Escherichia coli infection: mechanisms and rapid detection of resistance to the association of ß-lactams with ß-lactamase inhibitors, and impact of the Sars-Cov-2 pandemic on the morbidity and mortality of intra-abdominal infection

Escherichia coli infection MECHANISMS AND RAPID DETECTION OF RESISTANCE TO THE ASSOCIATION OF ß-LACTAMS WITH ß-LACTAMASE INHIBITORS, AND IMPACT OF THE SARS-COV-2 PANDEMIC ON THE MORBIDITY AND MORTALITY OF INTRA-ABDOMINAL INFECTION PhD Tesis of LIDIA GÁLVEZ BENÍTEZ Directed by: Younes Smani José Antonio Lepe Jiménez Mentored by: María Elisa Cordero Matía Sevilla, 2023 El Prof. Dr. José Antonio Lepe Jiménez, Jefe de Servicio de Microbiología en el Hospital Universitario Virgen del Rocío y Profesor Asociado en el Departamento de Microbiología de la Universidad de Sevilla, el Prof. Dr. Younes Smani, Profesor Titular del Departamento de Biología Molecular e Ingeniería Bioquímica “Área Microbiología” de la Universidad Pablo de Olavide,, como Directores de la Tesis Doctoral, y la Profa. Dra. María Elisa Cordero Matía, Profesora Titular (v) del Departamento de Medicina de la Universidad de Sevilla, como Tutora de la Tesis Doctoral, CERTIFICAN Que la Tesis para optar al Grado de Doctor por la Universidad de Sevilla titulada “Infección por Escherichia coli: mecanismos y detección rápida de la resistencia a la asociación de ß-lactámicos con inhibidores de ß-lactamasas, e impacto de la pandemia por SARS-CoV-2 en la morbimortalidad de la infección intraabdominal” ha sido realizada por la Graduada Dña. Lidia Gálvez Benítez bajo nuestra supervisión, considerando que reúne los requisitos necesarios para su presentación. Y para que conste a los efectos oportunos, expiden la presente certificación en Sevilla a 12 de junio de 2023. Dr. Younes Smani Director Dr. José Antonio Lepe Jiménez Director Dra. María Elisa Cordero Matía Tutora AgrAdecimientos Parece mentira que al fin haya llegado este día, el día en que me toca darle los últimos retoques a este trabajo y ponerme a reflexionar sobre los agradecimientos de mi Tesis Doctoral. Este proyecto ha sido posible gracias a la contribución de muchas personas, y me siento muy afortunada porque tengo mucho y a muchos a los que agradecer. En primer lugar, a Younes Smani y Elisa Cordero, sin los que este proyecto no habría sido posible, por el tiempo dedicado y por la paciencia todos estos años. Pero si realmente hay alguien a quien tengo que agradecer el camino andado hasta llegar al día de hoy es a José Antonio Lepe, al que debo agradecerle todo: por su impulso imprescindible, por compartir conmigo sus conocimientos y por haber sido inasequible todas las veces que pensé que no era capaz de seguir. De él he aprendido mucho sobre Microbiología, mucho sobre resistencias a antibióticos, mucho sobre investigación… pero los mejores han sido sin duda los momentos en los que hemos compartido reflexiones sobre literatura, sobre música o sobre la vida que hay fuera del laboratorio. Por todo esto y mucho más no tengo palabras para decirle lo agradecida que me siento con él y el cariño inmenso que le profeso. Gracias. Estoy profundamente agradecida a todas las instituciones y personas que han participado, directa o indirectamente, en sacar adelante este trabajo. A los pacientes, por su participación altruista. A todo el equipo de profesionales del Hospital Universitario Virgen del Rocío que han hecho posible la recopilación de datos que aquí se presentan. También al Dr. Thierry Naas, que me permitió rotar en su laboratorio para aprender sobre secuenciación. Sin ellos habría sido imposible. A todos los compañeros que forman el laboratorio de investigación de Infecciones víricas y en pacientes inmunodeprimidos, que han colaborado conmigo y me han echado una mano cuando era incapaz de terminar los experimentos o no encontraba dónde estaban guardadas las pipetas. Muy especialmente me gustaría agradecer a Andrea Vila, que ha sido mi compañera de batallas en esta aventura y sin la que habría sido imposible que la mayoría de los experimentos de este trabajo hubieran visto la luz. Tengo que agradecerle al Dr. Jerónimo Pachón su inestimable colaboración: por abrirme las puertas de su laboratorio, por compartir conmigo su maestría, por sus valiosos consejos, por su disponibilidad permanente. A todos los compañeros que forman el Servicio de Microbiología del Hospital Universitario Virgen del Rocío, que me han acompañado durante estos años de formación. Muy especialmente a Marisa Calle, Marisa Cubero, Javi Dorado, Charo Durán, Pilar Durán, Mila López, Nati Márquez, | 3 Paloma Parras, Inma Ramírez, Isabel Vázquez y Candela Velasco: habéis sido mis manos y mis pies en ese laboratorio, me habéis enseñado a sembrar, a hacer antibiogramas, a calibrar y disparar el MALDI… pero sobre todo me habéis acompañado y alentado en los momentos en los que sentía que no iba a poder ganarle el pulso a este proyecto. A todos los facultativos del servicio, al equipo completo, especialmente a María José Gómez, Mª Carmen Lozano, Laura Merino e Inmaculada Pupo, por todo lo que me han enseñado e inspirado. Porque les debo la Microbióloga que soy hoy pero, sobre todo, son el ejemplo de la Microbióloga que me gustaría llegar a ser. Y por supuesto, a Verónica González y a Maite Ruiz, mis tutoras durante el periodo MIR aunque su tutela haya trascendido lo puramente académico: sin ellas, sin su generoso esfuerzo, sin su consejo, sin su esmero y sin su aliento jamás habría podido llegar hasta aquí. Gracias a todos vosotros, compañeros, el ir a trabajar cada día ha sido para mí un regalo, un juego y una aventura… incluso cuando las cosas venían torcidas. No quiero dejar de agradecer también al Dr. Javier Aznar, que fue mi primer jefe, la confianza que siempre puso en mí y cómo supo impulsarme hacia retos que a mí misma me parecían insalvables desde el primer día que pisé el laboratorio. Gracias, Jefe. Te prometí que no abandonaría la especialidad y no lo hice. Tengo que agradecer también a José Miguel Cisneros aquella clase de tercero de Medicina cuando oí hablar por primera vez del uso racional de los antibióticos y a José Molina, que dirigió mi Trabajo Fin de Grado: aquellos momentos de mi formación como médico fueron sin duda fueron los cimientos de mi amor por las Enfermedades Infecciosas y la razón de que acabase dedicándome a esto. A Guillermo Martín y Ángel Rodríguez, que me orientaron, guiaron e inspiraron para seguir el camino del clínico-investigador. A José Manuel Ortiz, por todo lo que me ha enseñado pero, sobre todo, lo fácil y divertido que hace trabajar con él. A mis compañeros, los residentes del Servicio de Microbiología: Carmen Cintora, Valentina Hoffmann, Adrián Pulido y Carmen Soto, por acompañarme y alegrarme cada día. A María Aznar y a José Luis García, por la compañía constante, por la paciencia, por el cariño. Nunca les podré agradecer lo suficiente la fortuna de que apareciesen en mi camino, por haber sido capaces de arrancarme una sonrisa hasta en los momentos más difíciles, por su capacidad de hacer suyas mis alegrías y por haber confiado en mí mucho más de lo que yo era capaz de hacerlo. Finalmente, a mi familia. A mis padres, Juan Carlos e Inmaculada. Me habéis enseñado a querer incondicionalmente, a hacer mi trabajo bien para servir a los demás, me habéis educado con vuestro ejemplo en la importancia del esfuerzo, de la honestidad y la dedicación. A vosotros os debo 4 | las incontables horas que me ha requerido este trabajo porque sois vosotros los que me habéis sostenido en los momentos más difíciles. Es vuestra ternura y vuestra fuerza la que me sostiene y me alienta y es en vosotros en quienes mido el valor de las cosas. Todo proyecto que yo empiece y termine será siempre para vosotros. | 5 Phd thesis funding 1. Instituto de Salud Carlos III y Ministerio de Ciencia e Innovación, Gobierno de España. Programa Rio Hortega (CM22/00196). PI: Lidia Gálvez Benítez. 2. Instituto de Salud Carlos III y Ministerio de Ciencia e Innovación, Gobierno de España. Programas Miguel Servet tipos I y II (CP15/00132 y CP20/00018). PI: Younes Smani 3. Instituto de Salud Carlos III. Proyecto de Investigación en Salud (PI19/01009): “Estudio clínico y microbiológico del impacto del tratamiento con piperacilina/tazobactam (P/T) en pacientes con infecciones intraabdominales por Escherichia coli portadores de beta-lactamasas”. PI: Younes Smani. 4. Junta de Andalucía. Fundación Pública Andaluza para la Gestión de la Investigación en Salud de Sevilla (FISEVI). Premio para residentes del Hospital Universitario Virgen del Rocío con el objetivo de realizar rotaciones en el extranjero de, al menos, dos meses de duración en el periodo correspondiente al año 2022. PI: Lidia Gálvez Benítez. 5. Sociedad Española de Enfermedades Infecciosas y Microbiología Clínica. Ayuda SEIMC a la formación: rotatorios para formación de residentes de entre 1 y 3 meses. PI: Lidia Gálvez. scientific Production relAted to Phd thesis PUBLICATIONS 1. Gálvez Benítez L*, Ortiz de la Rosa JM*, Rodríguez-Villodres A, et al. Role of blaTEM and OmpC in the piperacillin-tazobactam resistance evolution by E. coli in patients with complicated intra-abdominal infection. Journal of infection. doi: 10.1016/j.jinf.2023.07.005. * Both authors contributed equally. 2. Rodríguez Villodres A*, Gálvez Benítez L*, Arroyo MJ, Méndez G, Mancera L, Vila Domínguez A, Lepe Jímenez JA, Smani Y. Ultrasensitive and rapid identification of ESRI developer- and piperacillin/tazobactam-resistant Escherichia coli by the MALDIpiptaz test. Emerg Microbes Infect. 2022 Dec;11(1):2034-2044. doi: 10.1080/22221751.2022.2113746 * Both authors contributed equally. 6 | 3. Rodríguez-Villodres Á, Gutiérrez Linares A, Gálvez-Benitez L, Pachón J, Lepe JA, Smani Y. Semirapid Detection of Piperacillin/Tazobactam Resistance and Extended-Spectrum Resistance to β-Lactams/β-Lactamase Inhibitors in Clinical Isolates of Escherichia coli. Microbiol Spectr. 2021 Oct 31;9(2):e0080121. doi: 10.1128/Spectrum.00801-21. 4. Gálvez-Benítez L, Rodríguez-Villodres Á, Álvarez-Marín R, Jiménez-Rodríguez R, Lepe-Jiménez JA, Pachón J, Smani Y. Impact of the COVID-19 Pandemic on Survival in the Patients With the Intra-Abdominal Infections. Front Med (Lausanne). 2021 Oct 20;8:687415. doi: 10.3389/ fmed.2021.687415. CONGRESS COMMUNICATIONS 1. Gálvez Benítez L, Ortiz de la Rosa JM, Rodríguez Villodres A, Molina Panadero I, Álvarez Marín R, Bonnin RA, Naas T, Pachón J, Cisneros JM, Lepe JA, Smani Y. Bases genéticas de la adquisición de la resistencia in vivo a piperacilina-tazobactam en Escherichia coli. XXVI Congreso Nacional de la Sociedad Española de -Enfermedades Infecciosas y Microbiología Clínica. Spain, 2023. 2. Gálvez Benítez L*, Ortiz de la Rosa JM*, Rodríguez Villodres A, Casimiro-Soriguer S. C, Molina Panadero I, Alvarez Marin, Bonnin RA, Naas T, Pachón J, Cisneros JM, Lepe JA, Smani Y. Genetic basis for in vivo piperacillin-tazobactam resistance. 33rd European Congress of Clinical Microbiology and Infectious Diseases. Copenhague, Denmark 2023. 3. Gálvez Benítez L, Rodríguez-Villodres A, Alvarez Marín R, Jiménez Rodríguez R, Lepe JA, Pachón J, Smani Y. Impacto de la pandemia COVID-19 en la mortalidad de pacientes con infecciones intraabdominales. XXIV Congreso de la Sociedad Española de Enfermedades Infecciosas y Microbiología Clínica, Spain 2021 | 7 index Introduction THE IMPORTANCE OF ESCHERICHIA COLI .................................................................... 26 The genus Escherichia .............................................................................................................................. 26 Reservoirs, transmission mechanisms and epidemiology of E. coli colonisation and infection ................................................................................................................................................... 27 Microbiological characteristics of E. coli of major clinical interest ..............................28 The plasticity of the E. coli genome .................................................................................................................28 Plasmids ............................................................................................................................................................31 Transposons .......................................................................................................................................................31 Gene cassettes and integrons .........................................................................................................................32 Genomic islands ...............................................................................................................................................33 Pathogenicity determinants ..............................................................................................................................33 Clinical presentation of E. coli infections ......................................................................................33 Intra-abdominal infections ...............................................................................................................................34 Bacteraemia ..........................................................................................................................................................34 Prognosis of E. coli infections ...............................................................................................................35 Mortality and morbidity associated with intra-abdominal E. coli infections ....................................... 35 Mortality and morbidity associated with E. coli bacteraemia ..................................................................35 8 | Antimicrobial treatment of E. coli infections ..............................................................................36 Antibiotics with activity against E. coli ............................................................................................................36 ß-lactams ...........................................................................................................................................................36 Piperacillin-tazobactam .................................................................................................................................... 40 Pharmacokinetics and pharmacodynamics of the ß-lactams .......................................................................41 THE THREAT OF ANTIBIOTIC RESISTANCE .......................................................................43 An historical perspective on the development of antimicrobial resistance ..........43 Impact of antimicrobial resistance ...................................................................................................44 Mechanisms of antimicrobial resistance in E. coli ..................................................................44 Main mechanisms of ß-lactam resistance in E. coli .....................................................................................44 Changes in the membrane permeability .......................................................................................................45 Active expulsion systems: efflux pumps .........................................................................................................47 Target modification: altered expression of PBPs ............................................................................................47 Antibiotic modification: the ß-lactamases ......................................................................................................47 Extended spectrum ß-lactamases ..................................................................................................................49 Piperacillin-tazobactam resistance in E. coli ..................................................................................................49 Epidemiological situation of piperacillin-tazobactam resistance in E. coli in our setting ................................................................................................................................................................ 50 Methods and strategies for early detection of antimicrobial resistance ..................51 The impact of the use of rapid diagnostic techniques for the detection of bacterial resistance to antibiotics .............................................................................................................................................................. 51 | 9 Promoter analysis .............................................................................................................................................137 Study of blaTEM copy number .......................................................................................................................... 138 ß-lactamase activity ..........................................................................................................................................139 Analysis of porins-expression in these isolates .........................................................................................139 P/T resistance remains stable .........................................................................................................................140 Study of growth rate ........................................................................................................................................140 To develop and to validate a system for rapid detection of high-level resistance to P/T and other ß-lactams in combination with ß-lactamase inhibitors (ESRI) in clinical isolates of E. coli ..........................................................................................................142 To analyse the impact of the SARS-CoV-2 pandemic on the morbidity and mortality of intra-abdominal E. coli infection ..........................................................................................144 Conclussions OBJECTIVE 1 ............................................................................................................................................148 OBJECTIVE 2 ............................................................................................................................................148 OBJECTIVE 3 ............................................................................................................................................149 Bibliography Annexes 16 | figures list FIGURE 1. Escherichia coli morphology ...........................................................................................26 FIGURE 3.Horizontal gene transfer. ................................................................................................29 FIGURE 4.Examples of mobile genetic elements (MGE) and processes involved in intracellular mobility or intercellular transfer of antibiotic resistance genes. ....................................................30 FIGURE 5. Chemical structure of the main representative ß-lactam antibiotics used in clinical practice. ...................................................................................................................................................37 FIGURE 6. Mechanism of action of ß-lactams on the bacterial wall. ...............................................38 FIGURE 7. Main mechanisms of resistance to ß-lactam antibiotics in Gram-negative bacteria. .....45 FIGURE 8. The two-component EnvZ-OmpR system. .......................................................................46 FIGURE 9. Diagram of sample processing in the laboratory of Microbiology Department at Virgen del Rocio University Hospital. .......................................................................................................... 