSB + SSW Mean SquaEre Between Groups: MSB = SSB / (k − 1) Mean ISVquare Within Groups: MSW = SSW / (N − k) F-NStatistic (or F-ratio): F = MSB / MSW U 207 Piperacillin Data Summary Groups N Mean Std. Dev. Std. Error MBL -ve 364 24.2253 6.1928 0.3246 MBL +ve 66 9.2273 13.692 1.6854 ANOVA Summary Degrees of Sum of Squares Mean Square Source F-Stat P-Value Freedom DF SS MS Between 1 12567.3457 12567.3457 206.0305 0 Groups Within 428 26106.9363 60.9975 Y Groups R Total: 429 38674.282 RA Piperacillin tazobactam B Data Summary I Groups N Mean Std. Dev. Std . LError MBL -ve 364 24.2253 6.1928 N0.3246 MBL +ve 66 9.2273 7.8483 A 0.9661 ANOVA Summary D Degrees of Sum of Squares Mean Square Source F-Stat P-Value Freedom DF SS AMS Between B 1 1495F6.6 28I8 14956.6288 357.1223 0 Groups Within 428 O17925.058 41.881 Groups Total: 429 TY 32881.6868 Ceftazidime I Groups SN Mean Std. Dev. Std. Error MBL -veE R 364 25.6319 5.2415 0.2747 MBILV +ve 66 4.3485 5.8533 0.7205 N ANOVA Summary Degrees of Freedom Sum of Squares Mean Square U Source F-Stat P-Value DF SS MS Between 1 25308.0613 25308.0613 887.8719 0 Groups Within 428 12199.7888 28.5042 Groups Total: 429 37507.8502 208 Cefoperazone Data Summary Groups N Mean Std. Dev. Std. Error MBL -ve 364 25.1374 5.8801 0.3082 MBL +ve 66 4.3485 5.8533 0.7205 ANOVA Summary Degrees of Freedom Sum of Squares Mean Square Source F-Stat P-Value DF SS MS Between 1 24145.7046 24145.7046 699.3116 0 Groups Y Within Groups 428 14777.9069 34.5278 R Total: 429 38923.6116 R A Imipenem Data Summary IB Groups N Mean Std. Dev. LStd. Error MBL –ve 364 25.1374 5.8801 N 0.3082 MBL +ve 66 4.3485 5.8A533 0.7205 ANOVA SummaDry Degrees of Sum of Squares Mean Square Source Freedom BA F-Stat P-Value SS MS DF I Between 1 24145.7046 24145.7046 699.3116 0 Groups F Within 428 OY 14777.9069 34.5278 Groups Total: 429 IT 38923.6116 Meropenem S Data Summary GERroups N Mean Std. Dev. Std. Error IV MBL –ve 364 25.2885 5.7031 0.2989 N MBL +ve 66 4.3485 5.8533 0.7205 U ANOVA Summary Degrees of Mean Source Sum of Squares SS F-Stat P-Value Freedom DF Square Between Groups 1 24497.9768 24497.9768 747.141 0 Within Groups 428 14033.6748 32.789 Total: 429 38531.6515 209 Doripenem Data Summary Groups N Mean Std. Dev. Std. Error MBL –ve 364 25.2143 5.7341 0.3005 MBL +ve 66 4.3485 5.8533 0.7205 ANOVA Summary Degrees of Sum of Squares Mean Source F-Stat P-Value Freedom DF SS Square MS 24324.66 735.11 Between Groups 1 24324.6693 0 93 37 Within Groups 428 14162.3776 33.0897 Y R Total: 429 38487.0469 A Cefepime BR Data Summary I Groups N Mean Std. Dev. LStd. Error MBL –ve 364 21.9286 10.12N11 0.5305 MBL +ve 66 4.3485 5A.8533 0.7205 ANOVA SummarDy Degrees of Sum of Squares Mean Source F-Stat P-Value Freedom DF IBS AS Square MS 17267.10 187.51Between Groups 1 F 17267.1056 0 56 7 Within Groups 428 O 39411.4823 92.0829 Total: Y429 56678.588 Ampicillin SI T R Data Summary Groups N Mean Std. Dev. Std. Error MBEL –ve 364 0.2418 1.4476 0.0759 V I MBL +ve 66 0 0 0 N ANOVA Summary U Degrees of Sum of Mean Square Source F-Stat P-Value Freedom DF Squares SS MS Between Groups 1 3.2666 3.2666 1.8379 0.1759 