Optimisation of the therapeutic potential of fosfomycin against Enterobacteriaceae: characterisation of genetic and physiological factors related to resistance and antimicrobial activity

UNIVERSIDAD DE SEVILLA Facultad de Medicina Departamento de Microbiología TESIS DOCTORAL Optimisation of the therapeutic potential of fosfomycin against Enterobacteriaceae: characterisation of genetic and physiological factors related to resistance and antimicrobial activity. Trabajo realizado para alcanzar el grado de Doctor por Dña. Miriam Ortiz Padilla Sevilla, marzo de 2023 UNIVERSIDAD DE SEVILLA Facultad de Medicina Departamento de Microbiología Dr. D. Fernando Docobo Pérez, Profesor Contratado Doctor de Microbiología de la Universidad de Sevilla; y el Dr. D José Manuel Rodríguez Martínez, Profesor Titular de Microbiología de la Universidad de Sevilla, CERTIFICAN: Que la tesis para optar al grado de Doctor por la Universidad de Sevilla que lleva por título “Optimisation of the therapeutic potential of fosfomycin against Enterobacteriaceae: characterisation of genetic and physiological factors related to resistance and antimicrobial activity” ha sido realizada por Dña. Miriam Ortiz Padilla bajo nuestra supervisión, reuniendo todos los requisitos necesarios para su presentación mediante la modalidad de compendio de publicaciones. Y para que conste donde proceda firmamos el presente certificado en Sevilla, marzo 2023. Dr. D Fernando Docobo Pérez Dr. D. José Manuel Rodríguez Martínez UNIVERSIDAD DE SEVILLA Facultad de Medicina Departamento de Microbiología Dr. D José Manuel Rodríguez Martínez, Profesor Titular de Microbiología de la Universidad de Sevilla, CERTIFICA: Que la tesis para optar al grado de Doctor por la Universidad de Sevilla que lleva por título “Optimisation of the therapeutic potential of fosfomycin against Enterobacteriaceae: characterisation of genetic and physiological factors related to resistance and antimicrobial activity” ha sido realizada por Dña. Miriam Ortiz Padilla bajo mi tutela y es apta para ser presentada mediante la modalidad de compendio de publicaciones. Y para que conste donde proceda firmo el presente certificado en Sevilla, marzo 2023. Dr. D. José Manuel Rodríguez Martínez AGRADECIMIENTOS Este es el resultado, mi tesis. Cuesta pensar que tantas horas de trabajo se resuman en las páginas que vienen a continuación, pero claro, aquí solo se cuenta una parte. Y como en toda historia hay mucha gente detrás de ésta. No voy a empezar a contar desde el principio, sería muy largo, aunque desde el principio siempre han estado mi familia. Ellos han apoyado desde el primer momento que me viniera a vivir a un sitio con menos olivos, que viniera a vivir y que quisiera quedarme, acabando por entender que me gustara tanto esta ciudad. Y lo más importante, aguantando el gran repertorio de malas caras que puedo poner cuando estoy agobiada…que han sido muchas veces durante los últimos años; pero como bien dijo mi amiga Irene, estamos en estado “pretésico” y nos convertimos en personas un poco más desagradables de lo que las personas de alrededor se merecen, pero se pasa. Llevo ya muchos años en esta ciudad, solo unos pocos menos en este laboratorio, que ha acabado por ser mi casa, y la gente, un poco mi familia también. Y como por algún lugar hay que empezar, empezaremos a agradecer a Jose Manuel, co-director de esta tesis, que me acogió en su equipo para hacer el TFG (aunque no tenía más remedio), pero ya voluntariamente, para hacer también el TFM. Durante la carrera, suponía que me gustaba la microbiología, pero estar en este laboratorio con Jose Manuel y Paula enseñándome todo lo que dio tiempo, vi que me encantaba, y que mientras la ciencia me dejara, era el camino que quería seguir. Y es que Paula me enseñó como se trabajaba en micro, desde cero, dedicando horas y horas conmigo, saliendo tarde si hacía falta. Mi mano en el laboratorio es gracias a ella, y algunas manías como que cuando una placa lleva kanamicina, se rotula en rojo. Jose Manuel también fue el que me propuso para realizar la tesis, en este caso con Fernando, el otro co-director de esta tesis. Para agradecer todo a Fernando me faltarían páginas. Mi mano en el laboratorio es gracias a Paula, pero en este caso, mi cabeza es gracias a Fernando, porque cuando pasas horas pensando sobre la fosfomicina con alguien, al final eso te enseña a pensar. Horas pensando, resolviendo