64 FIGURE 10.Schematic summary of the most relevant clinical data of the episode in which the samples were obtained from patient #1. ...............................................................................................94 FIGURE 11.Schematic summary of the most relevant clinical data of the episode in which the samples were obtained from patient #2. ...............................................................................................95 FIGURE 12.Schematic summary of the most relevant clinical data of the episode in which the samples were obtained from patient #3. ...............................................................................................96 FIGURE 13.Schematic summary of the most relevant clinical data of the episode in which the samples were obtained from patient #4. ...............................................................................................97 FIGURE 14. BLAST atlas of the blaTEM-1 harbouring plasmids comparison. ...................................102 FIGURE 15. BLAST atlas of the blaTEM-1 harbouring plasmids comparison. ...................................103 FIGURE 16. BLAST atlas of the blaTEM-1 harbouring plasmids comparison. ...................................104 | 17 FIGURE 17. BLAST atlas of the blaTEM-1 harbouring plasmids comparison. ...................................105 FIGURE 18. (A) ompC gene transcription (B) ...............................................................................106 FIGURE 19. ß-lactamase activity of the four couples of E. coli isolates . ........................................107 FIGURE 20.blaTEM gene transcription in the four couples of E. coli isolates. .................................107 FIGURE 21. Impact of the P/T resistance on bacterial fitness. .......................................................109 FIGURE 22. Distance plot of RF and PLS-DA for discrimination between ESRI developers susceptible to P/T and their isogenic isolates resistant to P/T. ........................................................................118 FIGURE 23. Distance plot of RF and PLS-DA for discrimination between P/T-susceptible and P/T- resistant E. coli (first step analysis). ..............................................................................................119 FIGURE 24. Distance plot of RF and PLS-DA for discrimination between P/T-susceptible and ESRI developer Escherichia coli (second step analysis). ....................................................................... 120 FIGURE 25. Feature importance from RF analysis from each assay. .............................................122 FIGURE 26. Average ROC curves from first 10-folds and their AUC values obtained from RF using Full Spectrum method to discriminate ........................................................................................123 FIGURE 27.Kaplan-Meier plots among patients with intraabdominal infections during pre- COVID-19 and COVID-19 periods. ................................................................................................125 18 | tAbles list TABLE 1.Characteristics of ß-lactamase inhibitors of clinical interest. ....................................................39 TABLE 2. Classification of ß-lactamases ..........................................................................................................48 TABLE 3.Specific primers for the amplification of the E. coli reference genes ......................................67 TABLE 4. Specific primers used for ß-lactamase gene detection. ............................................................. 70 TABLE 5. Migration gel composition (10%) ................................................................................................... 77 TABLE 6. Concentration gel composition .......................................................................................................77 TABLE 7. Specific primers used for real-time detection of ompC. ...........................................................80 TABLE 8. Specific primers used for real-time detection of blaTEM. .............................................................81 TABLE 9. Parameters applied for peak matrices built in each assay and method. ............................... 87 TABLE 10. Microbiological characteristics of E. coli clinical isolates. .......................................................98 TABLE 11. Coverage of selected plasmidic and chromosomal genes in clinical isolates. ..............100 TABLE 12. MIC of P/T in pretreated E. coli isolates after in vitro pressure study with subinhibitory concentrations of P/T. .......................................................................................................................................108 TABLE 13. MIC of P/T in post-therapy E. coli isolates after removal of antimicrobial pressure. ...... 108 TABLE 14. Demographical, epidemiological and microbiological data of the patients and the E. coli isolates. ...............................................................................................................................................................110 TABLE 15. Clinical isolates of E. coli susceptible to P/T.............................................................................112 TABLE 16. Clinical isolates of E. coli that were ESRI developers. ...........................................................112 TABLE 15. Clinical isolates of E. coli susceptible to P/T (continued) ...................................................... 113 TABLE 16. Clinical isolates of E. coli that were ESRI developers (continued) ......................................113 | 19 TABLE 15. Clinical isolates of E. coli susceptible to P/T (continued) ...................................................... 114 TABLE 16. Clinical isolates of E. coli that were ESRI developers (continued) ......................................114 TABLE 15. Clinical isolates of Es. coli susceptible to P/T (continued) ....................................................115 TABLE 16. Clinical isolates of E. coli that were ESRI developers (continued) ......................................115 TABLE 17.Clinical isolates of E. coli that were ESRI developers with already acquired resistance to P/T. .......................................................................................................................................................................116 TABLE 18. Average results in accuracy terms by repeating 20 times a 10-fold cross validation analysis for all ML supervised algorithms for each method and step ..................................................117 TABLE 19. Demographics, clinical and microbiological variables of the patients with IAI in the pre-COVID-19 pandemic period and during the COVID-19 pandemic period. .................................126 TABLE 19. Demographics, clinical and microbiological variables of the patients with IAI in the pre-COVID-19 pandemic period and during the COVID-19 pandemic period (continued). .........127 TABLE 20. Univariate analysis of the factors related to death in a combined cohort of the pre- COVID-19 pandemic cohort and the COVID-19 pandemic cohort. .......................................................129 TABLE 20. Univariate analysis of the factors related to death in a combined cohort of the pre- COVID-19 pandemic cohort and the COVID-19 pandemic cohort (continued). ...............................130 TABLE 21. Multivariate analysis of the factors related to death in a combined cohort of the pre- COVID-19 pandemic cohort and the COVID-19 pandemic cohort. ......................................................130 20 | AbbreViAtions • ABC: ATP-binding cassette • AMC: amoxicillin/clavulanic acid • AMP: ampicillin • APS: ammonium persulphate • ARDS: acute respiratory distress syndrome • ASP: antimicrobial stewardship programme • ATP: adenosine triphosphate • ATU: area of technical uncertainty • AUC: area under curve • BHI: brain heart infusion • BL/BLI: ß-lactam/ß-lactamase inhibitor • BLAST: basic local alignment search tool • BRIG: basic local alignment search tool ring image generator • CC: clonal complex • CFU: colony-forming unit • CG: chromosomal genes • CGE: Centre for genomic epidemiology • CIP: ciprofloxacin • COVID-19: The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-associated coronavirus diseases 2019 • DDD: defined daily dose • DMT: drug-metabolite transporter • DNA: deoxyribonucleic acid • ESBL: extended-spectrum ß-lactamase • ESRI: Extended-spectrum ß-lactams/ß-lactamase inhibitors resistance • EUCAST: European Committee of Antimicrobial Susceptibility Testing • HGT: horizontal gene transfer • HUVR: Virgen del Rocío University Hospital • IAI: intra-abdominal infection • ICE: integrative conjugative elements • ICT: Immunochromatography • In: integron • IR: inverted repeat • IRT: inhibitor resistant TEM • IS: insertion sequence • IV: intravenous • KPC: Klebsiella pneumoniae carbapenemase • LEV: levofloxacin | 21 • LPS: lipopolysaccharide • MALDI-TOF: Matrix Assisted Laser Desorption/Ionization • MBL: metallobetalactamase • MCC: mean chromosomal coverage • MDS-EB: myelodysplastic syndrome with excess of blasts • MFS: major facilitator system • MGE: mobile genetic elements • MHB: Mueller Hinton Broth • MIC: minimum inhibitory concentrations • MLST: multilocus secquencing typing • MOP: multidrug-oligosaccharidyl-lipid/polysaccharide export • MS: mass spectrometry • OMP: outer membrane protein • P/T: piperacillin/tazobactam • PAI: pathogenicity islands • PBP: penicillin binding-protein • PCA: principal component analysis • PIP: piperacillin • PIRASOA: comprehensive programme on prevention and control of healthcare associated infections and appropriate use of antimicrobials. (Programa integral de prevención y control de las infecciones relacionadas con la asistencia sanitaria y uso apropiado de los antimicrobianos). • PK/PD: pharmacokinetics/pharmacodynamics • PLS-DA: partial least squares discriminant analysis • PRIOAM: Institutional Antimicrobial Stewardship Programme. Programa Institucional para la Optimización de los Antimicrobianos. • RAST: rapid antimicrobial susceptibility testing • RF: random forest • RNA: ribonucleic acid • RND: resistance-nodulation-cell division. • ROC: receiver operating characteristic • SARS-CoV-2: severe acute respiratory syndrome coronavirus 2 • SDS: Sodium dodecyl sulfate • SEM: stardard error of the mean • SHV: shulfhydryl reagent variable • SI: surgical intervention • SMART: Study for Antimicrobial Resistance Trends. • SNP: single nucleotide polymorphisms • SNV: single nucleotide variation • SO: symptoms onset • ssDNA: single-stranded deoxyribonucleic acid 22 | • ST: sequence type • SVM: Linear Support Vector Machines • TCC: ticarcillin/clavulanate • TIC: ticarcillin • TMP-SMX: trimethoprim-sulfamethoxazole • Tn: transposons • Tnp: transposase • tRNA: transfer ribonucleic acid • TSD: target site duplication • TU: translocable unit • WGS: whole genome secquencing • WHO: World Health Organization | 23 Introduction as a class I or compound transposon, more ISs were identified. IS6/IS26-family elements are characterised by a unique transposase and a consensus sequence at the inverted-repeats (IR) terminal end, at -35 base pairs, that can form a promoter when inserted into another fragment (21). It spreads by replicative transposition, a copy-paste mechanism that allows it to form composite transposons at the site of insertion, rather than moving alone to the new location. Upon insertion into the new region, it allows the formation of a target site duplication (TSD), a duplication of the region adjacent to the insertion site. IS6/IS26 is the insertion sequence most frequently associated with the development of antimicrobial resistance in Gram-negative organisms (21–23). GENE CASSETTES AND INTEGRONS Gene cassettes are small mobile elements (0.5 to 1 kb) consisting of a single gene -occa- sionally twowith no promoter and attC-type recombination sites. Although they can exist in free circular form, they are non-replicative and are usually inserted into an integron (16). Integrons are genetic elements that can capture genes from different sources and are usually defined by the presence of a gene encoding an integrase (intI gene), responsible for the acquisition of genetic material, a primary recombination site (attI) and a promoter (Pc) (24). Integrons can be located on transposons or plasmids, but also on the bacterial chromosome itself. The intI gene encodes a site-specific tyrosine recombinase that catalyses recombination between the attI site of the integron and the attC site of the cassette (16). The amino acid sequences of IntI integrases allow the classification of integrons into 5 different classes (I to V). Integron classes I and III (25) are associated with Tn402 (26,27), whereas class II integrons are almost exclusively linked to Tn7 (28). These three classes are the ones that are historically involved in the spread of antibiotic resistance, and are the most important from a medical point of view. Class I integrons are the most ubiquitous in nature and the most frequently found in clinical isolates (29). They do not have their own mobility, although their close association with the Tn3 integron family and conjugative plasmids suggests that these mobile elements serve as vehicles for them. 32 | Introduction GENOMIC ISLANDS Genomic islands comprise large regions of genomic DNA, more than 10,000 base pairs length, which can be easily exchanged between bacterial isolates. They encode factors that allow transfer, recombination and restriction or modification of proteins, or groups of genes that allow metabolic adaptation or encode virulence or resistance factors (16). Genomic islands encoding virulence factors are known as pathogenicity islands or PAIs (30,31). PAIs are usually flanked by other MGEs and are often inserted near transfer RNA (tRNA)-encoding genes (3). PATHOGENICITY DETERMINANTS A variety of virulence factors have been described in E. coli that allow for an increased ability to adapt to new niches and cause infection. These factors facilitate adhesion, invasion, attachment, iron uptake, motility and toxigenic activity, among others (32). Some of the bacterial surface structures, present in commensal strains, can act as virulence factors in combination with the appropriate cellular receptor: this is the case of the LPS or flagellin, the major protein of the flagellum (3), although most of the pathogenicity determinants are acquired and encoded by genetic elements acquired through MGEs (1,3,11,32). Traditionally, E. coli strains with pathogenic capacity have been classified into pathotypes or pathovars, identified by acronyms, based on the virulence factors contained in each isolate, although the progressive introduction of whole genome sequencing techniques has allowed the in-depth study of the acquisition of pathogenic factors and phylogenetic relationships between different populations (11). CLINICAL PRESENTATION OF E. COLI INFECTIONS E. coli is the most frequently isolated microorganism in patients with bacteraemia (33), is responsible for more than 90% of urinary tract infections (3,34) and is a major cause of intestinal disease worldwide causing gastroenteritis of varying severity (3,34), as well as extraintestinal disease when enteric contents infect other tissues. This PhD thesis has focused on the study of intra-abdomi- nal infections caused by E. coli, as well as the bacteraemia associated with this entity. Introduction | 33 INTRA-ABDOMINAL INFECTIONS The term intra-abdominal infection (IAI) refers to a wide number of processes that affect the abdominal cavity and cause inflammation and infection of the gastrointestinal tract and adjacent tissues. They can be divided into (i) uncomplicated infections, which are those in which the infectious process is limited to the organ or tissue of origin, without anatomical disruption of the gastrointestinal tract (appendicitis, diverticulitis and cholecystitis); and (ii) complicated infections, when the infection extends into the peritoneal cavity, as in diffuse peritonitis or intra-abdominal abscesses. Although most uncomplicated infections can be resolved with simple interventions, delayed or inappropriate implementation and the involvement of more virulent infectious agents increase the risk of progression to complicated infection (35,36). Most intra-abdominal infections occur in the context of inflammation or perforation of the bowel wall and are caused by components of the gastrointestinal microbiome, and are therefore often polymicrobial and mixed infections, involving anaerobic bacteria. The management of IAIs is based on three pillars: focus control strategies, antimicrobial treatment, and supportive measures. Strategies for focus control are all those measures that are taken to eliminate the source of infection and minimise its impact on other tissues. In IAI the four most important strategies are: debridement, removal of exogenous devices, drainage of purulent collections and decompression of the abdominal cavity (37). Inappropriate focus control in IAI increases the 28-day mortality rate from 27% to 43% (38,39). Antimicrobial treatment in IAI serves two purposes: on the one hand, local treatment and, on the other hand, prevention of the spread of infection to surrounding and distant tissues through haematogenous spread. It should be initiated as soon as the diagnosis is suspected and should be implemented within the first hour in case of presentation in the form of severe sepsis (40) (41). The choice of antimicrobial therapy is made according to the most likely focus of infection, the epidemiological situation conditioning the most likely aetiological agent and the immunological status of the patient, as well as the severity of the clinical presentation (40,42). BACTERAEMIA E. coli is the microorganism most frequently involved in cases of bacteraemia (33,43,44), representing between 24% and 30% of all isolates (45), with an incidence that shows an upward 34 | Introduction trend in recent series (45–48). E. coli bacteraemia is considered the haematogenous extension of an infection with a primary focus in another location, the most frequent being the urinary and abdominal foci in first and second place, respectively (43,45,48). From the primary focus of infection, it spreads through the action of several factors including bacterial virulence factors, antimicrobial resistance, suboptimal treatment of the primary focus and insufficient venous access control (44). It is estimated that between 21% and 32% of all bacteraemia by E. coli are originally nosocomial (45). As for bacteraemia of abdominal origin, haematogenous dissemination occurs more frequently in patients with anatomical alterations or obstructions, such as in cases of biliary lithiasis or neoplastic obstructions (45,49). Bacteraemia may also develop as a complication of an abdominal procedure or surgery, often in connection with processes requiring urgent intervention (45,48). PROGNOSIS OF E. COLI INFECTIONS MORTALITY AND MORBIDITY ASSOCIATED WITH INTRA-ABDOMINAL E. COLI INFECTIONS The risk of mortality in IAI is directly related to the extent and severity of organ dysfunction it causes. The overall mortality rate ranges from 7.6% to 29% (50–52). Numerous risk factors for mortality of intra-abdominal E. coli infection have been identified, including: advanced age, poor nutritional status, organ dysfunction, a baseline immunosuppressed state, clinical severity of the IAI or baseline clinical situation as measured by different indices (50,53– 57), delay in initial intervention beyond the first 24 hours, inappropriate focus control, prolonged hospital stay or having received antimicrobial therapy prior to surgery (50,58). Mortality of IAI is also higher in cases where the causative agent is a Gram-negative bacillus producing extended-spectrum ß-lactamase (ESBL) or carbapenemase (50). MORTALITY AND MORBIDITY ASSOCIATED WITH E. COLI BACTERAEMIA Mortality of E. coli bacteraemia is high, with figures varying between 12% and 40% in recent studies. This variability depends on the series, the primary focus of infection and the distribution of risk factors (33,43,45,48,50,59). Introduction | 35 The mortality of E. coli bacteraemia is influenced by multiple factors, including older age, nosocomial acquisition and previous comorbidities; as well as initial presentation as sepsis or septic shock (60), the aetiological agent being antibiotic resistant (43,46), as well as a non-urinary focus (59,61–63). Antibiotic resistance has also been linked to higher rates of treatment failure, longer hospital stays and higher healthcare costs (46). Every hour of delay in administering antimicrobial therapy in abdominal sepsis increases mortality by 2% (64,65), and by 