Within Groups 428 760.6831 1.7773 Total: 429 763.9497 210 Amoxicillin-clavulanic acid Data Summary Groups N Mean Std. Dev. Std. Error MBL –ve 364 0.4615 2.0358 0.1067 MBL +ve 66 0 0 0 ANOVA Summary Degrees of Sum of Mean Square Source F-Stat P-Value Freedom DF Squares SS MS Between Groups 1 11.8993 11.8993 3.3852 0.0665 Within Groups 428 1504.4468 3.5151 Y Total: 429 1516.3461 RA R Cefuroxime B Data Summary I Groups N Mean Std. De vL. Std. Error MBL -ve 364 0.4615 2.N0358 0.1067 MBL +ve 66 0.3636 DA1.6792 0.2067 ANOVA Summary Degrees of Sum of MAean Square Source IB F-Stat P-Value Freedom DF Squares SS MS Between 1 0.5355 0.5355 0.1358 0.7127 Groups F Within Groups 428 O1687.7282 3.9433 Total: 42T9 Y 1688.2636 Cephalothin SI Data Summary Groups ER N Mean Std. Dev. Std. Error MBILV -ve 364 0.4615 2.0358 0.1067 NMBL +ve 66 0.7879 2.5269 0.311 U ANOVA Summary Degrees of Freedom Sum of Squares Mean Square Source F-Stat P-Value DF SS MS Between Groups 1 5.9522 5.9522 1.3272 0.2499 Within Groups 428 1919.4864 4.4848 Total: 429 1925.4386 211 Ticarcillin Data Summary Groups N Mean Std. Dev. Std. Error MBL -ve 364 12.4725 9.3143 0.4882 MBL +ve 66 1.1061 3.548 0.4367 ANOVA Summary Degrees of Sum of Squares Mean Square Source F-Stat P-Value Freedom DF SS MS Between Groups 1 7218.0973 7218.0973 95.6136 0 Within Groups 428 32310.7347 75.4924 Total: 429 39528.8321 AR Y Ceftriaxone BR Data Summary I Groups N Mean Std. Dev. Std. E rrLor MBL -ve 364 11.9753 9.4442 N0.495 MBL +ve 66 1.1061 3.548D A0.4367 ANOVA SummarAy Degrees of Sum of Squares Mean Square Source IB F-Stat P-Value Freedom DF SS MS Between Groups 1 6600.4 269 6600.4269 85.102 0 Within Groups 428 O3319 F5.2674 77.559 Total: 429 Y 39795.6943 Aztreonam T SI Data Summary Groups RN Mean Std. Dev. Std. Error MBL -veE 364 21.9286 10.1211 0.5305 MBLI V+ve 66 7.0909 8.4065 1.0348 N ANOVA Summary U Degrees of Sum of Squares Mean Square Source F-Stat P-Value Freedom DF SS MS Between 1 12300.1395 12300.1395 126.0103 0 Groups Within Groups 428 41778.0102 97.6122 Total: 429 54078.1497 212 Gentamicin Data Summary Groups N Mean Std. Dev. Std. Error MBL -ve 364 21.8736 10.1014 0.5295 MBL +ve 66 5.6364 7.3456 0.9042 ANOVA Summary Degrees of Sum of Mean Square Source F-Stat P-Value Freedom DF Squares SS MS Between Groups 1 14729.8778 14729.8778 155.4829 0 Within Groups 428 40547.1559 94.7363 Y R Total: 429 55277.0337 A Amikacin BR Data Summary I Groups N Mean Std. Dev. L St d. Error MBL -ve 364 21.5495 10.4114A N0.5457 MBL +ve 66 5.6364 7.3456 0.9042 ANOVA Summary D Degrees of Sum of MAean Square Source F-Stat P-Value Freedom DF Squares SIS B MS Between Groups 1 1414F7.71 97 14147.7197 141.2941 0 Within Groups 428 4O2855.4613 100.1296 Total: 429 57003.181 Tobramycin ITYS Data Summary Groups N Mean Std. Dev. Std. Error MBL -vEe R 364 20.3571 11.1994 0.587 MBILV +ve 66 7.5152 7.6986 0.9476 N ANOVA Summary U Sum of Mean Degrees of P-Source Squares Square F-Stat Freedom DF Value SS MS Between Groups 1 9213.7289 9213.7289 79.8561 0 Within Groups 428 49382.2901 115.3792 Total: 429 58596.0191 213 Ciprofloxacin Data Summary Groups N