1% for every hour that focus control is delayed. In cases presenting as sepsis, implementation of a focus control strategy within the first 6 hours is associated with a decrease in 90-day mortality of 29% (42). ANTIMICROBIAL TREATMENT OF E. COLI INFECTIONS ANTIBIOTICS WITH ACTIVITY AGAINST E. COLI There are a multitude of options for the treatment of severe infections caused by susceptible E. coli, including ß-lactams, fluoroquinolones, aminoglycosides and trimethoprim-sulfamethoxazole (TMP-SXT), among others. Antibacterial agents whose mechanism of action involves interaction with the bacterial wall or cell membranes are some of the most widely used in clinical practice (66). By acting on this essential structure for bacterial survival, they have traditionally been regarded as bactericidal agents. Among these agents are the ß-lactams. ß -LACTAMS ß-lactam antibiotics have been used safely and effectively for more than 70 years. The presence of the ß-lactam ring chemically defines this family of antibiotics (Figure 5) and is the determinant of their mechanism of action, their low toxicity and their main mechanism of resistance. For the ß-lactam ring to be active, it needs to be bound to other radicals, which modify the properties of the resulting compound and give rise to the different groups of ß-lactams: penicillins, cephalosporins, carbapenems, and monobactams (6). All ß-lactams exert their effect in a similar way through a dual mechanism: inhibition of bacterial wall synthesis and induction of bacterial autolysis by activation of endolysins (68). Penetration 36 | Introduction of ß-lactam antibiotics into Gram-negative bacilli requires passage through porins in the outer cell membrane. The synthesis of the bacterial wall occurs through a complex metabolic pathway that begins with the synthesis of its precursors in the cytoplasm and their assembly and subsequent transport to the outer face of the cytoplasmic membrane, where peptidoglycan formation takes place. Peptidoglycan is the essential structural protein of the bacterial wall, and its skeleton is made up of long glucose chains, formed by the repetition of N-acetylmuramic acid and N-acetylglucosamine molecules. In turn, N-acetylmuramic acid attaches pentapeptide chains containing a terminal dipeptide (D-Alanine- D-Alanine). This terminal dipeptide is incorporated into the forming peptidoglycan structure using specific enzymes that facilitate cross-bridging of the bacterial wall and its mesh-like configuration (6,68,69). The enzyme target for ß-lactams acts at the final stage of peptidoglycan synthesis, when the N-acetylmuramic acid-binding pentapeptides lose one of the amino acid termini through the action of enzymes called penicillin-binding proteins or PBPs (Figure 6). PBPs normally act as bifunctional FIGURE 5. Chemical structure of the main representative ß-lactam antibiotics used in clinical practice. R-groups indicate modifications to the main structure of each group. Adapted from (68). Introduction | 37 enzymes with a combination of transpeptidase and transglycosylase or transpeptidase and carboxypeptidase activity. The ß-lactam ring has structural similarity to the region of the pentapeptide to which PBPs bind, so its covalent attachment to this space prevents PBP binding to the pentapeptide and results in virtual inactivation of these enzymes. This prevents peptidoglycan elongation and, consequently, wall formation (70). FIGURE 6. Mechanism of action of ß-lactams on the bacterial wall. PBP: Penicillin-Binding Proteins. Adapted from (68). 38 | Introduction TABLE 1.Characteristics of ß-lactamase inhibitors of clinical interest. beta-lactamase inhibitor chemical structure beta-lactam with which they are associated Clavulanic acid Amoxicillin Sulbactam Ampicillin Tazobactam Piperacillin Ceftolozane Avibactam Ceftazidime Relebactam Imipenem Vaborbactam Meropenem Nacubactam Meropenem Introduction | 39 Functional alterations caused by PBP-ß-lactam binding trigger an autolysin activation mechanism, which is normally repressed, resulting in cell lysis. Among the ß-lactams there are different activity spectra, from least to most active: penicillins, monobactams and narrow-spectrum cephalosporins, extended-spectrum cephalosporins and carbapenems (70). In Gram-negative bacteria, the main mechanism of resistance to ß-lactams is the production of ß-lactamase enzymes, which act by hydrolysing the ß-lactam ring and thus inactivate the antibiotic prior to binding to PBPs (6). The emergence of ß-lactam-resistant isolates due to ß-lactamase production has progressively compromised the use of narrow-spectrum ß-lactams in monotherapy (69). In response to this growing threat, ß-lactamase inhibitors have emerged: molecules that have reduced antimicrobial activity on their own, but have the ability to hydrolyse ß-lactamases, allowing the ß-lactam to perform its activity when applied in combination (71). The characteristics of the main ß-lactamase inhibitors are summarised in Table 1. Both clavulanic acid, the first inhibitor discovered, and the synthetic derivatives sulbactam and tazobactam share structural similarities with the other ß-lactams and have a similar spectrum of activity (70), with the ability to hydrolyse Ambler class A ß-lactamases. In response to the increasing risk of antimicrobial resistant bacterial infections, new generations of ß-lactamase inhibitors including avibactam and vaborbactam have been added to our arsenal in recent years. The avibactam/ceftazidime combination was approved in 2015 for the treatment of complicated intra-abdominal infection and complicated urinary tract infection. When avibactam was evaluated against many ß-lactamases including TEM, CTX-M, KPC, AmpC (Ambler class C), OXA- 10 and OXA-48 (Ambler class D) it was confirmed a general mechanism of inhibition by avibactam (74,75). Vaborbactam in combination with meropenem was approved in the United States in 2017 for treating complicated urinary tract infections and has been shown activity against KPC-producing Enterobacteriaceae, however did not reduce the MIC of meropenem against bacterial strains expressing metallobetalactamases (Ambler class B) (72,76). PIPERACILLIN-TAZOBACTAM The combination of piperacillin with tazobactam first became available in the United States in 1993. Piperacillin is a broad-spectrum penicillin with bactericidal activity against many Grampositive and Gram-negative microorganisms. Piperacillin, a ureidopenicillin, is derived from ampicillin by the addition of a hydrophilic heterocyclic group to the α-amino group of ampicillin. This 40 | Introduction substituent improves the activity of piperacillin against Gram-negative bacteria, partly due to increased affinity for penicillin-binding protein (PBP)-3, and thus leads to a broader spectrum of activity (77). Piperacillin has been shown to bind quite selectively to PBP-3 in E. coli (78,79). Tazobactam inactivates ß-lactamase enzymes by irreversibly binding to them through the formation of a stable acyl-enzyme complex, that is identical to the acyl-enzyme intermediate formed with the normal ß-lac- tam substrates of these enzymes. The tazobactam-ß-lactamase complex protects piperacillin from the hydrolytic activity of ß-lactamases. Tazobactam increases the spectrum of piperacillin against most ß-lactamase-producing Enterobacterales, both narrow-spectrum and extended-spectrum (71). Piperacillin-tazobactam shows equivalent and sometimes slightly superior efficacy to other broad-spectrum antibiotics (e.g. carbapenems, third generation cephalosporins) for the treatment of pneumonia, skin and soft tissue infections, intra-abdominal infections, polymicrobial infections and febrile neutropenia (80). The usual adult dose is 4 g of piperacillin and 0.5 g of tazobactam every 6-8 hours for patients with creatinine clearance greater than 40 ml/min. ESBL-producing strains of enterobacteria are susceptible in vitro to piperacillin-tazobactam. It has recently been demonstrated that bacteraemia caused by ESBL-producing microorganisms can be effectively treated with piperacillin-tazobactam and that it is non-inferior to the use of carbapenems (81–83). PHARMACOKINETICS AND PHARMACODYNAMICS OF THE ß -LACTAMS Although each ß-lactam has a specific half-life, all have been shown to have time-dependent activity (84): that is, the killing effect against an organism is produced by the amount of time of the drug binds at the binding site. Duration of exposure is often measured as the percentage of the dose interval in which the free drug concentration remains above the pathogen’s minimum inhibitory concentration (MIC) (% fT > MIC). Maximum bactericidal effect is achieved when the free drug concentration is four times the MIC of the pathogen during 40-60% of the dosing interval (85). The optimal target for exceeding the % fT>MIC is variable depending on the specific drug. Maximising the duration of exposure can be achieved in three ways: increasing the dose, shortening the dosing interval or prolonging the infusion time (86,87). The efficacy of extended infusions of piperacillin-tazobactam for the goal of microbiological eradication has been demonstrated in Gram-negative infections (88) and have also been associated Introduction | 41 TABLE 2. Classification of ß-lactamases. Adapted from (69). Ambler class subtype (includes bush-jacoby-medei- ros classification) preferred substrate inhibitor representative examples A 2a (Gram positives) Penicillins Clavulanic acid PC1 2b (Gram negatives) Penicillins 1st generation cephalosporins Clavulanic acid TEM-1 SHV-1 2be (ESBLs) Penicillins Extended-spectrum cephalosporins Aztreonam Clavulanic acid TEM-24 SHV-12 CTX-M-15 2br (IRT) Penicillins Clavulanic acid TEM-30 SHV-10 2c Carbenicillin Clavulanic acid PSE-1 CARB-3 2e (ESBLs) Extended-spectrum cephalosporins Clavulanic acid CepA 2f Carbapenems Avibactam KPC-2 SME-1 B3a (MBL) All except monobactams EDTA Divalent cation chelators IMP-1 VIM-2 NDM-1 C1 (AmpC) Cephalosporins Cloxacillin Avibactam AmpC CMY-2 D 2d Oxacillin Clavulanic acid OXA-1 OXA-10 2de (ESBLs) Extended-spectrum cephalosporins Clavulanic acid OXA-11 OXA-15 2df Carbapenems None OXA-23 OXA-40 OXA-48 The first ß-lactamase was identified in 1940, a “penicillinase” isolated from E. coli (109). In the 1960s, the spread of ampicillin resistance in Gram-negative bacteria was attributed to the emergence of the TEM-1 ß-lactamase, a plasmid ß-lactamase named after the patient from whom the first strain was obtained: Temoniera. TEM-1 is the most frequently found ß-lactamase in Gramnegative bacteria, it has hydrolytic capacity for penicillins and narrow-spectrum cephalosporins in Enterobacterales and it is responsible for more than 90% of ampicillin resistance in E. coli, although it is not active against cephalosporins with an oxyimino side chain, such as cefotaxime, ceftazidime, ceftriaxone or cefepime. The TEM family of ß-lactamases spread worldwide, as did the chromosomally encoded, plasmid-mediated, SHV-type ß-lactamases with a TEM-like molecular structure 48 | Introduction (110). SHV, another common ß-lactamase found primarily in K. pneumoniae and was named from “shulhydryl reagent variable”. Some ß-lactamases arising from point mutations in TEM have acquired the ability to hydrolyse also ß-lactams/ß-lactams inhibitor (BL/BLI): these are called IRTs (inhibitor-resistant TEM), the most widespread mechanism of resistance to ß-lactams/ß-lactams inhibitors (111). EXTENDED SPECTRUM ß -LACTAMASES Extended spectrum ß-lactamases (ESBLs) are ß-lactamases with the ability to hydrolyse penicillins, broad-spectrum cephalosporins and aztreonam (110,112). Shortly after the introduction of cefotaxime, strains of Enterobacterales with cephalosporin resistance derived from SHV and TEM enzymes have emerged (113). TEM-derived ESBLs were discovered in the 1980s in Europe and the USA and arise from amino acid substitutions that broaden the spectrum of activity against oxyimino-ß-lactams. Other ESBLs of great importance due to their wide distribution are those belonging to the CTX-M family, which arise from plasmid acquisition of ß-lactamases originally encoded in chromosomal genes (114). In general, ESBLs are sensitive to ß-lactamase inhibitors such as clavulanic acid, sulbactam or tazobactam. PIPERACILLIN-TAZOBACTAM RESISTANCE IN E. COLI Resistance to piperacillin-tazobactam in E. coli can be conditioned by different mechanisms: (i) TEM-type ß-lactamase-dependent (111); (ii) OXA-type ß-lactamase-dependent (115); or (iii) deficiency in OmpC or OmpF porins (116). The most important mechanisms involved in the emergence of the piperacillin-tazobactam resistance phenotype while maintaining susceptibility to third generation cephalosporins and carbapenems are those dependent on TEM-type ß-lactamases (117,118). While tazobactam has the capacity under normal conditions to inhibit the TEM enzyme (119,120), it has been suggested that hyperproduction of the enzyme allows it to escape the hydrolytic capacity of tazobactam and thus facilitates hydrolysis by piperacillin (121). The evolution of TEM to IRT and other variants with higher hydrolytic capacities would also facilitate the acquisition of the piperacillin-tazobactam resistance phenotype (111,122). In this context, the ESRI (extended spectrum resistant inhibitor) concept arises, defined as the acquisition of high-level resistance to BL/BLI from a low- or moderate-level BL/BLI resistance Introduction | 49 49 | Introduction phenotype, i.e. from phenotypes with resistance to ampicillin-sulbactam and/or amoxicillin-clavula- nate, but susceptibility to piperacillin-tazobactam (123). It has been shown that the ESRI phenomenon is related to the increased hyperproduction of TEM, which consequently leads to an increase in ß-lactamase activity that translates into increases in MIC of piperacillin-tazobactam (123). Mechanisms that allow TEM hyperproduction include mutations in the blaTEM gene promoter, changing a weak (P3) to a strong (P4 or P5) promoter (124) or an upstream point mutation resulting in overlapping sequences and generation of the strong Pa/Pb promoter (125). Another proposed mechanism explaining TEM hyperproduction is the increased copy number of blaTEM, present on the bacterial chromosome or on plasmids, although the mechanism of the duplication is not well described, blaTEM amplification has been localised to a region of approximately 10 kb that may include other resistance genes as well as sequences related to transposition and integration mechanisms (121). Tandem amplification of this region would condition TEM hyperproduction. Another proposed mechanism for blaTEM amplification involves the IS26 insertion sequence, which is often associated with the mobility of antibiotic resistance genes, such as in cases of translocatable unit (TU) regions (117,123). Piperacillin-tazobactam has been shown in vitro to play a role in inducing the emergence of resistance in E. coli strains carrying blaTEM (121,123). EPIDEMIOLOGICAL SITUATION OF PIPERACILLIN-TAZOBACTAM RESISTANCE IN E. COLI IN OUR SETTING In Spain, the rate of piperacillin-tazobactam resistance has increased in recent years. Analysis of a total of 4824 and 2857 E. coli isolates causing intra-abdominal infections in the SMART studies between 2002-2010 (126) and 2011-2015 (127), respectively, showed an increase in the piperacillin-tazobactam resistance rate from 7.75% to 10%. This increase was higher when these isolates were ESBLs producers, being 24.6% and 29.5%, respectively. The increase in piperacillin-tazobactam resistance rates is probably related to antibiotic consumption in some hospitals. For example, in 35 Andalusian hospitals participating in the PIRASOA programme (Programa Integral de prevención y control de las infecciones relacionadas con la asistencia sanitaria y uso apropiado de los antimicrobianos), the annual average piperacillin-tazobactam 50 | Introduction consumption expressed in defined daily doses (DDD) per-1000 stays increased from 48.8% in 2014 to 54.2% in the first three quarters of 2020 (128). Data from PRIOAM (Programa Institucional para la Optimización de los Antimicrobianos) at the Virgen del Rocío University Hospital in Seville showed 9% P/T resistance in E. coli isolates in 2022, excluding urine cultures and epidemiological surveillance cultures; and 5% in E. coli isolates recovered from blood cultures (129). Moreover, in a prospective observational study of patients admitted to the Hospital Virgen del Rocío in Seville, in which 184 had intra-abdominal infections and 247 had bacteraemia caused by E. coli, P/T resistance was observed to be 9.3% and 2%, respectively (130). Forty-six percent (70 out of 153) of patients with IAI received initial treatment with piperacillin-tazobactam and 24.3% (17 out of 70) of them died and/or needed to extend antimicrobial coverage due to poor clinical outcome (130). Therefore, there is accumulating evidence that piperacillin-tazobactam treatment in patients with IAIs caused by E. coli could generate piperacillin-tazobac- tam-resistant strains and, consequently, piperacillin-tazobactam treatment failure. METHODS AND STRATEGIES FOR EARLY DETECTION OF ANTIMICROBIAL RESISTANCE THE IMPACT OF THE USE OF RAPID DIAGNOSTIC TECHNIQUES FOR THE DETECTION OF BACTERIAL RESISTANCE TO ANTIBIOTICS The contribution of different antibiotic resistance mechanisms on poor patient prognosis is probably multifactorial and complex (131,132). These resistance mechanisms have an impact on empirical treatment and often lead to delays in the administration of appropriate treatment. Time to administration of appropriate therapy is an independent predictor of 30-day mortality in patients with bacteraemia (133) and it is recommended starting antibiotic therapy within 24 hours of taking a blood culture (133). Rapid diagnostic tests, in combination with antimicrobial optimisation programs (ASP), have been shown to improve clinical prognosis compared to the use of conventional diagnostic methods alone, as well as benefit in cost-effectiveness studies (134–137). Rapid diagnostic methods not only allow early identification of the infection-causing microorganism, but also allow approximation of antimicrobial susceptibility and aid monitoring of anti-infective therapy (138,139). Although culture-based methods remain the gold-standard for the diagnosis of bacterial infections, they are time-consuming as they usually require 24-72 hours for the identification of Introduction | 51 microorganisms and/or susceptibility studies; therefore, there is a need for the development of methods that allow early identification of antimicrobial resistance profiles. Currently, two types of rapid methods can be found to detect antibiotic resistance: (i) those that detect compounds that indicate bacterial growth or antibiotic degradation; (ii) genetic and/or molecular methods that detect nucleic acid sequences; and (iii) methods that analyse proteomic profiles of resistance genes or products of their expression. While molecular techniques provide rapid and reliable results, they are complex techniques that require trained staff, are costly and detect only a pre-defined number of genes which detection does not necessarily imply expression. It is therefore necessary to create new methods with minimal processing and low cost that allow rapid detection of resistance mechanisms and are easy to implement in clinical microbiology laboratories. STRATEGIES POR EARLY ANTIBIOTIC RESISTANCE DETECTION The clinical microbiology laboratory has been challenged with the difficulty of identifying the presence of antibiotic resistance mechanisms and developing methods for their early identification. These resistance mechanisms can be identified using standard antimicrobial susceptibility testing (phenotypic methods), as in the rapid antimicrobial susceptibility testing (RAST) standardised by the EUCAST (140), or using genotypic methods from colonies, clinical specimens or positive blood cultures. Specific microorganism antigen can be quickly identified from clinical samples through immunoassays. These tests can be completed as point-of-care-procedures in as little as 15 to 30 minutes and have been devised in a number of formats. The immunochromatographic (ICT) test, commonly referred to as the lateral-flow immunoassay, is one of the most widely used formats (141,142). The detection of one or more RNA or DNA sequences unique to a single pathogen is the goal of nucleic acid amplification assays, which are accessible for quick testing as both PCR-based and isothermal nucleic acid amplification procedures. They are more sensitive than ICT, but also require a higher degree of technicality and training. Methicillin, vancomycin, extended-spectrum ß-lactamases and carbapenemase resistance genes can be identified directly from clinical specimens using methods based on nucleic acid amplification testing. Clinical applications for whole-genome sequencing performed directly on specimens are being developed (143,144). 52 | Introduction USING MALDI-TOF MASS SPECTROMETRY FOR ANTIBIOTIC RESISTANCE DETECTION MALDI-TOF (matrix-assisted-laser desorption/ionization and time-of-flight) mass spectrometry is a proteomic technique used in clinical microbiology laboratories to identify microorganisms by analysing ribosomal proteins between 2 and 20 kDa in size from whole cells. The mass spectrum obtained consists of hundreds of thousands of mass-to-charge ratio (m/z) peaks for each bacterial isolate, as well as intensity levels and areas under the curve for each mass peak. Since the introduction of MALDI-TOF technology in laboratories, its use has been proposed for the characterisation of resistance mechanisms based on antibiotic hydrolysis detection methods. The analysis of MALDI-TOF spectra is presented as a simple and rapid procedure with the universal advantages of phenotypic assays but combined with the speed and precision of molecular assays. Some of the approaches that have been put forward include quantitative growth assays (145) and methods based on peak analysis (146–149). For this purpose, different home-made protocols using different combinations of antibiotics, buffers and incubation times have been described, presenting different inconsistencies and discrepancies between the different methods (148,150), making standardisation and widespread application difficult. None of these methods are endorsed by the EUCAST committee. Introduction | 53 the imPAct of the sArs-coV-2 PAndemic on mortAlitY in intrA-AbdominAl e. coli infections THE COVID-19 PANDEMIC The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-associated coronavirus diseases 2019 (COVID-19) emerged in China in December 2019 and spread globally, creating a worldwide pandemic that was declared as such by the WHO on 11th March, 2020. This public health crisis reached historic dimensions (151). The clinical spectrum of COVID-19 varies widely, from asymptomatic disease to pneumonia and life-threatening complications, including acute respiratory distress syndrome (ARDS), multisystem organ failure and ultimately death (152). On 14th March 2020, the Spanish Government activated the state of alarm and national lockdown, limiting the mobility of all citizens except health workers and those working in financial institutions or selling food, pharmaceuticals or basic necessities (153). IMPACT OF COVID-19 PANDEMIC ON OTHER DISEASES The impact of this new disease and the measures taken to mitigate its effects on society and on the performance of the health system are unprecedented in recent history. The COVID-19 pandemic has been an important threat to health due to its direct effects and the difficulties to attend other diseases (154). Significant reduction in the admission of patients with other diseases (cancer, cardiovascular disease or stroke) and delay in their diagnosis has been detected in Spain and in other European, American and Asian countries (155–158). Consequently, higher number of deaths in patients with these diseases have been reported during the COVID-19 pandemic. Although healthcare services continue to function with adaptative measures and patient acceptance, many patients are hesitant to visit hospitals. This hesitance is related to inadequate information regarding COVID-19, particularly with respect to hospital safety measures focused on COVID-19. Therefore, patients should receive timely information about COVID-19 from official sources, through various channels (in example, the internet and social media), regardless of their age and socio-economic status. In response to the suspension of non-urgent medical services brought 54 | Introduction on by the COVID-19 pandemic, appointments have been rescheduled to concentrate on COVID- 19-related treatments, and reducing patient motivation to seek care. Colorectal, cervical lung and prostate cancer screening are among the services that are affected (159). Due to their higher risk of infection, cancer patients are more conscious of the effects of COVID-19. The necessity for frequent visits to medical facilities for treatment and related-care, as well as the immunocompromised condition of many of these patients, all contribute to the risk of infection. It has been described previously in this PhD thesis that the management of patients with intra-abdominal infections is increasingly complicated due to the lack or the optimal therapeutical options, which boost the risk of therapeutic failure and mortality, especially in the patients with the secondary bloodstream infection, when they have developed sepsis or septic shock. The COVID-19 pandemic has hindered the availability and access to the hospital for patients with many pathologies as a result of lack of resources and mobility limitations, making early treatment difficult. Introduction | 55 Hypothesis & Objectives SAMPLES COLLECTION AND PROCESSING All intra-abdominal and blood samples obtained by aseptic procedures were sent immediately after collection to the HUVR microbiology service laboratory. There they were processed according to the usual protocol (Figure 9), the first step of which is a Gram staining of the sample and quantification of the number of epithelial cells and polymorphonuclear cells per 10x field under light microscopy to assess the quality of the sample. FIGURE 9. Diagram of sample processing in the laboratory of Microbiology Department at Virgen del Rocio University Hospital. Intra-abdominal samples cultures For bacterial isolation of the samples obtained, cultures were grown on Columbia agar media with 5% sheep blood and McConkey II agar (Becton Dickinson, NJ, USA), which were incubated under aerobic conditions for 18 hours at 37 ºC. 64 | Materials and Methods Blood cultures At least 10 ml of blood sample was inoculated into blood culture vials (BACTEC™Standard/10Aerobic/F and BACTEC™ Lytic/10Anaerobic/F) (Becton Dickinson, NJ, USA), incubated in the BACTEC FX system (Becton Dickinson, NJ, USA) located in the laboratory of the Microbiology service of the HUVR. Identification and selection of clinical isolates of E. coli Isolates were identified by matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry using a MALDI biotyper (Bruker Daltonics GmbH, Germany). The P/T-resistant isolate obtained from an intra-abdominal sample from each patient was selected for the experiments belonging to the study. In those patients who had a sensitive sample prior to P/T treatment, it was also selected for analysis. Antimicrobial susceptibility testing An antibiogram was performed using the semi-automated MicroScan Walkaway microdilution system (Beckman Coulter Inc., CA, USA), using the NM51 and NMR-1 panels for gram-ne- gatives. The antibiotics tested were ampicillin, amoxicillin/clavulanic acid, piperacillin/tazobactam, cefoxitin, cefotaxime, ceftazidime, cefepime, ertapenem, imipenem, meropenem, ciprofloxacin, levofloxacin, gentamicin, tobramycin, amikacin, tigecycline, trimethoprim-sulfamethoxazole, ceftazidime-avibactam, ceftolozane-tazobactam and colistin. For the definition of susceptibility and resistance, the criteria defined by the European Committee of Antimicrobial Susceptibility Testing (EUCAST), in its guidelines for the years 2020 and 2021 (160,161), were used. The minimum inhibitory concentration (MIC) for P/T was confirmed by broth microdilution method, according to the criteria established by the EUCAST. To this end, 96-well U-bottom plates were used, containing 50 µl of Mueller-Hinton broth (MHB) (Sigma-Aldrich Inc., MO, USA) in the first column of the plate. Subsequently, 1:2 serial dilutions were made from the first column of wells. Thus, a total of 10 wells were used for each sample containing decreasing concentrations of antimicrobial in the range of 256 mg/l to 0.5 mg/l concentration. Finally, 50 µl of inoculum of each E. coli strain, obtained after incubation in MHB under aerobic conditions for 24 hours at 37 ºC, was added until a turbidity of 0.5 on the McFarland scale was achieved, diluted 1:100 (final concentration Materials and Methods | 65 of 106 colony forming unit (CFU)/ml). Growth controls were performed with bacterial inoculum but without antibiotic. Sterility control without antibiotic and without inoculum were also performed. The control strain used was the reference strain E. coli ATCC 25922. All plates were incubated in aerobic conditions at 37°C for 24 hours. After this period, the plates were read, and the MIC was defined as the lowest concentration at which the broth showed no macroscopically visible turbidity. TO CHARACTERISE THE CLONAL RELATIONSHIP BETWEEN P/T- RESISTANT AND P/T-SUSCEPTIBLE CLINICAL ISOLATES To analyse the clonal relationship between the isolates, the multilocus sequence typing technique (MLST) was used according to the Atchman protocol (162). This method uses seven constituti- ve genes of the bacterium that allow isolates to be separated into groups with identical allelic profiles or sequences. DNA EXTRACTION Each isolate was seeded on a Columbia agar plate with 5% sheep blood and cultured under aerobic conditions for 18 hours at 37°C to isolate individual colonies. After incubation, four to five colonies from each plate were suspended in 50 µl of sterile distilled water in 1.5 ml Eppendorf® tubes. To produce cell lysis and DNA collection, these tubes were heated at 95°C for 5 minutes in thermoblock and then cooled directly on ice. After this process, they were centrifuged for 60 seconds at 10,000 revolutions per minute (rpm) in a microcentrifuge Sorvall Legend Micro 21R (Themo Scientific, CA, USA) in order to deposit the cellular detritus and separate them from the extracted DNA suspension. These fragments were kept on ice for immediate use. AMPLIFICATION Seven E. coli reference genes: adk, fumC, gyrB, icd, mdh, purA and recA were amplified using specific primers (Thermo Scientific Inc., MA, USA), described in Table 3. For amplification, a sample was made including: 5 µl of DNA extract, 10 µl of buffer (10X Buffer, BioVision Inc, CA, USA [containing: diluted 1:10 KCl at 10 mM, Tris-Cl at 20 mM, MgSO4 at 2 mM (NH4)2SO4 at 10 66 | Materials and Methods mM, Triton X-100 at 1%, BSA at 1 mg/ml], 2.5 µl Forward primer, 2.5 µl Reverse primer, 0.5 µl Taq-polymerase (Taq DNA Polymerase, BioVision Inc, CA, USA) and Milli-Q water to complete 50 µl. In all runs, a negative control was performed to detect possible contamination of the reagents from a mix containing all the components referred, except for the DNA extract, which was replaced by 5 µl of Milli-Q water. The amplification was performed on a 2720 Thermal Cycler (Applied Biosystems, Thermo Fisher Scientific, MA, USA) with the following thermal profile: an initial denaturation step for two minutes at 95 ºC; 30 amplification cycles, consisting of: one minute at 95 ºC, one minute at the corresponding alignment temperature for each primer (54 ºC for adk, fumC, icd and purA, 60 ºC for gyrB and mdh and 58 ºC for recA) and two minutes at 72 ºC; and finally, a final five minutes at 72 ºC. TABLE 3.Specific primers for the amplification of the E. coli reference genes Primer Sequence (5’- 3’) adk-F ATTCTGCTTGGCGCTCCGGG adk-R CCGTCAACTTTCGCGTATTT fumC-F TCACAGGTCGCCAGCGCTTC fumC-R GTACGCAGCGAAAAAGATTC gyrB-F TCGGCGACACGGATGACGGC gyrB-R ATCAGGCCTTCACGCGCATC icd-F ATGGAAAGTAAAGTAGTTGTTCCGGCACA icd-R GGACGCAGCAGGATCTGTT mdh-F ATGAAAGTCGCAGTCCTCGGCGCTGCTGGCGG mdh-R TTAACGAACTCCTGCCCCAGAGCGATATCTTTCTT purA-F CGCGCTGATGAAAGAGATGA purA-R CATACGGTAAGCCACGCAGA recA-F CGCATTCGCTTTACCCTGACC recA-R TCGTCGAAATCTACGGACCGGA F: forward; R: reverse. Materials and Methods | 67 ELECTROPHORESIS After amplification, 1% agarose gel electrophoresis was performed (1 g D1 Low EEO agarose [Conda-Pronadisa, Condalab, Spain] and 100 ml of 1x buffer (Ultrapure 10x TAE Buffer, ThermoFisher Scientific Inc, MA, USA), containing 5 µl of staining solution (RedSafe™ Nuclei Acid Staining Solution [20,000x, iNtRON Biotechnology Inc., South Korea]) per 100 ml of gel. For this purpose, a 10 µl aliquot of each sample, including the negative control, was deposited. The molecular weight marker used was HypperLadder™ (Bioline, GmbH, Germany). The electrophoresis result of the gels was visualised using a UVP 2UV™ Benchtop Transilluminator (UVP Inc., CA, USA). DNA PURIFICATION DNA purification was performed from the electrophoresis gel following the protocol of the E.Z.N.A.® Gel Extraction kit (Omega Bio-tek Inc., GA, USA). For this, the fragment of interest was excised from the electrophoresis gel using a scalpel blade and placed in an Eppendorf® tube, to which 700 µl of the XP2 Binding Buffer reagent provided by the kit was added. This mix was incubated at 60°C for 7 minutes in a thermoblock and then deposited on a HiBind® DNA Mini Column inserted into a 2 ml collection tube. It was centrifuged at 10,000 x g for 1 minute at room temperature and the supernatant was discarded. 300 µl of XP2 Binding Buffer Reagent was added and centrifuged at 14,000 x g for 1 minute at room temperature and the supernatant was discarded. 700 µl of SPW Buffer Reagent, previously diluted with 100% ethanol, was added and centrifuged at maximum speed for one minute and the supernatant was discarded; this step was repeated for a second time. The empty column was centrifuged at full speed for two minutes to dry the column and remove any remaining ethanol that could interfere with the subsequent sequencing reaction. Each column was then transferred to a new Eppendorf® tube and 50 µl of elution buffer provided by the kit was added directly to the centre of the column matrix and allowed to stand at room temperature for two minutes. The eluate was then centrifuged at maximum speed for one minute to recover the eluate. SEQUENCING REACTION AND ANALYSIS Sequences were analysed at the Genomics and Sequencing service of the Instituto de Biomedicina de Sevilla (IBiS), equipped with the Sanger ABI3500 Genetic 121 Analyzer sequencer (Applied biosystems, Foster, VA, USA). Sequences were analysed using MySequence software 68 | Materials and Methods version 1.2.14 for MacOS (telethon Co., Ltd., Japan) and the BLAST (Basic Local Alignment Search Tool) sequence alignment web services hosted at http://www.ncbi.nlm.nih.gov/BLAST. The data were compared with those hosted on the Enterobase website (hosted at http://enterobase.warwick. ac.uk), where the strains were classified into different sequence types (ST) according to the reference MLST sequences. TO STUDY THE MECHANISM UNDERLYING THE ACQUISITION OF RESISTANCE TO P/T DETECTION OF BETA-LACTAMASES The presence of the ß-lactamases blaTEM, blaOXA-1, blaSHV and blaCTX-M in the clinical study samples was studied by conventional PCR with specific primers (Thermo Scientific Inc., MA, USA) as shown in Table 4. Bacterial DNA extraction was performed by heat, following the technique previously described on Page 48. For amplification, a mix was made including: 5 µl of the DNA extract, 10 µl of buffer (10X Buffer, BioVision Inc, CA, USA [containing: diluted 1:10 KCl at 10 mM, Tris-Cl at 20 mM, MgSO4 at 2 mM (NH4)2SO4 at 10 mM, Triton X-100 at 1%, BSA at 1 mg/ml]), 2.5 µl of the suspension of each of the primers, 0.4 µl of Taq polymerase (Taq DNA Polymerase, BioVision Inc., CA, USA) and Milli-Q water to complete 50 µl. In all PCR runs, a negative control was performed to detect possible contamination of the reagents from a mixture containing all the components of the mixture, except for the DNA extract, which was replaced by 5 µl of Milli-Q water. Amplification was performed on a Thermo Cycler 2720 Thermo Cycler (Applied biosystems, Thermo Fisher Scientific Inc, CA, USA) with the following amplification conditions: a four-minute denaturation cycle at 95°C; followed by 35 amplification cycles of 30 seconds at 95°C, 30 seconds at 58°C for blaTEM, 46°C for blaOXA-1, 62°C for blaSHV and 50°C for blaCTX-M, and one minute and 15 seconds at 72°C; and a final extension cycle of seven minutes at 7°C. Gene amplification was checked by agarose gel electrophoresis (1%) and the result was visualised according to the protocol described previously. The reference strain of E. coli ATCC 25922 was used as a control. Materials and Methods | 69 TABLE 4. Specific primers used for ß-lactamase gene detection. primer sequence (5’ - 3’) blaTEM-F TTCTTGAAGACGAAAGGGC blaTEM-R ACGCTCAGTGGAACGAAAAC blaOXA-1-F GGATAAAACCCCCAAAGGAA blaOXA-1-R TGCACCAGTTTTCCCATACA blaSHV-F GGGTTATTCTTATTTGTCGC blaSHV-R TTAGCGTTGCCAGTGCTC blaCTX-M-F ATGTGCAGYACCAGTAARGT blaCTX-M-R TGGGTRAARTARGTSACCAGA F: forward; R: reverse. WHOLE-GENOME SEQUENCING Whole-genome sequencing (WGS) was performed with two different techniques: longread Oxford Nanopore® technology and short-read Illumina® sequencing to obtain a consensus sequence. Oxford Nanopore Technologies BACTERIAL DNA EXTRACTION Bacterial cultures were incubated on Columbia agar plates with 5% sheep blood under aerobic conditions at 37 °C for 18 hours to obtain isolated colonies. Four to five colonies were suspended in LB broth (Conda-Pronadisa, Condalab, Spain) and incubated under aerobic conditions at 37 °C for 18 hours with continuous shaking at 180 rpm. One millilitre of the bacterial culture was taken and then centrifuged at 5000 x g for 10 minutes. An automated MagCore® HF16 Plus System extractor (RBC Bioscience Corp., Zhonghe, Taiwan) and MagCore® Genomic DNA Large Volume Whole Blood Cartridge Kit Cartridges Coded 102 were used for extraction and purification of bacterial DNA into 20-30 kb fragments. Two-hundred microliters of each sample were taken in a 1.5 ml Eppendorf® 70 | Materials and Methods tube in which 4 µl RNase A at a concentration of 50 mg/ml was added and incubated at room temperature for 20 min. Then, 20 µl of Proteinase K at a concentration of 10 mg/ml was added. This mixture was introduced into the cartridge according to the manufacturer’s instructions and the autoextractor was programmed to perform the cartridge numbered 102 protocol. LIBRARY PREPARATION For genomic DNA end repair, 3 µg of genomic DNA was taken and transferred into an Eppendorf® DNA LoBind tube and then adjusted to 48 µl with Milli-Q water. In a 0.2 ml PCR tube, a mix was prepared containing 48 µl genomic DNA, 3.5 µl of NEBNext FFPE DNA Repair Buffer (New England BioLabs, Inc., MA, USA), 3.5 µl of Ultra II End-prep reaction buffer (New England BioLabs, Inc, MA, USA), 3 µl of Ultra II End-prep enzyme mix (New England BioLabs, Inc., MA, USA) and 2 µl of NEBNext FFPE DNA Repair Mix (New England BioLabs, Inc., MA, USA). This mix was incubated at 20°C for 5 minutes and at 65°C for 5 minutes in a 2720 Thermo Cycler (Applied biosystems, Thermo Fisher Scientific Inc., CA, USA), leaving the lid at room temperature. The incubation product was then transferred to a new 1.5 ml Eppendorf® DNA LoBind tube and 60 µl of AMPure XP magnetic beads (Beckman Coulter Inc., CA, USA) were added. This mixture was incubated for 5 min under gentle agitation at room temperature. A magnet was used to separate the magnetic beads and the supernatant was discarded. Then, 200 µl of 70% ethanol was added without resuspending the pellet. The pellet was again separated with the magnet and then the supernatant was discarded. This step was repeated at a second time. The ethanol residue was removed and allowed to dry for 30 seconds. The pellet was resuspended in 25 µl Milli-Q water and incubated at room temperature. The 25 µl of eluate was transferred to a new 1.5 ml Eppendorf® DNA LoBind tube. The Native Barcoding Expansion 1 - 12 EXP-NBD104 kit (Oxford Nanopore Technologies Limited, Oxford, UK) was used to bind the barcodes. One barcode was selected for each sample. 