Mean Std. Dev. Std. Error MBL –ve 364 19.7445 11.1218 0.5829 MBL +ve 66 6.9394 6.9035 0.8498 ANOVA Sumary Degrees of Sum of Squares Mean Square Source F-Stat P-Value Freedom DF SS MS Between Groups 1 9160.9985 9160.9985 81.6875 0 Within Groups 428 47998.8703 112.1469 Y Total: 429 57159.8688 AR Levofloxacin R Data Summary IB Groups N Mean Std. Dev. LStd. Error MBL –ve 364 19.467 11.294 N 0.592 MBL +ve 66 7.8485 7.452A9 0.9174 ANOVA Summary D Degrees of Sum of Squares Mean Square Source IB A F-Stat P-Value Freedom DF SS MS Between Groups 1 754 1.8343 7541.8343 64.671 0 Within Groups 428 OF49912.732 116.6185 Total: 429Y 57454.5663 Ofloxacin IT Data Summary Groups RN S Mean Std. Dev. Std. Error MBL –ve E 366 19.0301 11.5551 0.604 MBLI V+ve 66 6.5152 6.9332 0.8534 N ANOVA Summary U Degrees of Sum of Squares Mean Square P-Source F-Stat Freedom DF SS MS Value Between 1 8757.8205 8757.8205 72.6167 0 Groups Within Groups 430 51859.4247 120.6033 Total: 431 60617.2452 214 Colistin sulphate Data Summary Groups N Mean Std. Dev. Std. Error Group 1 364 12.2005 1.6848 0.0883 Group 2 66 11.7273 2.853 0.3512 ANOVA Summary Degrees of Sum of Squares Mean Square Source Freedom F-Stat P-Value SS MS DF Between 1 12.5103 12.5103 3.4335 0.0646 Y Groups R Within 428 1559.4686 3.6436 A Groups BRTotal: 429 1571.9789 I Polymixin B L Data Summary AN Std. Groups N Mean SDtd. Dev. Error MBL -ve 366 14.5984 BA 1.9217 0.1005 I MBL +ve 64 15.5781 12.832 1.604 ANOVFA Summary Degrees of Sum of Mean Source FreedYom O Squares Square F-Stat P-Value TDF SS MS Between SI 0.167 1 52.2852 52.2852 1.9091 Groups 8 Within GroupsR 428 11721.5139 27.3867 Total:E 429 11773.7991 V httNps:/I/goodcalculators.com/one-way-anova-calculator/ U 215 Appendix IV Selected genbank flatfiles and BLASTn of blaVIM and blaNDM sequences Preliminary genbank flatfile(s: LOCUS Seq1 343 bp DNA linear BCT 21-JUL-2019 DEFINITION metallo-beta-lactamase VIM-5 (blaVIM) gene, partial cds. ACCESSION MH201592 VERSION KEYWORDS SOURCE Pseudomonas aeruginosa ORGANISM Pseudomonas aeruginosa Y Bacteria; Proteobacteria; Gammaproteobacteria; Pseudomonadales; R Pseudomonadaceae; Pseudomonas. A REFERENCE 1 (bases 1 to 365) R AUTHORS Olaniran,O.B., Adeleke,O.E. and Bukhari,S.H. TITLE Pseudomonas aeruginosa VIM-5 metallo beta-lIaBctamase JOURNAL Unpublished L REFERENCE 2 (bases 1 to 365) AUTHORS Olaniran,O.B., Adeleke,O.E. and BukhaNri,S.H. TITLE Direct Submission JOURNAL Submitted (21-JUL-2019) PharmaceutAical Microbiology, University D of Ibadan, Ibadan, Nigeria, AIbadan, Oyo +234, Nigeria COMMENT Bankit Comment: LocalID:ISBeq1. Bankit Comment: BankIt 2246580. ##Assembly-Data-STARFT## Sequencing Technology :: Sanger dideoxy sequencing ##Assembly-DaYta- E OND## FEATURES LTocation/Qualifiers source 1..365 S I/organism="Pseudomonas aeruginosa" /mol_type="genomic DNA" R /strain="BT1168" /isolate="PS168" V E /isolation_source="Homo sapiens" I /bio_material="Wound" N /db_xref="taxon:287" /country="Nigeria" U /collection_date="2015" /biotype="Bacteria" /note="[cultured bacterial source]" gene <1..>365 /gene="blaVIM-5 metallo-beta-lactamase" /allele="blaVIM-5" CDS <1..