750 ng of each sample was taken and then diluted in 22.5 µl of Milli-Q water. A mixture containing 25 µl ligase master mix (Blunt/TA Ligase Master Mix, New England BioLabs, Inc., MA, USA), 2.5 µl barcoding (NBD104 Native Barcoding Expansion, Oxford Nanopore Technologies Limited, Oxford, UK) and 22.5 µl of sample was mixed to give a final volume of 50 µl. This mixture was incubated for 10 minutes at room temperature and then, 50 µl of AMPure XP magnetic beads were added. This mixture was incubated for 5 minutes under gentle agitation at room temperature. A magnet was used to separate the magnetic beads and the supernatant was discarded. Then, 200 µl of 70% ethanol was Materials and Methods | 71 added without resuspending the pellet. The pellet was separated again with the magnet and the supernatant was discarded. This ethanol wash step was repeated a second time. The ethanol residue was removed and allowed to dry for 30 seconds. The pellet was resuspended in 26 µl Milli-Q water and incubated at room temperature. Equimolar amounts of each sample were collected to a pool of 1050 ng total in a 1.5 ml Eppendorf® DNA LoBind tube. This mixture was diluted with Milli-Q water to a final volume of 65 µl. For adapter binding, a mixture was made including 65 µl of the barcoded-sample pool, 5 µl of Adapter Mix II (New England BioLabs, Inc., MA, USA), 20 µl of 5X NEBNext Quick Ligation Reaction Buffer (New England BioLabs, Inc, MA, USA) and 10 µl of Quick T4 DNA Ligase (New England BioLabs, Inc., MA, USA) for a final volume of 100 µl, which was incubated at room temperature for 10 min. To this mixture, 50 µl of AMPure XP magnetic beads were added and then incubated for 5 minutes under gentle agitation at room temperature. A magnet was used to separate the magnetic beads and the supernatant was discarded. The pellet was resuspended in 250 µl of Long Fragment Buffer [LFB] (Oxford Nanopore Technologies Limited, Oxford, UK.). A magnet was used to separate the magnetic beads and the supernatant was discarded. This step was repeated at a second time. The pellet was resuspended in 15 µl of Elution Buffer [EB] (Oxford Nanopore Technologies Limited, Oxford, UK.) and incubated for 10 minutes at room temperature. The magnetic beads were separated using the magnet and then, the 15 µl of the eluate was transferred to a new 1.5 ml Eppendorf® DNA LoBind tube. SEQUENCING REACTION A MinION™ Mk1B device (Oxford Nanopore Technologies Limited, Oxford, UK) equipped with an R9.4 Flow Cell was used to perform the sequencing reaction. First, the mixture composed of 30 µl of Flush Tether [FLT] (Oxford Nanopore Technologies Limited, Oxford, UK) and 970 µl of Flush Buffer [FB] (Oxford Nanopore Technologies Limited, Oxford, UK) was prepared and then added. The library was prepared in a new tube, with a mixture containing: 37.5 µl of Sequencing Buffer [SQB] (Oxford Nanopore Technologies Limited, Oxford, UK), 25.5 µl of Loading Beads [LB] (Oxford Nanopore Technologies Limited, Oxford, UK) and 12 µl of the DNA library, for a final volume of 75 µl, which were loaded into the MinION Flow Cell (R9.4). 72 | Materials and Methods ASSEMBLY Basecalling were performed in a GPU cluster with four Tesla v100 using Guppy 5.0.16 in high accuracy mode to obtain the Nanopore reads. Then and to perform a genomic analysis, nanopore long reads were assembled using flye 2.9. Illumina® sequencing BACTERIAL DNA EXTRACTION Bacterial cultures were incubated on MHB agar plates (Bio-Rad Laboratories Inc., CA, USA) under aerobic conditions at 37°C for 18 hours to obtain isolated colonies. Four to five colonies were suspended in BHI-T broth (bioMérieux, Marcy-l’Etoile, France) and incubated under aerobic conditions at 37°C for 18 hours with continuous shaking at 180 rpm. One milliliter of the bacterial culture was taken and centrifuged at 5000 x g for 10 minutes. For DNA extraction and purification, the protocol specified by the PureLink™ Genomic DNA Mini Kit (Thermo Fisher Scientific Inc., MA, USA) was used. The bacterial pellet was resuspended in 180 ml of digestion buffer, and 20 µl of Proteinase K provided by the kit was added. The mixture was incubated at 55 °C for 45 minutes in a water bath. Next, 20 µl of RNAase was added and incubated for 2 minutes at room temperature. Next, 200 µl of Binding Buffer provided by the kit was added and mixed until a homogeneous solution was obtained. To this solution, 200 µl of 100% ethanol was added. Each sample was transferred to a GeneJET Genomic DNA Purification Column provided in the kit, inserted into a collection tube, and centrifuged at 6000 x g for 1 minute. The column was transferred to a new collecting tube and 500 µl of Wash Buffer I previously dissolved in 100% ethanol was added. The column was centrifuged again at 8000 x g for 1 minute and then transferred to a new collecting tube. 500 µl of Wash Buffer II, previously dissolved in 100% ethanol was added and centrifuged at 12000 x g for 1 minute. The column was then transferred to a new collecting tube and centrifuged at 12000 x g for three minutes. The column was then transferred to an empty 1.5 ml Eppendorf® tube and 35 µl of elution buffer was added. It was incubated at room temperature for 2 minutes and centrifuged for 1 minute at 12000 x g at room temperature. The concentration of genomic DNA obtained was measured using a Nanodrop spectrophotometer (Thermo Fisher Scientific Inc., CA, USA), measuring the A260/A280 and A260/ A230 ratios as well as the amount of DNA recovered expressed in ng/µL. A minimum genomic DNA concentration of 20 ng/µl was considered for library preparation. Materials and Methods | 73 The mRNA expression for the ompC gene was normalised to that of the constitutive rrsG gene, whose primers are specified in Table 7. The normalised ompC expression of each strain was calibrated against the expression of the E. coli ATCC 25922 strain with the ∆∆CT method and using the formula: 2−ΔΔCT − ΔΔCT =ΔC T(ATCC 25922)− ΔC T (Study strain)2−∆∆CT − ∆∆CT =∆C T(ATCC 25922)− ∆C T (Study strain). TABLE 7. Specific primers used for real-time detection of ompC. Primer Sequence (5’—3’) ompC-F CAGGATGTGGGTTCTTTCG ompC-R GAAGTCAGTGTTACGGTAG rrsG-F TGCATCTGATACTGGCAAGC rrsG-R ACCTGAGCGTCAGTCTTCGT F: forward; R: reverse. TO DEMONSTRATE THAT THE ACQUISITION OF P/T RESISTANCE IS DUE TO blaTEM ACTIVITY IN E. COLI ANALYSIS OF BETA-LACTAMASE ACTIVITY The study of ß-lactamase activity by colorimetric assay was carried out in duplicate for each isolate, using the protocol of the ß-lactamase Activity Colorimetric Assay Kit (Bio-Vision Inc., CA, USA) as previously published by our research group (123). First, bacterial cultures were incubated on 5% Columbia agar plates with 5% sheep blood under aerobic conditions at 37°C for 18 hours to obtain isolated colonies. Three to four colonies were then resuspended in 1 ml of physiological saline 80 | Materials and Methods and centrifuged at 5000 x g for 20 minutes. The bacterial culture pellet obtained after centrifugation was weighed and resuspended in the ß-lactamase assay buffer reagent, provided in the kit, to obtain a final concentration of 1 mg pellet in 5 µl of buffer. The samples (1 mg pellet in 5 µl buffer) were sonicated to produce bacterial wall lysis in a Vibra-CellTM75115 sonicator (Thermo Fisher Scientific, MA, USA) for five minutes and centrifuged at 16,000 x g at 4 °C for twenty minutes. Five µl of each sample was dispensed into each well of a 96-well flat-bottom plate and the nitrocefin and ß-lactamase assay buffer reagents included in the kit were added to a final volume of 100 µl. The optical density was immediately measured at 490 nm in kinetic measurement mode from the start time to a final time of 30 minutes, consecutively every five minutes on a CLARIOSTAR Plus multimodal reader (BGM LABTECH, Ortenberg, Germany). A standard curve was generated using different concentrations of nitrocefin and the specific ß-lactamase activity of each sample was expressed in mU/mg protein. The reference strain of E. coli ATCC 25922 was used as a control. ANALYSIS OF blaTEM ACTIVITY Expression of the blaTEM gene by real-time RT-PCR was analysed in a selection of isolates, following the technique previously described for the analysis of ompC transcription. The mRNA expression for the blaTEM gene was normalised to that of the constitutive rrsG gene, whose primers are specified in Table 8. The normalised blaTEM expression of each strain was calibrated against the expression of the E. coli ATCC 25922 strain with the ∆∆CT method and using the formula: 2−ΔΔCT − ΔΔCT =ΔC T(ATCC 25922)− ΔC T (Study strain)2−∆∆CT − ∆∆CT =∆C T(ATCC 25922)− ∆C T (Study strain). TABLE 8. Specific primers used for real-time detection of blaTEM. Primer Sequence (5’—3’) blaTEM-F GAGGACCGAAGGAGCTAACC blaTEM-R TTGCCGGGAAGCTAGAGTAA rrsG-F TGCATCTGATACTGGCAAGC rrsG-R ACCTGAGCGTCAGTCTTCGT F: forward; R: reverse. Materials and Methods | 81 TO DETERMINE WHETHER P/T-INDUCED RESISTANCE ACQUISITION IS STABLE IN E. COLI IN VITRO PRESSURE STUDY WITH SUBINHIBITORY CONCENTRATIONS OF P/T Susceptible E. coli strains collected prior to P/T treatment were selected for pressure analysis with sub-inhibitory P/T concentrations as previously published by our research group (123). Bacterial cultures were incubated on Columbia agar with 5% sheep blood plates under aerobic conditions at 37°C for 18 hours to obtain isolated colonies. Four to five colonies were taken from each isolate and resuspended in MHB containing a fixed tazobactam concentration of 4 mg/l and a piperacillin concentration at half its MIC. This suspension was incubated under aerobic conditions at 37°C for 24 hours under continuous agitation at 180 rpm. Ten microliters of this culture was taken and added to 10 ml of MHB containing a fixed concentration of 4 mg/l tazobactam and a concentration of piperacillin at its original MIC. In turn a growth control was performed by taking 10 µl of the culture and seeding it on a Columbia agar with 5% sheep blood plate. This process was repeated until the maximum concentration of piperacillin that allowed bacterial growth was reached on the control plate or until a piperacillin concentration of 256 mg/l was reached. REMOVAL OF ANTIMICROBIAL PRESSURE Bacterial cultures were incubated on P/T free Columbia agar with 5% sheep blood plates under aerobic conditions at 37°C for 18 hours to obtain growth of isolated colonies, from which successive daily subcultures were obtained for 15 days. The MIC at P/T was determined on the original clinical isolates and on days 5, 10 and 15 according to the method previously described. GROWTH RATE ASSAY To determine the relative fitness between the P/T-susceptible and P/T-resistant isolate, fresh colonies were precultured and incubated on broth during 2-3 hours with shaking at 37 ºC to reach the exponential phase. Samples were tested in LB broth (Sigma-Aldrich Inc., MO, USA) and MH broth (Sigma-Aldrich Inc., MO, USA). After incubation, each culture was adjusted until a turbidity of 0.5 on the McFarland scale was reached, and then diluted 1:100 in fresh broth to reach a final concentration of 105 - 106 CFU/ml. Then, 200 µl aliquots were transferred into 3 separate wells of a flat-bottom 82 | Materials and Methods 96-well microtiter plate. The growth rate was measured for 12 hours with reads at 20-minutes intervals using a CLARIOSTAR Plus (BGM LABTECH, Ortenberg, Germany) densitometer (optical density at 600 nm [OD600]). The doubling time (Td) was determined using growth rate (amount of doubling in one unit of time): Were N(t) being the number of cells at time t; N0 being the number of cells at time 0; r being the growth rate and t being the time. The equation simplifies to: STATISTICAL ANALYSIS The group data are presented as mean ± standard error of the mean (SEM). Student’s t-test was used to determine the differences between means. The difference was considered significant at P < 0.05. All the statistical analyses were performed by using the SPSS software package (version 25.0, IBM Corporation, Armonk, New York, USA) and R Statistical Software (R Core Team [2020]. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria). Materials and Methods | 83 OBJECTIVE 2 TO DEVELOP AND TO VALIDATE A SYSTEM FOR RAPID DETECTION OF HIGHLEVEL RESISTANCE TO P/T AND OTHER SS-LACTAMS IN COMBINATION WITH SS-LACTAMASE INHIBITORS (ESRI) IN CLINICAL ISOLATES OF E. COLI BACTERIAL STRAINS COLLECTION AND CULTURE CONDITIONS A collection of non-duplicated E. coli isolates was used, obtained from bloodstream and intraabdominal infections at HUVR of Seville, A Coruña and Son Espases of Mallorca from 2016 to 2021. Bacterial isolates were stored at –80 ºC using cryopreservation vials (Deltalab, Barcelona, Spain). Bacterial isolates were seeded on a blood agar with a sterile loop before incubation in aerobic atmosphere at 37 ºC for 18 hours. TO CHARACTERISE E. COLI ISOLATES, PHENOTYPICALLY AND GENOTYPICALLY, BASED ON THEIR SUSCEPTIBILITY TO P/T P/T SUSCEPTIBILITY ASSAY The P/T susceptibility profile was initially tested by broth microdilution using MicroScan Walk Away NM44 panels (Beckman Coulter, Inc., USA). MICs of P/T were subsequently confirmed using the standard broth microdilution method, as previously described in Objective 1. MICs were determined according to the EUCAST standards being susceptible ≤ 8 mg/l, resistant greater than 8 mg/l and in the area of technical uncertainty (ATU) equals to 16 mg/l. MLST ASSAY Seven housekeeping genes (adk, fumC, gyrB, icd, mdh, purA and recA) were amplified by conventional PCR and analysed as previously described in Objective 1. 84 | Materials and Methods TO ANALYSE THE CAPACITY TO ACQUIRE HIGH-LEVEL RESISTANCE TO BL/IBL (ESRI) The antimicrobial selection pressure experiment was performed as described previously in Objetive 1. Overnight bacterial cultures grown in 10 ml of Mueller-Hinton broth (MHB) at 37 ºC were adjusted to 0.5 standard (108 CFU/ml). The diluted inoculum were incubated with subinhibitory concentrations of P/T (8:1 ratio) corresponding to 1-fold dilutions below the MICs at 37 ºC for 24 hours. Positive bacterial growth was readjusted to 105 CFU/ml, and adjusted cultures were incubated with a 2-fold-increased concentration of P/T. These steps were repeated until reaching a P/T concentration of 256/32 mg/l or until a P/T concentration that did not allow bacterial growth was reached. At the end of the process, MICs were determined for all the pressured isolates. TO DETECT THE blaTEM, blaOXA-1 AND blaSHV GENES IN THESE ISOLATES The blaTEM, blaOXA-1 and blaSHV genes were determined and analysed by PCR, respectively, as we previously described in. TO EVALUATE THE ABILITY OF AN AUTOMATED SOFTWARE FOR ANALYSING MALDI-TOF MS SPECTRA TO DETECT P/T RESISTANCE SPECTRUM ACQUISITION Bacterial strains were cultured on Columbia agar with 5% sheep blood. A single colony was smeared as a thin film directly onto a MALDI-TOF MS target (MBT Biotargets-96, Bruker Daltonik, Germany) as a homogeneous distribution of material. After the sample was dried, 1 µl of a a-cyano-4-hydorxycinnamic acid matrix solution (Bruker Daltonik, Germany) was resuspended in the standard solution (acetonitrile 50%, water 47,5% and trifluoroacetic acid 2,5%) and overlayed in each sample position. The bacterial solution and matrix were mixed directly on the target by pipetting, and the mix was air dried for at least 2 min. MALDI-TOF MS analysis was performed on a Bruker MicroFlex LT mass spectrometer (Bruker Daltonics Inc., GmbH, Germany). Spectra were acquired over a mass/ charge (m/z) ratio ranging from 2,000 to 20,000. Each spot was measured using 120 laser shots at 60 Materials and Methods | 85 Hz. FlexAnalysis and MBT Compass software (v4.1; Bruker Daltonik, Germany) were used in the MS data analysis. The external calibration was performed following the Bruker’s recommendation by including the bacterial test standard (BTS; Bruker Daltonik, Germany) in each run before spectral acquisition. PEAK ANALYSIS For the classification of the P/T-susceptible and resistant isolates and ESRI developer isolates, their protein spectra were processed using Clover MS Data Analysis software (Clover Biosoft, Spain), as summarized in Table 9. The spectrum features from isolates were analysed by peak matrices built in the range of 2000 m/z to 20000 m/z. The spectra were first preprocessed to remove the noise by applying a Savitzky-Golay filter as smoothing and a top-hat filter for baseline substruction. A spectral alignment with a replicate union process was then performed as follows: 1) shot spectra within the same spot were aligned to create an average spot spectrum per target position; 2) these average spectra were aligned with those from other spots of the same sample, and thus, one average spectrum per sample was obtained; and 3) average spectra from different isolates were then aligned together. Once the data were normalized in this step by total ion current (TIC) normalization, two different methods were tested for building the peak matrices. In the full spectrum method, peak matrices were built with all intensities of the entire spectra. On the other hand, using the threshold method, the peak matrices were built from all peaks with intensities higher than 1.0% of the maximum peak intensity (0.01 factor) of each spectrum. For the classification of the isolates into selected categories, the peak matrices were then used as input data for three different machine learning supervised algorithms: partial least squares discriminant analysis (PLS-DA), linear support vector machines (SVMs) and random forest (RF). Stratified k-fold cross validation was used as internal validation, in which the dataset was randomly split into (k) balanced folds. From the folds obtained, k-1 were used for training the algorithm, and the remaining values were used as a blind set. This process was iterated k times, so each fold was tested in a blinded manner. Furthermore, this k-fold was repeated 20 times to obtain the average score and its relative error. 