>365 /gene="blaVIM-5 metallo-beta-lactamase" /allele="blaVIM-5" /note="[intronless gene]; carbapenemase" /codon_start=1 216 /transl_table=11 /product="VIM-5 metallo-beta-lactamase" /translation="WSHIATQSFDGAVYPSNGLIVRDGDELLLIDTAWGAK NTA ALLAEIEKQIGLPVTRAVSTHFHDDRVGGVDVLRKAGVATYASPSTRRLAEAE GNEIPTHSLEGLSSSGDAVRFGPVELFY" BASE COUNT 74 a 100 c 110 g 81 t ORIGIN 1 tggtcgcata tcgcaacgca gtcgtttgat ggcgcggtct acccatccaa tggtctcatt 61 gtccgtgatg gtgatgagtt gcttttgatt gatacagcgt ggggtgcgaa aaacacagcg 121 gcccttctcg cggagattga gaagcaaatt ggacttcccg tgacgcgtgc agtctccacg 181 cactttcatg acgaccgcgt cggcggcgtt gatgtccttc ggaaggctgg agtggcaacYg 241 tacgcatcac cgtcgacacg ccggctagcc gaggcagagg ggaacgagat tcccacgcac 301 tctctagaag gactctcatc gagcggggac gcagtgcgct tcggtccagt agaAgctRcttc 361 tatcc // BR LI DA N IB A F Y O IT S R IV E UN 217 LOCUS Seq10 454 bp DNA linear BCT 20-JUL-2019 DEFINITION metallo-beta-lactamase NDM-1 (blaNDM) gene, partial cds. ACCESSION MN193051 VERSION KEYWORDS . SOURCE Pseudomonas aeruginosa ORGANISM Pseudomonas aeruginosa Bacteria; Proteobacteria; Gammaproteobacteria; Pseudomonadales; Pseudomonadaceae; Pseudomonas. REFERENCE 1 (bases 1 to 454) AUTHORS Olaniran,O.B., Adeleke,O.E. and Bukhari,S.H. TITLE Pseudomonas aeruginosa NDM-1 metallo beta-lactamase Y JOURNAL Unpublished REFERENCE 2 (bases 1 to 454) R AUTHORS Olaniran,O.B., Adeleke,O.E. and Bukhari,S.H. A TITLE Direct Submission JOURNAL Submitted (20-JUL-2019) Pharmaceutical R Microbiology,University IB of Ibadan, Ibadan, Nigeria, Ibadan, Oyo +L234, Nigeria COMMENT Bankit Comment: LocalID:Seq10. N Bankit Comment: BankIt2246115. ##Assembly-Data-START## A Sequencing Technology :: SangerD dideoxy sequencing ##Assembly-Data-END## A FEATURES Location/QualifIiers source 1..454 B /organism="Pse udomonas aeruginosa" /mol_type="Fgenomic DNA" /strain="BT1154" /isoYlat e O="PS154" /Tisolation_source="Homo sapiens" I/bio_material="Urine" S /db_xref="taxon:287" /country="Nigeria" R /collection_date="2015" E /biotype="Bacteria" V /note="[cultured bacterial source]" genIe <1..>454 N /gene="blaNDM-1 metallo-beta-lactamase" /allele="blaNDM-1" U CDS <1..>454 /gene="blaNDM-1 metallo-beta-lactamase" /allele="blaNDM-1" /note="[intronless gene]; carbapenemase" /codon_start=1 /transl_table=11 /product="blaNDM-1 metallo-beta-lactamase" /translation="SNGLIVRDGGRVLVVDTAWTDDQTAQILNWIKQEINLP VALAVVTHAHQDKMGGMDALHAAGIATYANALSNQLAPQEGMVAAQHSLTFA ANGWVEPATAPNFGPLKVFYPGPGHTSDNITVGIDGTDIAFGGCLIKDSKAK 218 SLGNLGDAD" BASE COUNT 92 a 139 c 141 g 82 t ORIGIN 1 tccaacggtt tgatcgtcag ggatggcggc cgcgtgctgg tggtcgatac cgcctggacc 61 gatgaccaga ccgcccagat cctcaactgg atcaagcagg agatcaacct gccggtcgcg 121 ctggcggtgg tgactcacgc gcatcaggac aagatgggcg gtatggacgc gctgcatgcg 181 gcggggattg cgacttatgc caatgcgttg tcgaaccagc ttgccccgca agaggggatg 241 gttgcggcgc aacacagcct gactttcgcc gccaatggct gggtcgaacc agcaaccgcg 301 cccaactttg gcccgctcaa ggtattttac cccggccccg gccacaccag tgacaatatc 361 accgttggga tcgacggcac cgacatcgct tttggtggct gcctgatcaa ggacagcaag 421 gccaagtcgc tcggcaatct cggtgatgcc gaca // RY RA LIB AN AD F I B O SI TY VE R NIU 219