86 | Materials and Methods TABLE 9. Parameters applied for peak matrices built in each assay and method. Process Parameters applied Baseline Substraction Tophat filter Factor: 0.02 Smoothing Sviztky-Golay filter Window length: 11 Polynomial order: 3 Alignment Allowed shift: Medium Mass tolerance for aligning the spectra: Constant tolerance: 3 Da Linear Mass tolerance: 900 ppm Normalisation Total Ion Count (TIC) Peaks a) Full Spectrum. Skip this step b) By threshold: peak threshold: 0.01 Da Merge a) Full Spectrum Method: constant tolerance: 0.5 Da b) Threshold method: - Contant tolerance: 1 Da. - Linear Mass tolerance: 300 ppm This classification assay was first applied to discriminate pairs of ESRI developers susceptible to P/T (original isolates) from those that were pressed in vitro to acquire P/T resistance (pressed isolates), as a previous study to test the potential of the approach described above. The main assay was then divided into two steps as follows: 1) discrimination of the P/T-susceptible and ESRI developer isolates from P/T-resistant isolates and 2) discrimination of the P/T-susceptible isolates from ESRI developer isolates. Materials and Methods | 87 OBJECTIVE 3 TO ANALYSE THE IMPACT OF THE SARS-COV-2 PANDEMIC ON THE MORBIDITY AND MORTALITY OF INTRA-ABDOMINAL E. COLI INFECTION TO IDENTIFY COHORTS OF INTRA-ABDOMINAL E. COLI INFECTION REQUIRING HOSPITAL ADMISSION, IN PRE-PANDEMIC AND PANDEMIC PERIODS DESIGN This is a post-hoc analysis of the two prospective observational cohorts of hospital-based patients. The study was approved by the Ethics Committee of the Virgen Macarena and Virgen del Rocio University Hospitals (approval no. 0023-N-16 and 0088-N-20). SCOPE AND STUDY POPULATION. The study population included all adult patients admitted to the HUVR who had intra-ab- dominal infection confirmed by isolation of E. coli from an intra-abdominal sample or blood culture. Only the first IAI episode was included in this study during the admissions of the patient. STUDY PERIOD All patients fulfilling the inclusion criteria from February to November 2016, and from February to November 2020, respectively, were recruited for the study INCLUSION AND EXCLUSION CRITERIA Inclusion criteria All patients admitted to the Virgen del Rocio University Hospital aged 18 years or older who met the following criteria were eligible to participate in the study: 88 | Materials and Methods • Intra-abdominal infection: (a) intestinal-peritoneal (appendicitis, diverticulitis, diffuse and localised peritonitis secondary to perforation, previous surgery, ischaemia, intestinal obstruction or others); (b) hepato-bilio-pancreatic (cholangitis, cholecystitis, hepatic or pancreatic abscesses, infected pancreatic necrosis or others); (c) others at the discretion of the medical staff. • E. coli isolate from representative site of infection (intra-abdominal sample taken in the operating room by percutaneous puncture or aspiration and/or blood culture). Exclusion criteria Patients were excluded if they had spontaneous bacterial peritonitis, pelvic, prostatic, tuboovarian or abdominal wall abscesses, device-associated infection, surgical wound infection, patients with previous surgeries carrying biomedical devices and samples obtained through post-surgical drainage. CASE DEFINITION Patients were included in the cohort when they had at least one E. coli isolation on culture of intra-abdominal specimen during the study period. SOURCES OF INFORMATION, CASE IDENTIFICATION AND DATA COLLECTION An active search was performed every working day of every week of the study period by screening samples received at the HUVR microbiology laboratory to identify potential candidates. After obtaining the isolate, baseline data were collected, and follow-up was initiated. The information was collected by analysing the patients’ medical records using a standardised form. FOLLOW-UP OF CASES Patients were followed up until 120 days from the date of inclusion or until death or hospital discharge if earlier. Follow-up included the collection of new clinical information. All clinical and care decisions were made by the responsible physicians and nursing staff in charge of the patients. Materials and Methods | 89 Patient #4 was a 63-year-old man admitted for febrile neutropenia of 39 ºC. The patient had a history of myelodysplastic syndrome with excess blast (MDS-EB) that was being treated with azacytidine. E. coli susceptible to P/T was isolated on blood culture taken on the day of admission, and the patient was treated with P/T dosed 4 g / 0.5 g intravenous 4-hour extended infusion for 8 days. Surgical debridement was performed on 2nd day of admission. The patient remained febrile at 38 ºC for 8 days and a new blood culture was taken, where E. coli P/T resistant was isolated (Figure 13). BACTERIAL ISOLATES A couple of E. coli isolated were recovered from each single patient (AEC-11 and AEC- 51 from patient #1; AEC-15 and AEC-24 from patient #2; AEC-99 and AEC-106 from patient #3; PT3 and PT4 from patient #4) across two separate infection episodes, which display within-patient FIGURE 12.Schematic summary of the most relevant clinical data of the episode in which the samples were obtained from patient #3. 96 | Results evolution to P/T resistance. All these isolates were carbapenem susceptible. None of the isolates were producers of ESBL. The MIC in the second isolate (after intravenous therapy with P/T) increased compared to the MIC of the original isolate (before P/T treatment) in all four cases (Table 10). TO CHARACTERISE THE CLONAL RELATIONSHIP BETWEEN PIPERACILLIN-TAZOBACTAM-RESISTANT ISOLATES AND CLINICAL ISOLATES For the study of the relationship between P/T treatment and the occurrence of resistant isolates, all P/T-resistant strains were selected and compared with the reference strain ATCC 25922, and the ST and MIC of P/T were determined (Table 9). Thus, 4 P/T-resistant strains, corresponding to 4 different patients, were taken as P/T-resistant with MICs ranging from 128 to 256 mg/L. FIGURE 13.Schematic summary of the most relevant clinical data of the episode in which the samples were obtained from patient #4. Results | 97 TABLE 10. Microbiological characteristics of E. coli clinical isolates. patient isolate st mic of p/t (mg/l)blatem blaoxa-1 blashv blactx-m #1 AEC-11 162 4 Detected Not detected Not detected Not detected AEC-51 162 128 Detected Not detected Not detected Not detected #2 AEC-15 48 8 Detected Not detected Not detected Not detected AEC-24 48 256 Detected Not detected Not detected Not detected #3 AEC-99 73 8 Detected Not detected Not detected Not detected AEC-106 73 256 Detected Not detected Not detected Not detected #4 PT3 88 8 Detected Not detected Not detected Not detected PT4 88 256 Detected Not detected Not detected Not detected The MLST analysis performed with the assembled sequence showed a clonal relationship among each couple of isolates from the same patient, being ST162, ST48, ST73, and ST88 found in both strains recovered from the patient #1, #2, #3 and #4, respectively. TO STUDY THE MECHANISM UNDERLYING THE ACQUISITION OF RESISTANCE TO PIPERACILLIN-TAZOBACTAM. To better understand and describe the underlying mechanisms responsible for the substantial increase in P/T MIC, the four coupled isolates were sequenced by Oxford nanopore technologies™ (long reads) and Illumina® (short reads), making possible the complete assembly of the plasmids and their subsequent analyses. DETECTION OF BETA-LACTAMASES All the strains expressed blaTEM. Other ß-lactamase genes were not detected in these isolates. 98 | Results WGS AND BIOINFORMATIC ANALYSIS Promoter analysis Since blaTEM-1b is the only gene apparently related to ß-lactam resistance that was detected in all four pairs of isolates, the genetic background of blaTEM-1b was analysed. First, upstream regions of blaTEM-1b were analysed. In all cases, the promoter sequence upstream the blaTEM-1b gene was Pa/Pb. Moreover, comparison of the same region between isolates recovered before and after exposition to P/T showed no difference as well. Copy number To determine whether the isolates recovered after the treatment contain a higher copy number of blaTEM-1b, a simple comparison of the coverage of the blaTEM-1b gene vs. the coverage of dnaA, rpoB and mdh constitutive genes were performed (Table 11). Results showed different modes of action: (i) In AEC-11 isolated, the dnaA, rpoB and mdh (chromosomal genes (CG)) coverages were of 6.75x, 27.43x and 1.99x, respectively, providing a mean chromosomal coverage (MCC) of 12.06x. The coverage of the blaTEM-1b gene was 9.07x, representing a single copy per cell. Meanwhile, in AEC- 51 isolate, the chromosomal genes coverages were 8.14x, 24.55x and 3.94x, respectively, providing a MCC of 12.21x. The 233.75x coverage of blaTEM-1b represents 25 copies per cell. (ii) In the AEC-15 isolate, coverages of chromosomal genes of 53.21x, 41.64x and 59.47x, respectively, providing a MCC of 51.44x. The coverage of the blaTEM-1b gene was 4.63x, representing a single copy per cell. Meanwhile, in AEC-24 isolate, the coverages of chromosomal genes were 50.62x, 35.6x and 58.19x, respectively, providing a MCC of 48.14x. The 108.67x coverage of bla- TEM-1b represents 25 copies per cell. Similarly, in the AEC-99 isolate chromosomal genes coverages were 5.31x, 6.02x and 4.85x, respectively, providing a MCC of 5.39x. The blaTEM-1b coverage was 2.51x, representing a single copy per cell. Meanwhile, in AEC-106 isolate, chromosomal genes coverages were 3.6x, 3.48x and 0.59x, respectively, providing a MCC of 2.56x. The 45.17x coverage of blaTEM-1b represents 35 copies per cell. Genetic context analysis reveals that AEC-24 and AEC-106 isolates suffer same modifications with slight differences. While a single copy of blaTEM-1b gene and no single nucleotide polymorphisms (SNPs) were identified in the four E. coli isolates resistant to P/T, a consistent deletion of approximately 60 kb and 30 kb was observed respectively. In AEC-24, Results | 99 total read count average coverage (x) Patient 1 Patient 2 Patient 3 Patient 4 Patient 1 Patient 2 Patient 3 Patient 4 AEC11 AEC51 AEC15 AEC24 AEC99 AEC106 PT3 PT4 AEC11 AEC51 AEC15 AEC24 AEC99 AEC106 PT3 PT4 blaTEM-1b 69 1891 34 801 21 338 302 2173 9.07 233.75 4.63 108.67 2.51 45.17 21.66 517.89 dnaA 66 83 511 485 62 36 897 251 6.75 8.14 53.21 50.62 5.31 3.6 29.89 37.53 rpoB 769 703 1145 991 202 105 2500 943 27.43 24.55 41.64 35.6 6.02 3.48 48.09 47.89 mdh 14 28 387 377 38 5 598 10 1.99 3.94 59.47 58.19 4.85 0.59 0.74 0.84 Coverage patient 1 AEC11 AEC15 Comparison AEC15 AEC24 Comparison AEC99 AEC106 Comparison PT3 PT4 Comparison Mean chromosome 12.06 12.21 1.01 51.44 48.14 0.94 5.39 2.56 0.47 26.24 28.75 1.10 blaTEM-1b 9.07 233.75 25.77 4.63 108.67 23.47 2.51 45.17 18.00 21.66 517.89 23.91 Ratio blaTEM/ chromosome 0.75 19.14 25.45 0.09 2.26 25.08 0.47 17.67 37.96 0.83 18.01 21.82 TABLE 11. Coverage of selected plasmidic and chromosomal genes in clinical isolates. 100 | Results the consistently deleted region encoded SOS system inhibitor genes, plasmid partition system, methyltransferases genes, and the main conjugation machinery. On the other hand, AEC-106 lost regions encoding transporter system, iron elated genes, and two resistance determinants. (iii) In the PT3 isolate, chromosomal genes coverages were of 29.89x, 48.09x and 0.74x, respectively, providing a MCC of 26.24x. The blaTEM-1b coverage was 21.66x, representing a single copy per cell. Meanwhile, in PT4 isolate, the coverages of chromosomal gene were 37.53x, 47.89x and 0.84x, respectively, providing a MCC of 28.75x. The blaTEM-1b coverage was 517.89x representing18 copies per cell. Genetic context analysis of PT4. Plasmid content analysis The blaTEM-1-harboring plasmid backbones comparison from AEC-11 and AEC-51 isolates showed different sizes, being 150 kb for the pAEC-11 and 200 kb for pAEC-51. Alignment of both plasmid sequence revealed that pAEC-51 contained 5 copies of blaTEM-1 on the plasmid IncFIB. The 5 copies of blaTEM-1 gene have been generated through a series of ~ 15 kb DNA fragment translocation events leaded by IS26. The repeated fragment is composed by resistance genes (blaTEM-1, tet(B), dfrA and ant(3)-la), and MGEs (IS4, IS26 and IntI) (Figure 14). In AEC-15/AEC-24 and AEC-99/AEC-106 isolates, the same modification was observed with slight differences. While a single copy blaTEM-1 gene and no single nucleotide polymorphisms (SNPs) were identified in the four E. coli isolates, a consistent deletion of approximately 60 kb (from pAEC-15) and 30 kb (from pAEC-99) was observed for pAEC-24 and pAEC-106, respectively (Figures 16 and 17). In pAE-24, the consistently deleted region encoded SOS system inhibitor genes, plasmid partition system, methyltransferases genes, and the main conjugation machinery. On the other hand, pAEC-106 lost regions encoding transporter system, iron related genes, and two resistance determinants. In this case, deletions were surrounded by IS1, IS4 and IS26 (Figure 15 and Figure 16). For PT3 and PT4 isolates, BLAST analysis found two copies of blaTEM-1 gene in the isolate PT4 localized in the pPT4-IncFIB plasmid (150 kb), which showed 100% identity with the pPT3-Inc- FIB plasmid (150 kb), and in the pSmPT4 plasmid (8 kb), which was generated through the action of the IS26 and harbour resistance genes, such as blaTEM-1, dfrA, aph(6)-lc and aph(3)-l, and two origin of replication (repA and repC) (Figure 17). Results | 101 FIGURE 14. BLAST atlas of the blaTEM-1 harbouring plasmids comparison. pAEC-11 plasmid (IncFIB) compared to the pAEC-51 plasmid (IncFIB). GC content is represented on the distance scale (in kbp) on the inner map. Resistance genes (red), Mobile Genetic Elements [MGEs] (blue), replication genes (orange), conjugative transfer gene (green) and other genes (black) are also indicated in the external ring. The maps were drawn using BLAST Ring Image Generator (BRIG). 102 | Results FIGURE 15. BLAST atlas of the blaTEM-1 harbouring plasmids comparison. pAEC-15 plasmid (IncFII) compared to the pAEC-24 plasmid (IncFII). GC content is represented on the distance scale (in kbp) on the inner map. Resistance genes (red), Mobile Genetic Elements [MGEs] (blue), replication genes (orange), conjugative transfer gene (green) and other genes (black) are also indicated in the external ring. The maps were drawn using BLAST Ring Image Generator (BRIG) (http://sourceforge.net/projects/brig/). Results | 103 FIGURE 16. BLAST atlas of the blaTEM-1 harbouring plasmids comparison. pAEC-99 plasmid (IncFII/FIA/FIB) compared to the pAEC-106 plasmid (IncFII/FIA/FIB). GC content is represented on the distance scale (in kbp) on the inner map. Resistance genes (red), Mobile Genetic Elements [MGEs] (blue), replication genes (orange), conjugative transfer gene (green) and other genes (black) are also indicated in the external ring. The maps were drawn using BLAST Ring Image Generator (BRIG) (http://sourceforge.net/projects/brig/). 104 | Results FIGURE 17. BLAST atlas of the blaTEM-1 harbouring plasmids comparison. pPT3 plasmid (IncFIB) compared to the pPT4 (IncFIB) and pSmPT4 plasmids. GC content is represented on the distance scale (in kbp) on the inner map. Resistance genes (red), Mobile Genetic Elements [MGEs] (blue), replication genes (orange), conjugative transfer gene (green) and other genes (black) are also indicated in the external ring. The maps were drawn using BLAST Ring Image Generator (BRIG). Results | 105 ISOLATE (N=76) ORIGIN ST CC RESISTANCE MECHANISM C1-74 Blood 101 101 NEG C1-83 Blood 12 12 NEG C1-90 Blood 359 UK NEG C1-95 Blood 428 428 NEG C1-97 Blood 69 69 NEG C1-100 Blood UK UK NEG C1-110 Blood 697 10 NEG C1-111 Blood 10 10 NEG C1-128 Blood 12 12 NEG C1-138 Blood 345 UK NEG C1-144 Blood 315 38 NEG C1-146 Blood 131 131 NEG C1-152 Blood 533 UK NEG C1-155 Blood 101 101 NEG C1-157 Blood 13 13 NEG C1-168 Blood 74 73 NEG C1-169 Blood ND ND NEG C1-172 Blood 569 UK NEG C1-177 Blood 372 372 NEG C2-4 Bile 58 155 NEG C2-8 Bile 3640 UK NEG TABLE 15. Clinical isolates of E. coli susceptible to P/T. isolate (n=81) ORIGIN ST CC RESISTANCE MECHANISM C1-8 Blood 3640 UK NEG C1-31 Blood 95 95 TEM-40 C1-38 Blood 12 12 TEM-35 C1-48 Blood 58 155 TEM-1 C1-65 Blood 744 10 TEM-1 C1-72 Blood 533 UK TEM-1 C1-81 Blood 131 131 TEM-30 C1-91 Blood 197 UK NEG C1-93 Blood 428 428 SHV C1-102 Blood 95 95 TEM-1 C1-103 Blood 131 131 TEM-1 C1-106 Blood 131 131 TEM-1 C1-108 Blood 648 UK TEM-1 C1-118 Blood 131 131 TEM-1 C1-120 Blood 73 73 TEM-1 C1-121 Blood 80 586 TEM-1/SHV C1-126 Blood 12 12 TEM-1 C1-129 Blood 131 131 TEM-1 C1-130 Blood 10 10 TEM-1 C1-141 Blood 131 131 TEM-1/SHV C1-142 Blood 58 155 TEM-135/SHV TABLE 16. Clinical isolates of E. coli that were ESRI developers. 112 | Results ISOLATE (N=76) ORIGIN ST CC RESISTANCE MECHANISM C2-12 Bile 12 12 NEG AEC-2 Peritoneal fluid 95 95 NEG AEC-3 Bile 131 131 NEG AEC-4 Bile 58 155 NEG AEC-5 Intra-abdominal abscess 429 UK NEG AEC-6 Bile UK UK NEG AEC-7 Peritoneal fluid ND ND NEG AEC-8 Peritoneal fluid 3640 UK NEG AEC-10 Bile UK UK NEG AEC-11 Peritoneal fluid 53 53 NEG AEC-12 Peritoneal fluid 73 73 NEG AEC-14 Intra-abdominal abscess 36 5 NEG AEC-15 Peritoneal fluid 10 10 NEG AEC-16 Peritoneal fluid 10 10 NEG AEC-17 Pus ND ND NEG AEC-18 Peritoneal fluid 648 648 NEG AEC-19 Intra-abdominal abscess 131 131 NEG AEC-20 Intra-abdominal abscess 567 UK NEG AEC-22 Peritoneal fluid 44 10 NEG AEC-23 Peritoneal fluid 131 131 NEG AEC-25 Intra-abdominal abscess 73 73 NEG TABLE 15. Clinical isolates of E. coli susceptible to P/T (continued) isolate (n=81) ORIGIN ST CC RESISTANCE MECHANISM C1-153 Blood 533 UK TEM-1/SHV C1-159 Blood 95 95 TEM-1 C1-160 Blood UK UK TEM-1 C1-162 Blood ND ND NEG C1-164 Blood 10 10 TEM-1 C1-171 Blood 3948 131 NEG C1-174 Blood ND ND TEM-1 C1-175 Blood ND ND TEM-1 C1-183 Blood 648 UK TEM-1 C1-187 Blood 10 10 TEM-1 C1-190 Blood 65 UK SHV C1-191 Blood 372 372 TEM-1 C1-306 Blood 746 10 TEM-1 C2-47 Intra-abdominal abscess 569 UK TEM-1 C2-49 Peritoneal fluid 69 69 TEM-1 C2-54 Intra-abdominal abscess 69 69 TEM-1 PT3 Blood 88 23 TEM-1 ES-1 Blood ND ND TEM-1 ES-2 Blood ND ND TEM-1 ES-3 Blood ND ND TEM-1 ES-4 Blood ND ND TEM-1 TABLE 16. Clinical isolates of E. coli that were ESRI developers (continued) Results | 113 ISOLATE (N=76) ORIGIN ST CC RESISTANCE MECHANISM AEC-26 Peritoneal fluid 825 UK NEG AEC-27 Peritoneal fluid 705 131 NEG AEC-29 Intra-abdominal abscess 131 131 NEG AEC-30 Intra-abdominal abscess UK UK NEG AEC-31 Bile 74 73 NEG AEC-33 Bile 1982 131 NEG AEC-34 Peritoneal fluid 475 10 NEG AEC-35 Peritoneal fluid 14 14 NEG AEC-36 Liver abscess 14 14 NEG AEC-37 Intra-abdominal abscess 493 12 NEG AEC-39 Intra-abdominal abscess 12 12 NEG AEC-40 Peritoneal fluid 372 372 NEG AEC-41 Peritoneal fluid UK UK NEG AEC-42 Peritoneal fluid 14 14 NEG AEC-43 Peritoneal fluid 372 372 NEG AEC-44 Peritoneal fluid UK UK NEG AEC-45 Intra-abdominal fluid 4508 UK NEG AEC-46 Peritoneal fluid ND ND NEG AEC 47 Peritoneal fluid UK UK NEG AEC-49 Intra-abdominal abscess 69 69 NEG AEC-50 Peritoneal fluid 131 131 NEG TABLE 15. Clinical isolates of E. coli susceptible to P/T (continued) isolate (n=81) ORIGIN ST CC RESISTANCE MECHANISM ES-5 Blood ND ND TEM-1 ES-6 Blood ND ND TEM-1 ES-7 Blood ND ND TEM-1 ES-8 Blood ND ND TEM-1 ES-9 Blood ND ND TEM-1 ES-10 Blood ND ND TEM-1 ES-11 Blood ND ND TEM-1 ES-12 Blood ND ND TEM-1 ES-13 Blood ND ND TEM-1 ES-14 Blood ND ND TEM-1 ES-15 Blood ND ND TEM-1 ES-16 Blood ND ND TEM-1 ES-17 Blood ND ND TEM-1 ES-18 Blood ND ND NEG ES-28 Blood ND ND TEM-1 ES-29 Blood ND ND TEM-1 ES-30 Blood ND ND TEM-1 ES-31 Blood ND ND TEM-1 ES-32 Blood ND ND TEM-1 ES-34 Blood ND ND TEM-1 ES-35 Blood ND ND TEM-1 TABLE 16. Clinical isolates of E. coli that were ESRI developers (continued) 114 | Results ISOLATE (N=76) ORIGIN ST CC RESISTANCE MECHANISM AEC-52 Peritoneal fluid ND ND NEG AEC-54 Intra-abdominal abscess 69 69 NEG AEC-55 Peritoneal fluid ND ND NEG AEC-56 Peritoneal fluid 14 14 NEG AEC-57 Peritoneal fluid 131 131 NEG AEC-58 Peritoneal fluid 307 93 NEG AEC-59 Intra-abdominal abscess ND ND NEG AEC-60 Intra-abdominal abscess 542 UK NEG AEC-61 Peritoneal fluid UK UK NEG AEC-62 Peritoneal fluid ND ND NEG AEC-63 Peritoneal fluid ND ND NEG AEC-64 Intra-abdominal abscess 141 UK NEG AEC-66 Peritoneal fluid 93 10 TEM-1 isolate (n=81) ORIGIN ST CC RESISTANCE MECHANISM ES-36 Blood ND ND TEM-1 ES-37 Blood ND ND TEM-1 ES-38 Blood ND ND TEM-1 ES-39 Blood ND ND TEM-1 ES-41 Blood ND ND TEM-1 ES-42 Blood ND ND TEM-1 ES-43 Blood ND ND TEM-1 ES-44 Blood ND ND TEM-1 ES-45 Blood ND ND TEM-1 ES-47 Blood ND ND TEM-1 ES-48 Blood ND ND TEM-1 ES-49 Blood ND ND TEM-1 ES-50 Blood ND ND TEM-1 ES-51 Blood ND ND TEM-1 ES-52 Blood ND ND TEM-1 ES-53 Blood ND ND TEM-1 ES-54 Blood ND ND TEM-1 ES-55 Blood ND ND TEM-1 TABLE 15. Clinical isolates of E. coli susceptible to P/T (continued) TABLE 16. Clinical isolates of E. coli that were ESRI developers (continued) ST: sequence type, CC: clonal complex, UK: unknown ST, ND: not determined. NEG: negative result in relation to studied resistance mechanisms (ß-lactamases). ST: sequence type, CC: clonal complex, UK: unknown ST, ND: not determined. NEG: negative result in relation to studied resistance mechanisms (ß-lactamases). Results | 115 Isolate PRESSED (n=37) Origin ST CC Resistance mechanism C1-8 Blood 3640 UK NEG C1-31 Blood 95 95 TEM-40 C1-38 Blood 12 12 TEM-35 C1-48 Blood 58 155 TEM-1 C1-65 Blood 744 10 TEM-1 C1-72 Blood 533 UK TEM-1 C1-81 Blood 131 131 TEM-1 C1-91 Blood 197 UK NEG C1-93 Blood 428 428 SHV C1-102 Blood 95 95 TEM-1 C1-103 Blood ND ND TEM-1 C1-106 Blood 131 131 TEM-1 C1-108 Blood 648 UK TEM-1 C1-118 Blood 131 131 OXA-1 C1-120 Blood 73 73 TEM-1 C1-121 Blood 80 568 TEM-1/SHV C1-126 Blood 12 12 TEM-1 C1-129 Blood 131 131 TEM-1 C1-130 Blood 10 10 TEM-1 C1-141 Blood 131 131 TEM-1/SHV Isolate PRESSED (n=37) Origin ST CC Resistance mechanism C1-142 Blood 58 155 TEM-135/SHV C1-153 Blood 533 UK TEM-1/SHV C1-159 Blood 95 95 TEM-1 C1-160 Blood UK UK TEM-1 C1-162 Blood ND ND NEG C1-164 Blood 10 10 TEM-1 C1-171 Blood ND ND NEG C1-174 Blood ND ND TEM-1 C1-175 Blood ND ND TEM-1 C1-183 Blood 648 UK TEM-1 C1-187 Blood ND ND TEM-1 C1-190 Blood ND ND SHV C1-191 Blood 372 372 TEM-1 C1-306 Blood 746 10 TEM-1 C2-47 Intra-abdominal abscess 569 UK TEM-1 C2-49 Peritoneal fluid 69 69 TEM-1 C2-54 Intra-abdominal abscess 69 69 TEM-1 TABLE 17.Clinical isolates of E. coli that were ESRI developers with already acquired resistance to P/T. ST: sequence type, CC: clonal complex, UK: unknown ST, ND: not determined. NEG: negative result in relation to studied resistance mechanisms (ß-lactamases). 116 | Results TO EVALUATE THE ABILITY OF AN AUTOMATED SOFTWARE FOR ANALYSING MALDI-TOF MS SPECTRA TO DETECT P/T RESISTANCE. DIFFERENTIATION BETWEEN PHENOTYPIC ESRI DEVELOPERS SUSCEPTIBLE TO P/T AND THEIR ISOGENIC P/T-RESISTANT E. COLI Three machine learning supervised algorithms (PLS-DA, SVMs and RF), were applied to the peak matrix generated by the full spectrum and threshold methods of 37 ESRI developers susceptible to P/T and their isogenic P/T-resistant isolates. The average results after 20 repetitions of the 10-fold assays are recorded in Table 18. Correct separation between both categories was achieved by applying the RF algorithm using the full spectrum and threshold methods, with 90.13% ± 0.22% and 87.50% ± 0.40% accuracy, respectively. In turn, PLS-DA and SVM achieved accuracies lower than 81% in both methods. The 10-fold internal validation of the RF full spectrum method yielded an F1 score (the harmonic mean of the sensitivity and the accuracy of the model) of 91.43%, with a sensitivity and specificity of 86.49% and 97.3%, respectively. The positive and negative predictive values were 96.97% and 87.8%, respectively, assuming the isogenic P/T-resistant isolates as the positive category. Similar 10-folds scores, sensitivity and specificity percentages were obtained with the threshold method. Additionally, in the case of PLS-DA and SVM, the 10-fold precision was lower than 83%. TABLE 18. Average results in accuracy terms by repeating 20 times a 10-fold cross validation analysis for all ML supervised algorithms for each method and step PLS-DA SVM RF Full Spectrum (%) Threshold (%) Full Spectrum (%) Threshold (%) Full Spectrum (%) Threshold (%) ESRI (original vs. pressed) N=74 80.33 ± 0.68 78.51 ± 0.48 61.48 ± 0.22 57.43 ± 1.00 90.13 ± 0.22 87.50 ± 0.40 Step 1 (P/T-susceptible and ESRI –originalvs. P/T-resistant) N=209 92.85 ± 0.11 91.80 ± 0.10 77.14 ± 0.21 90.09 ± 0.04 96.92 ± 0.08 95.52 ± 0.14 Step 2 (P/T-susceptible vs. ESRI –original) N= 157 74.33 ± 0.23 71.52 ± 0.29 61.56 ± 0.34 55.66 ± 0.27 87.77 ± 0.25 81.75 ± 0.29 PLS-DA: Partial Least Square Discriminant Analysis; SVM: Linear Support Vector Machines; RF: Random Forest Results | 117 The distance plot of RF and PLS-DA analysis for discrimination between ESRI developers susceptible to P/T and their isogenic isolates resistant to P/T showed that the RF full spectrum and threshold methods achieved positive separation between both groups of isolates compared with the PLS-DA full spectrum and threshold methods (Figure 23). In the case of SVM full spectrum and threshold methods, distance plot analysis was not performed because principal component analysis (PCA) must be applied before, and it did not discriminate between samples. DIFFERENTIATION BETWEEN PHENOTYPIC P/T-SUSCEPTIBLE AND P/T-RESISTANT E. COLI The same three supervised algorithms were applied to the peak matrix generated by the full spectrum and threshold methods of 157 P/T-susceptible (53 of which were ESRI developers susceptible to P/T) and 53 P/T-resistant E. coli. FIGURE 22. Distance plot of RF and PLS-DA for discrimination between ESRI developers susceptible to P/T and their isogenic isolates resistant to P/T. A. C, Full spectrum method. B. D, Threshold method. ESRI: Extended spectrum resistance to BL/BLI susceptible to P/T; Pressed: ESRI isogenic isolates resistant to P/T after pressure with P/T; PLS-DA: partial least squares discriminant analysis; RF: random forest. T0 and t1 are two components of an algorithm called MDS that help to represent data from a distance matrix. 118 | Results In the first step of the assay for the determination of P/T resistance, RF achieved 96.92% ± 0.08% and 95.52% ± 0.14% for the full spectrum and threshold methods, respectively (Table 18). In this step, the PLS-DA (in full spectrum and threshold methods) and SVM in threshold method algorithms also showed more than 90% accuracy (Table 18). The 10-fold internal validation of the RF full spectrum method yielded an F1 score of 94.23%, with a sensitivity and specificity of 92.45% and 98.73%, respectively. The positive and negative predictive values were 98.08% and 97.48%, respectively, assuming resistant isolates as the positive category. Similar results were achieved with PLS in both methods. Additionally, in the case of SVM, the 10-fold precision was 70% and 90.09% in the full spectrum and threshold methods, respectively. The distance plot of RF and PLS-DA analysis for discrimination between P/T-susceptible and P/T-resistant E. coli showed that the RF full spectrum and threshold methods achieved positive separation between both groups of isolates compared with the PLS-DA full spectrum and threshold methods (Figure 24). In the case of SVM full spectrum and threshold methods, distance plot analysis was not performed because PCA did not discriminate between samples. FIGURE 23. Distance plot of RF and PLS-DA for discrimination between P/T-susceptible and P/T-resistant E. coli (first step analysis). A. C, Full Spectrum method. B. D, Threshold method. PLS-DA: Partial Least Square Discriminant Analysis; RF: Random Forest. T0 and t1 are two components of an algorithm called MDS that help to represent data from a distance matrix. Results | 119 DIFFERENTIATION BETWEEN PHENOTYPIC P/T-SUSCEPTIBLE AND ESRI DEVELOPER E. COLI To determine whether the software is able to differentiate between P/T-susceptible and ESRI- developer E. coli, we analysed two collections of P/T-susceptible E. coli (N=76) and ESRI developers (N=81). RF analysis achieved the best accuracy for the classification of ESRI developers from susceptible isolates compared with PLS-DA and SVM analyses (Table 18). The RF algorithm achieved 87.77% ± 0.25% and 81.75% ± 0.29% accuracy for the full spectrum and threshold methods, respectively. In turn, PLS-DA and SVM in both methods showed accuracies lower than 75% and 62%, respectively. The 10-fold internal validation of the RF full spectrum method yielded an F1 score of 86.79%, with 85.19% and 88.16% sensitivity and specificity, respectively, assuming ESRI developers were the positive category. The positive and negative predictive values were 88.46% and 84.81%, respectively. Additionally, in the case of PLS-DA and SVM, the 10-fold accuracy was lower than 75%. FIGURE 24. Distance plot of RF and PLS-DA for discrimination between P/T-susceptible and ESRI developer Escherichia coli (second step analysis). A. C, Full Spectrum method. B. D, Threshold method. PLS-DA: Partial Least Square Discriminant Analysis; RF: Random Forest. T0 and t1 are two components of an algorithm called MDS that help to represent data from a distance matrix. 120 | Results The distance plot of RF and PLS-DA analysis for discrimination between P/T-susceptible and ESRI developer E. coli showed that the RF full spectrum and threshold methods achieved positi- ve separation between both groups of isolates compared with the PLS-DA full spectrum and threshold methods (Figure 24). In the case of SVM full spectrum and threshold methods, distance plot analysis was not performed because PCA did not discriminate between samples. DIFFERENTIATION BETWEEN PHENOTYPIC P/T-SUSCEPTIBLE AND ESRI DEVELOPER OR P/T-RESISTANT E. COLI BY PEAK FEATURE IMPORTANCE STUDY AND ROC CURVES Further discrimination in the three assays between (i) ESRI developer isolates susceptible to P/T and their isogenic isolates resistant to P/T, (ii) P/T-susceptible isolates and P/T-resistant isolates, and (iii) P/T-susceptible isolates and ESRI developer isolates susceptible to P/T was attempted using a different approach for peak analysis by the feature importance of RF using full spectrum method (Figure 25). In these features, the higher the feature importance is, the more important it is when splitting the samples in the classifier trees. The importance of a feature is computed as the (normalised, the values of the array sum to 100%) total reduction of the criterion brought by that feature. It is also known as the Gini importance. Feature importance was obtained from RF between phenotypic ESRI developers susceptible to P/T and their isogenic P/T-resistant isolate assays, showing a high area of interest from 2000 to 3500 m/z (Figure 25-A). In addition, feature importance was also obtained from RF between phenotypic P/T susceptible isolates and P/T-resistant isolate assays, showing a high area of interest from 2000 to 3500 m/z (Figure 25-B) On the other hand, the distribution of the feature importance in differentiation between P/T-susceptible isolates and ESRI developers isolates susceptible to P/T was more homogeneous and did not show any characteristic range (Figure 25-C), but instead the entire spectrum contributes equally to the classification. Receiver Operating Characteristic (ROC) curves were obtained from RF for the same three assays described above (Figure 26). Furthermore, their area under the curve (AUC) were obtained for the first 10-folds of each assay. The average AUC-ROC was 0.995 ± 0.05 for ESRI developer isolates susceptible to P/T and their isogenic isolates resistant to P/T, 0.981 ± 0.02 for P/T-susceptible isolates and P/T-resistant isolates, and 0.929 ± 0.06 for P/T-susceptible isolates and ESRI developer isolates susceptible to P/T. Isogenic isolates resistant to P/T, the resistant isolates to P/T and ESRI developer isolates susceptible to P/T were labelled as positive categories from getting the previous results, respectively. Results | 121 0.001), septic shock (p < 0.001), sepsis (p < 0.001), and the presence of additional care support such as urinary catheter (p < 0.001), transitory central venous catheter (CVC) (p < 0.001), nasogastric tube (p < 0.001), and mechanical ventilation (p < 0.001). In contrast, a higher percentage of the patients with the polymicrobial infection and appropriate empirical therapy in the first 24 h was observed in the survival group (p = 0.01 and p < 0.001, respectively). With respect to the time intervals between the symptoms onset and the surgical intervention, the dead patients have diagnosed later; the time intervals between the symptoms onset and the surgical intervention and between the first medical contact and the surgical intervention were near and higher to the two times (p = 0.03 and p = 0.04) for the dead patients vs. survival patients (p = 0.03 and p = 0.04), respectively. Finally, the multivariate analysis in both the cohorts grouped together identify the patients diagnosed during the COVID-19 (p = 0.02), the longer period from the symptoms onset to the surgical intervention (p = 0.01), septic shock (p = 0.03), the Charlson comorbidity index (p = 0.001) as independent factors associated with death, and the appropriate empirical therapy during the first 24 h (p = 0.001) as a protective factor for death (Table 21). 128 | Results Factor Death (n = 23) Survival (n= 181) P-value Age, median (range) 72 (62 – 76) 63 (51 – 73) 0.02 Male sex, No. (%) 14 (60.9) 106 (58.6) 0.83 IAI patients in COVID-19 period, No. (%) 16 (69.6) 80 (44.2) 0.02 Emergency surgery, No. (%) 13 (56.5) 81 (44.7) 0.28 Post-surgical abscess 5 (21.7) 55 (30.4) 0.39 Post-surgical peritonitis 6 (26.1) 16 (8.8) 0.02 Cholecystitis 2 (8.7) 47 (26.0) 0.07 Appendicitis 1 (4.3) 20 (11.0) 0.31 Hollow viscus perforation 6 (26.1) 17 (9.4) 0.01 Diverticulitis 0 (0) 10 (5.5) 0.24 Others 3 (13.0) 16 (8.9) 0.51 Bacteremia, No. (%) 6 (26.1) 9 (5.0) <0.001 McCabe score, ultimately or rapidly fatal 16 (69.6) 40 (22.1) <0.001 Charlson index, median (range) 6 (4 – 7.5) 2 (0 – 4) < 0.001 Comorbid condition, No. (%) Cancer 9 (39.1) 55 (30.4) 0.39 Metastatic cancer 3 (13.0) 13 (7.2) 0.39 Diabetes 3 (13.0) 27 (14.9) 1.00 Autoimmune diseases 1 (4.3) 4 (2.2) 0.45 Dementia 0 (0.0) 1 (0.6) 1.00 Chronic lung disease 3 (13.0) 16 (8.8) 0.51 Vascular disease 1 (4.3) 7 (3.9) 1.00 Chronic kidney disease 3 (13.0) 4 (2.2) 0.03 Liver disease 3 (13.0) 5 (2.8) 0.01 Hematologic malignancies 0 (0) 2 (1.1) 0.88 Immunosuppression 4 (17.4) 16 (8.9) 0.25 Neutropenia 1 (4.3) 0 (0) 0.11 Acquisition type, No. (%) Community-acquired 10 (43.5) 105 (58.0) 0.19 Healthcare-related 3 (13.0) 13 (7.2) 0.40 Hospital-acquired 10 (43.5) 63 (34.8) 0.41 Polymicrobial infection, No. (%) 7 (30.4) 103 (56.9) 0.01 Septic shock, No. (%) 7 (30.4) 14 (7.7) <0.001 Sepsis, No. (%) 16 (69.6) 27 (14.9) <0.001 Previous intervention, No. (%) 7 (30.4) 14 (7.7) 0.39 Presence of devices, No. (%) Urinary catheter 21 (91.3) 100 (55.2) <0.001 TABLE 20. Univariate analysis of the factors related to death in a combined cohort of the pre-COVID-19 pandemic cohort and the COVID-19 pandemic cohort. Results | 129 Factor Death (n = 23) Survival (n= 181) P-value Transitory CVC 17 (73.9) 56 (30.9) <0.001 Permanent CVC 0 (0) 2 (1.1) 1.00 Nasogastric tube 15 (65.2) 48 (26.5) <0.001 Mechanical ventilation 10 (43.5) 24 (13.3) 0.001 Times intervals from the SO to the intervention, median of days (range) From SO to 1st medical contact 1 (0 – 4) 1 (0 –3.25) 0.88 From SO to intervention 11 (5 – 17.5) 6 (2 –10) 0.03 From 1st medical contact to admission 1 (0 – 6) 2 (0 – 10) 0.17 From 1st medical contact to intervention 8 (1.5 – 17) 3 (1 – 9) 0.04 From admission to intervention 1 (0 – 7.5) 1 (0 – 4) 0.76 Appropriate empirical thepry (24 h), No. (%) 7 (30.4) 134 (74) < 0.001 IAI, intraabdominal infection; CVC, central venous catheter; SO, symptoms onset. TABLE 20. Univariate analysis of the factors related to death in a combined cohort of the pre-COVID-19 pandemic cohort and the COVID-19 pandemic cohort (continued). TABLE 21. Multivariate analysis of the factors related to death in a combined cohort of the pre- COVID-19 pandemic cohort and the COVID-19 pandemic cohort. Factor Death (n = 23) Survival (n = 181) P-value OR (CI 95%) Age, median (range) 72 (62 – 76) 63 (51 – 73) 0.567 1.01 (0.97 – 1.06) Charlson index ≥ 3, No. (%) 19 (82.6) 53 (29.3) 0.001 9.83 (4.63 – 17.49) IAI patients in COVID-19 period, No (%) 16 (69.6) 80 (44.2) 0.02 6.23 (2.10 – 14.92) Days from SO to intervention, median (range) 11 (5 –17.5) 6 (2 – 10) 0.01 5.03 (2.98 – 8.74) Septic shock, No. (%) 7 (30.4) 14 (7.7) 0.03 5.69 (2.64 – 9.43) Post-surgical peritonitis, No. (%) 6 (26.1) 16 (8.8) 0.705 1.32 (0.31 – 5.53) Hollow viscus perforation, No. (%) 6 (26.1) 17 (9.4) 0.181 2.61 (0.39 – 4.47) Appropriate empirical therapy (24h), No. (%) 7 (30.4) 134 (74.0) 0.001 0.27 (0.01 – 0.41) So, symptoms onset; CI, confidence interval; OR, odds ratio. 130 | Results Results | 131 Discussion TO ANALYSE THE ASSOCIATION BETWEEN P/T TREATMENT AND THE ACQUISITION OF RESISTANCE IN ESCHERICHIA COLI IN PATIENTS WITH INTRA-ABDOMINAL INFECTION, WITH OR WITHOUT BACTERAEMIA Significant morbidity and mortality rates are still recorded for complicated intraabdominal infections (IAI), despite advance in early diagnosis and treatment. Prompt antimicrobial therapy and adequate source control are the cornerstones of ideal care. Adequate antimicrobial therapy has grown increasingly challenging due to the worldwide spread of multidrug-resistant bacteria. Antibiotic resistance has risen dramatically worldwide in recent years and the WHO, and several governmental organisations have been warning the loss of effectiveness of antibiotics due to the loss of bacterial susceptibility (95,99,100). Both public health organisations and the scientific community agree that the need to halt the advance of this phenomenon is a priority. The increasing emergence of resistant isolates means that therapeutic approaches are becoming increasingly limited and not only have a high clinical cost, which in many cases leads to the death of patients, but also have a high economic impact due to the cost of antibiotic use, the increased occurrence of clinical complications -which in turn are associated with a greater need for more complex therapeutic proceduresas well as longer periods of hospitalisation and need for human resources for these patients. P/T is used as a first-line empirical treatment for serious infections like pneumonia and intra-abdominal infections (171) because it has broad spectrum activity against Gram-negative and -positive bacteria (172), is well tolerated (173), and is available for paediatric use. Its broad spectrum makes it a crucial agent for lowering the use of carbapenems, which are crucial last-resort antibiotics globally. Limiting the use of carbapenem is an important aspect of antimicrobial stewardship and is necessary to stop the spread of resistance. Resistance to P/T can be caused by the production of multiple ß-lactamases enzymes (174), carbapenemases enzymes (175) or ESBLs in combination with increased efflux or porin loss (176), which can also lead to third generation cephalosporins resistance. The phenotype of resistance to P/T with susceptibility to third generation cephalosporins indicates the possibility of alternative resistance mechanisms. Some causes of this phenotype are the hyperproduction of class A ß-lactamases such as TEM-1 (121,123). Increased production of ß-lactamase overcomes the inhibitive effect of tazobactam, ostensibly through saturation of the inhibitor, allowing the excess enzyme to hydrolyse and degrade piperacillin (119). Other mechanisms have also been identified, including production class D ß-lactamases as OXA-1 (119) and inhibitor-resistant enzymes such as TEM-33. Understanding the mechanism of resistance is becoming increasingly important due to the emergence of a P/T-resistant but third-generation cephalosporin and carbapenem susceptible E. coli phenotype (177–179), as well 134 | Discussion as the growing reliance on P/T as an empirical treatment and the recent interest in using P/T as a carbapenem sparing treatment for extended-spectrum ß-lactamase infections (180). CLINICAL CASES Initially, three isolates of P/T-resistant, third generation cephalosporin and carbapenem susceptible E. coli from intraabdominal source were selected. Each patient had a corresponding P/T- susceptible isolate from the same or previous infection episode, recovered prior to P/T therapy. A fourth isolate of P/T-resistant, third generation cephalosporin and carbapenem susceptible E. coli was recovered from blood culture with abscess focus. The MIC in the second isolate, recovered after intravenous therapy with P/T for 12, 18, 14 and 8 days, respectively, increased 5-fold to the MIC in the first isolate, susceptible to P/T. These data are consistent with previous findings obtained following in vitro P/T pressure (123), where antibiotic pressure with subinhibitory P/T concentrations were associated with selection of P/T-resistant isolates in TEM-harbouring E. coli strains (123). Although TEM-1 can hydrolyse benzylpenicillins and narrow-spectrum cephalosporins, these enzymes are effectively inhibited by tazobactam (71,181). Thus, theoretically the production of TEM-1 should not confer resistance to P/T, as tazobactam has the ability to inhibit class A ß-lactamases (71). As an example, the E. coli P/T- susceptible harboured TEM-1. However, resistance could have been attributed to the hyperexpression of the encoded TEM-1, as previously described (119,121,123). In another recent study (117), a paired clinical isolates of E. coli with P/T resistance phenotype and susceptibility to third generation cephalosporins and carbapenems was identified, in which resistance seemed to have been acquired in vivo after a 12-day course of P/T treatment. Although all four patients had been treated with P/T in a dosing regimen that extended exposure (4 g / 0.5 g four-hour extended infusion), with the aim of optimising exposure by leveraging of the pharmacokinetic and pharmacodynamic characteristics of ß-lactams in combination with ß-lactamase inhibitors, all isolates were recovered from difficult-to-control foci of infection. Previous studies have reported that P/T concentrations reached in the abdomen may be lower compared to concentrations reached in blood (182). The concentrations reached in the abdomen may therefore be insufficient for the purpose of inhibiting bacterial growth and may exert the same antibiotic pressure effect as the sub-inhibitory concentrations used in the in vitro assay. In the case of the patient in whom Discussion | 135 the strains were recovered from blood culture, the focus of the bacteraemia was a perianal abscess, in which the penetration of the antibiotic could have been similarly hampered. While combinations of ß-lactams with ß-lactamase inhibitors have potentially interesting features for the treatment of complicated intra-abdominal infections, it may be expected that the resistance mechanisms observed over the last four decades following massive selective pressure by ß-lactamases may be maintained or even enhanced using these drugs. Furthermore, P/T has the same pharmacokinetic profile as other ß-lactams: a limited volume of distribution and renal clearance. Both conditions are largely impacted by severe infections where major pharmacokinetic changes are expected (183). An increased volume of distribution decreased peak concentrations and dose adjustments for renal dysfunction have been reported on many occasions during the use of ß-lactams as causes of under- o overdosage (183). These results suggest that prolonged treatment with P/T in patients with complicated focus control infections may exert selective pressure, leading to selection of P/T-resistant isolates. CLONAL RELATIONSHIP BETWEEN P/T-RESISTANT ISOLATES AND CLINICAL ISOLATES There are several genotyping techniques based on the study of chromosomal DNA, which differ in their reproducibility, genome coverage and discriminatory power (184). Each isolate recovered after treatment belonged to the same sequence type as the strain isolated before P/T treatment, demonstrating the clonal relationship between them. MLST analysis revealed the P/T-resistant after treatment occurred in a diverse number of sequence types: ST162 in patient #1, ST48 in patient #2, ST73 in patient #3, and ST88 in patient #4, suggesting the acquisition of resistance to P/T is not driven by a clonal outbreak of high-risk clones, but rather by multiple acquisitions of resistance mechanisms in the circulating population of clinical strains. Previous studies highlighting the phenomenon of acquisition of P/T resistance after antibiotic exposure also showed variability in the STs of the strains studied (117,123). This data suggest the impact of repeated or sustained antimicrobial pressure, rather than fixation in a certain lineage and subsequent spread. 136 | Discussion STUDY OF THE MECHANISM UNDERLYING THE ACQUISITION OF RESISTANCE TO PIPERACILLIN-TAZOBACTAM TEM-1 production Tazobactam is able to inhibit the activity of class A ß-lactamases (71), and therefore the presence of blaTEM-1 within the genome of an E. coli isolate should not result in resistance to P/T. However, several studies have demonstrated that isolates of E. coli carrying TEM enzymes could acquire P/T resistance by exposure to P/T in vitro (121,123,185), we suggested that the increase in P/T MIC values of the isolates after exposure to P/T was due to the hyperproduction of TEM enzymes. To evaluate this, we detected the presence of blaTEM-1 in these isolates. All the isolated expressed blaTEM-1b. It is noteworthy that blaOXA-1 was not detected in these isolates, although blaOXA-1 has been previously linked to P/T-resistant and third generation cephalosporines susceptibility phenotype (119,181). These data are consistent with previous studies in which the acquisition of resistance to P/T while maintaining susceptibility to third generation cephalosporins and carbapenems was associated with hyperproduction of TEM-1, both in vitro and in vivo (117,121,123,185). ß-lactamase hyperproduction can occur via increased gene expression modulated by a stronger promoter (125), or an increase in gene copy number mediated by insertion sequences (117,185) or plasmids (123). In Hubbard et al. (117), the authors also seem to find TEM-1 hyperproduction as the most likely mechanism for the acquisition of P/T resistance in vivo, in contrast to other works that link P/T resistance to the presence of OXA-1-type ß-lactamases (181). Promoter analysis In some cases, higher blaTEM-1 enzyme levels were attributed to a Pa/Pb mutation in the blaTEM-1 promoter that increased transcription (124,186,187). In order to verify whether differential expression was due to different promoter sequences, upstream regions of blaTEM-1 gene were analysed. In all cases, the promoter upstream the blaTEM-1 gene was Pa/Pb. The Pa/Pb promoter is a strong promoter characterized to be involved in TEM-1 hyperproduction and leading to resistance to BLI such as clavulanate and sulbactam (124) and also tazobactam (117,125). No differences in promoter sequence were found in the comparison between susceptible and resistant isolates, suggesting that the increase in TEM-1 production is not due to a change from a Discussion | 137 performed. That way, we have been able to achieve good results with no more than 1% of variability between k-fold tests by stratifying the folds of the method, considering the categories of the samples to always obtain a balanced fold. Nevertheless, further studies of this models, by including a validation set, will improve the models. TO ANALYSE THE IMPACT OF THE SARS-COV-2 PANDEMIC ON THE MORBIDITY AND MORTALITY OF INTRA-ABDOMINAL E. COLI INFECTION. This study shows that the mortality of IAI by E. coli was higher during the COVID-19 pandemic period compared to the COVID-19 pandemic period. To the best of our knowledge, this is the first study that analysed the impact of the COVID-19 on the IAI diagnosis, treatment and outcomes. The bivariate analysis reported that the patients with IAI during the COVID-19 pandemic had more comorbidities, required more emergency surgery, and had more severe infections. With regard to the lower percentage of the patients with bacteraemia during the COVID-19 pandemic, it is due to the decrease in the blood cultures requests by the physicians and consequently less bacteraemia detection in the microbiology service. According to the data, it is important to note that this study does not include bacteraemia in the multivariate analysis to avoid any bias in the independent variables introduced in the analysis. Moreover, the period of time between the symptoms onset and the surgical intervention was longer for the patients with IAI during the COVID-19 pandemic period compared to the patients with IAI in the pre-COVID-19 pandemic period and it was independently associated with death. Despite the logistical difficulties caused by the COVID-19 pandemic, no significant delay was detected between the hospital admission and the surgical intervention, a period that only includes in-hospital factors. Of note, the visit to the emergency department and the use of the operating room decreased in 2020, the COVID-19 pandemic period, in our hospital by 25% and 22%, respectively. In turn, there was an increase in the time between the symptoms onset and the first medical contact, a period that includes the prehospital and in-hospital factors and between the first medical contact and the hospital admission. Probably, these issues are owing to the time spent by the patients to go to the hospital or the primary care attention due to the necessary protection measure applied in Spain during the COVID-19 pandemic. In this study, the possible explanations for these results could be a combination of the avoidance of medical care due to the social distancing and the fear of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) hospital transmission and to the use of the telephone consultation with the physician, a process in which the physicians had no prior experience or training. 144 | Discussion This issue seems to be suggested in our study, in which the patients with IAI avoid going to the hospitals or they go later, as reflected by the high number of patients with more severe infections. In Italy, a similar observation was reported in which the urgent emergency interventions were dropped and were associated with a more severe presentation due to a diagnostic delay (210). In addition, this study also found that the patients during the COVID-19 pandemic had more severe infections with higher rates of septic shock and sepsis. This could be explained by the fact that during the COVID-19 pandemic, the patients were admitted to the hospital in an advance stage of the disease. The delay in the time interval between the first medical contact and the surgical intervention could be also another factor, which could imply an increase of the complications before the abdominal surgery, which can further lead to hemodynamic instability (e.g., perforations, complicated abscesses). In this PhD thesis, it is important to highlight the worrying increase in the mortality of patients with IAI during the COVID-19 pandemic. Recent epidemiological data indicate a significant increase in mortality during this period that cannot be fully explained by the patients with the COVID-19 alone (211). Lack or delay in access to the specific antimicrobial treatment also would imply an increase in the severity of infections. Thus, it may be useful that recommendation statements devote attention to the management of the patients with IAI and the other severe infections during the COVID-19 health crisis. This study also has several limitations. The first one is the small sample size used and the patients are recruited from a unicenter cohort. A larger sample size and multicentre cohorts would help to confirm our results. Moreover, this study has only been able to analyse the available data in the pre-COVID-19 pandemic period and during the COVID-19 pandemic period. However, this study provides some advantages. We have compared the demographic, clinical and microbiological variables of the patients with IAI in the two periods differing only in the presence or absence of the COVID-19 pandemic situation and also identified the impact of the days before the diagnosis and surgical interventions and their effects on the death between the pre-COVID-19 pandemic period and the COVID-19 pandemic period. Discussion | 145 Conclussions OBJECTIVE 1 1. Treatment with piperacillin-tazobactam (P/T) can generate resistance to this antimicrobial and treatment failure in high-inoculum infections due to E. coli producing TEM-1 ß-lactamase. 2. Evolution towards the emergence of P/T-resistant isolates in E. coli during P/T treatment is a process linked to an increase in blaTEM expression. 3. There are three different mechanisms of higher blaTEM expression: the increase in blaTEM copy number mediated by IS26 insertion sequences, generation of a small multicopy plasmid and adaptive evolution through reduction of plasmid size and fitness cost, leading to increase plasmid copy number. 4. Long read sequencing is crucial in the elucidation of the resistance mechanisms, mainly those involving numerous IS-elements and repeat sequences. 5. The acquisition of resistance to P/T induced by P/T treatment is not driven by a clonal outbreak of high-risk clones, but rather by multiple acquisitions of resistance mechanisms in the circulating population of clinical strains. 6. Acquired resistance to P/T induced by P/T treatment is a stable process that is long-term stable, even after withdrawal of the antibiotic. 7. Exposure to P/T has been shown to induce adaptive mutations in OmpC and its regulator OmpR, reducing its expression. OBJECTIVE 2 8. Analysis of MALDI-TOF MS spectra data using an automated system based on machine learning algorithms shows high sensitivity and specificity to differentiate ESRI-developing and P/T-resistant E. coli isolates. 148 | Conclussions 9. The use of this system allows early detection of P/T resistant isolates, as well as isolates with P/T susceptibility but with ability to acquire resistance. 10. The RF analysis directly identifies ESRI-developer isolates and differentiates them from P/T-susceptible and P/T-resistant isolates. More specifically, it has been identified a specific signature (series of peaks between 2000 and 3500 m/z) associated with ESRI-developer and P/T-resistant isolates, leading to the precise differentiation between ESRI-developer, P/T-re- sistant and susceptible isolates. OBJECTIVE 3 11. The COVID-19 pandemic had a negative impact on the care of patients with intra-abdominal E. coli infections. 12. Delayed care of patients with intra-abdominal E. coli infections during the COVID-19 pandemic led to increased mortality in these patients. Conclussions | 149 fected Control Subjects. J Infect Dis. 2004 Dec 15;190(12):2121–8. 14. von Wintersdorff CJH, Penders J, van Niekerk JM, Mills ND, Majumder S, van Alphen LB, et al. Dissemination of Antimicrobial Resistance in Microbial Ecosystems through Horizontal Gene Transfer. Front Microbiol [Internet]. 2016 Feb 19 [cited 2022 Jun 2];7. Available from: http://journal.frontiersin.org/Article/10.3389/fmicb.2016.00173/ abstract 15. Furuya EY, Lowy FD. Antimicrobial-resistant bacteria in the community setting. Nat Rev Microbiol. 2006 Jan 1;4(1):36–45. 16. Partridge SR, Kwong SM, Firth N, Jensen SO. Mobile Genetic Elements Associated with Antimicrobial Resistance. Clin Microbiol Rev. 2018 Oct;31(4):e00088-17. 17. Datta N, Hedges RW. Compatibility Groups among fi− R Factors. Nature. 1971 Nov;234(5326):222–3. 18. Couturier M, Bex F, Bergquist PL, Maas WK. Identification and classification of bacterial plasmids. MICROBIOL REV. 1988;52:21. 19. Villa L, García-Fernández A, Fortini D, Carattoli A. Replicon sequence typing of IncF plasmids carrying virulence and resistance determinants. J Antimicrob Chemother. 2010 Dec;65(12):2518–29. 20. Tansirichaiya S, Rahman MA, Roberts AP. The Transposon Registry. Mob DNA. 2019;10:40. 21. Harmer CJ, Hall RM. An analysis of the IS6/IS26 family of insertion sequences: is it a single family? Microb Genomics [Internet]. 2019 Sep 1 [cited 2022 Jul 13];5(9). Available from: https://www.microbiologyresearch.org/ content/journal/mgen/10.1099/mgen.0.000291 22. He S, Hickman AB, Varani AM, Siguier P, Chandler M, Dekker JP, et al. Insertion Sequence IS26 Reorganizes Plasmids in Clinically Isolated Multidrug-Resistant Bacteria by Replicative Transposition. mBio. 2015 Jun 9;6(3):e00762. 23. Varani A, He S, Siguier P, Ross K, Chandler M. The IS6 family, a clinically important group of insertion sequences including IS26. Mob DNA. 2021 Dec;12(1):11. 24. Escudero* JA, Loot* C, Nivina A, Mazel D. The Integron: Adaptation On Demand. Microbiol Spectr. 2015 Mar 12;3(2):3.2.10. 25. Arakawa Y, Murakami M, Suzuki K, Ito H, Wacharotayankun R, Ohsuka S, et al. A novel integron-like element carrying the metallo-beta-lactamase gene blaIMP. Antimicrob Agents Chemother. 1995 Jul;39(7):1612–5. 26. Collis CM, Kim MJ, Partridge SR, Stokes HW, Hall RM. Characterization of the class 3 integron and the site-spe- cific recombination system it determines. J Bacteriol. 2002 Jun;184(11):3017–26 27. Xu H, Davies J, Miao V. Molecular characterization of class 3 integrons from Delftia spp. J Bacteriol. 2007 Sep;189(17):6276–83. 28. Ramírez MS, Piñeiro S, Argentinian Integron Study Group, Centrón D. Novel insights about class 2 integrons from experimental and genomic epidemiology. Antimicrob Agents Chemother. 2010 Feb;54(2):699–706. 29. Deng Y, Bao X, Ji L, Chen L, Liu J, Miao J, et al. Resistance integrons: class 1, 2 and 3 integrons. Ann Clin Microbiol Antimicrob. 2015 Dec;14(1):45. 30. Croxen MA, Finlay BB. Molecular mechanisms of Escherichia coli pathogenicity. Nat Rev Microbiol. 2010 Jan;8(1):26–38. Bibliography 1. Tenaillon O, Skurnik D, Picard B, Denamur E. The population genetics of commensal Escherichia coli. Nat Rev Microbiol. 2010 Mar;8(3):207–17. 2. Vila J, Sáez-López E, Johnson JR, Römling U, Dobrindt U, Cantón R, et al. Escherichia coli : an old friend with new tidings. Gerdes K, editor. FEMS Microbiol Rev. 2016 Jul;40(4):437–63. 3. Kaper JB, Nataro JP, Mobley HLT. Pathogenic Escherichia coli. Nat Rev Microbiol. 2004 Feb;2(2):123–40. 4. Brenner DJ, Krieg NR. Bergey’s Manual of Systematic Bacteriology. 2nd ed. Vol. 2. East lansing, MI, USA: Springer International Publishing; 2005. 5. Strockbine NA, Boopp CA, Fields PI, Kaper JB, Nataro JP. Escherichia, Shigella and Salmonella. In: Manual of Clinical Microbiology. 11th ed. 2015. p. 685–6. 6. Enterobacteriaceae. In: Microbiología Médica. 8a Edición. Barcelona: Elsevier; 2017. p. 251–71. 7. Priest FG, Barker M. Gram-negative bacteria associated with brewery yeasts: reclassification of Obesumbacterium proteus biogroup 2 as Shimwellia pseudoproteus gen. nov., sp. nov., and transfer of Escherichia blattae to Shimwellia blattae comb. nov. Int J Syst Evol Microbiol. 2010 Apr 1;60(4):828–33. 8. Hata H, Natori T, Mizuno T, Kanazawa I, Eldesouky I, Hayashi M, et al. Phylogenetics of family Enterobacteriaceae and proposal to reclassify Escherichia hermannii and Salmonella subterranea as Atlantibacter hermannii and Atlantibacter subterranea gen. nov., comb. nov.: Atlantibacter hermannii gen. nov., comb. nov. Microbiol Immunol. 2016 May;60(5):303–11. 9. Alnajar S, Gupta RS. Phylogenomics and comparative genomic studies delineate six main clades within the family Enterobacteriaceae and support the reclassification of several polyphyletic members of the family. Infect Genet Evol. 2017 Oct;54:108–27. 10. Adeolu M, Alnajar S, Naushad S, S. Gupta R. Genome-based phylogeny and taxonomy of the ‘Enterobacteriales’: proposal for Enterobacterales ord. nov. divided into the families Enterobacteriaceae, Erwiniaceae fam. nov., Pectobacteriaceae fam. nov., Yersiniaceae fam. nov., Hafniaceae fam. nov., Morganellaceae fam. nov., and Budviciaceae fam. nov. Int J Syst Evol Microbiol. 2016 Dec 1;66(12):5575–99. 11. Denamur E, Clermont O, Bonacorsi S, Gordon D. The population genetics of pathogenic Escherichia coli. Nat Rev Microbiol. 2021 Jan;19(1):37–54. 12. Savageau MA. Escherichia coli Habitats, Cell Types, and Molecular Mechanisms of Gene Control. Am Nat. 1983 Dec;122(6):732–44. 13. Sannes MR, Kuskowski MA, Owens K, Gajewski A, Johnson JR. Virulence Factor Profiles and Phylogenetic Background of Escherichia coli Isolates from Veterans with Bacteremia and Unin- 152 | Bibliography 31. Desvaux M, Dalmasso G, Beyrouthy R, Barnich N, Delmas J, Bonnet R. Pathogenicity Factors of Genomic Islands in Intestinal and Extraintestinal Escherichia coli. Front Microbiol. 2020 Sep 25;11:2065. 32. Pakbin B, Brück WM, Rossen JWA. Virulence Factors of Enteric Pathogenic Escherichia coli: A Review. Int J Mol Sci. 2021 Sep 14;22(18):9922. 33. Verway M, Brown KA, Marchand-Austin A, Diong C, Lee S, Langford B, et al. Prevalence and Mortality Associated with Bloodstream Organisms: a Population-Wide Retrospective Cohort Study. Ledeboer NA, editor. J Clin Microbiol. 2022 Mar 7;e02429-21. 34. Dale AP, Woodford N. Extra-intestinal pathogenic Escherichia coli (ExPEC): Disease, carriage and clones. J Infect. 2015 Dec;71(6):615–26. 35. García-Sánchez JE, García-García MI, García-Garrote F, Sánchez-Romero I. Diagnóstico microbiológico de las infecciones intraabdominales. Enfermedades Infecc Microbiol Clínica. 2013 Apr;31(4):230–9. 36. Spellberg B. Basic principles of antiinfective therapy. In: Mandell, Douglas and Bennet’s: Principles and practice of Infectious Diseases. Elsevier; 2021. p. 211–21. 37. Hecker A, Reichert M, Reuß CJ, Schmoch T, Riedel JG, Schneck E, et al. Intra-abdominal sepsis: new definitions and current clinical standards. Langenbecks Arch Surg. 2019 May;404(3):257–71. 38. Marshall JC, Maier RV, Jimenez M, Dellinger EP. Source control in the management of severe sepsis and septic shock: an evidence-based review. Crit Care Med. 2004 Nov;32(11 Suppl):S513-526. 39. Cecconi M, Evans L, Levy M, Rhodes A. Sepsis and septic shock. Lancet Lond Engl. 2018 Jul 7;392(10141):75–87. 40. Busch LM, Kadri SS. Antimicrobial Treatment Duration in Sepsis and Serious Infections. J Infect Dis. 2020 Jul 21;222(Supplement_2):S142–55. 41. Reitz KM, Kennedy J, Li SR, Handzel R, Tonetti DA, Neal MD, et al. Association Between Time to Source Control in Sepsis and 90-Day Mortality. JAMA Surg [Internet]. 2022 Jul 13 [cited 2022 Jul 14]; Available from: https://jamanetwork.com/journals/jamasurgery/fullarticle/2794180 42. Shahrami B, Sharif M, Sefidani Forough A, Najmeddin F, Arabzadeh AA, Mojtahedzadeh M. Antibiotic therapy in sepsis: No next time for a second chance! J Clin Pharm Ther. 2021 Aug;46(4):872–6. 43. Blandy O, Honeyford K, Gharbi M, Thomas A, Ramzan F, Ellington MJ, et al. Factors that impact on the burden of Escherichia coli bacteraemia: multivariable regression analysis of 2011–2015 data from West London. J Hosp Infect. 2019 Feb;101(2):120–8. 44. Abernethy JK, Johnson AP, Guy R, Hinton N, Sheridan EA, Hope RJ. Thirty day all-cause mor- Bibliography | 153