problemas, sonriendo cada vez que entraba en el laboratorio diciendo que los resultados eran raros, y sonriendo aún más cuando decían que los resultados tenían buena pinta. También tengo que agradecer a Álvaro Pascual la oportunidad de formar parte de su grupo durante tantos años. Y por supuesto a Sofía Ballesta, siempre intentando ayudar en todo, en el trabajo y en la vida. Atendiendo con una sonrisa sin importar la de veces que entremos en su despacho. Tiene que tener un párrafo, aunque para poner todo necesitaría un par canario de párrafos (que puede ser cualquier número par entre 2 y el infinito), y es que Sara ha acabado por ser parte de mi vida, dentro y fuera del laboratorio. Porque es la persona con la que puedo pasar todo el día en el trabajo, ir a entrenar, luego salir y cuando llego a casa, estar una hora mandándonos audios para comentar el día… Muchas gracias al resto del “equipo fosfo”, Inés y Belén, y luego Ana. Dispuestas siempre a pensar conmigo y frustrarnos juntas, y hacer todos los experimentos que se nos ocurrían, incluso de un día para otro. Y al resto del laboratorio los de antes y los de ahora, Ana Duque y Adri, que ya no están en el lab, pero están fuera, que es más importante. A Marinilla, que igual te llama la atención y te pone a leer los protocolos originales, te da un abrazo y te cuida. A Andrea, que sigue viniendo a desayunar los viernes y con la que ojalá hubiese compartido un zulito bioinformático, lleno de fotos de gatos. A Marisa, que siempre ayuda en el lab y además, te manda fotos y videos de Mechi, siempre que es necesario, y si no también los manda, y nunca están de más. A Laurita y María, que han aguantado todo este colapso final, todo el estrés, viendo un poco lo que les espera, y animando como mejor han sabido, aunque fuera comiendo sándwich de nocilla a las 6 de la tarde. A Jose Luis, que viene y hace “ciencia de tarde”, también conocida como cosas absurdas que hay que hacer (medir tubos o calibrar bombas, el Hollow-Fiber tiene esas cosas), no necesitan mucha concentración y te puedes reírte un poco mientras, porque a veces el día ya ha sido muy largo, pero aún no te puedes ir. Y la gente de fuera del laboratorio, Irene y Maribel, que llevan conmigo desde el primer día de carrera, que no siempre tenemos muy claro qué va a ser de nosotras, pero ahí vamos…Siempre puedes mandarles una decena de audios, y tener conversaciones largas que duran días. Y por su puesto a las niñas del jiuji, que aparecieron hace nada, me acogieron y me descubrieron un maravilloso mundo, y ahora son absolutamente necesarias en mi vida… Seguramente me dejo gente y cosas importantes que decir…Pero soy rancia jiennense, así que es complicado escribir estas cosas para mí. Así que a todas las personas que han estado ahí estos últimos años, ayudando en la tesis, o a que se me pase la frustración por la misma, de verdad, gracias. Tesis como compendio de artículos previamente publicados. Esta tesis, de acuerdo con el informe correspondiente y la autorización de los directores de la misma, y del órgano responsable del Programa de Doctorado, se presenta como un compendio de tres artículos. Las referencias completas a los artículos que componen el cuerpo de la tesis son las siguientes: • Ortiz-Padilla M, Portillo-Calderón I, de Gregorio-Iaria B, Blázquez J, Rodríguez-Baño J, Pascual A, Rodríguez-Martínez JM, Docobo-Pérez F. Interplay among Different Fosfomycin Resistance Mechanisms in Klebsiella pneumoniae. Antimicrob Agents Chemother. 2021 Feb 17;65(3):e01911-20. doi: 10.1128/AAC.01911-20. PMID: 33361305; PMCID: PMC8092493. • Ortiz-Padilla M, Portillo-Calderón I, Maldonado N, Rodríguez-Martínez J, de Gregorio-Iaria B, Merino-Bohórquez V, Rodríguez-Baño J, Pascual Á, Docobo-Pérez F. Role of inorganic phosphate concentrations in in vitro activity of fosfomycin. Clin Microbiol Infect. 2022 Feb;28(2):302.e1-302.e4. doi: 10.1016/j.cmi.2021.09.037. Epub 2021 Oct 8. PMID: 34634457. • Ortiz-Padilla M, Portillo-Calderón I, Velázquez-Escudero A, Rodríguez-Baño J, Pascual Á, Rodríguez-Martínez JM, Docobo-Pérez F. Effect of Glycerol on Fosfomycin Activity against Escherichia coli. Antibiotics (Basel). 2022 Nov 12;11(11):1612. doi: 10.3390/antibiotics11111612. PMID: 36421256; PMCID: PMC9686493 4 permeability or the presence of fosfomycin-modifying enzymes. Chromosomal mediated fosfomycin resistance usually occurs in a stepwise mode, often generating complex phenotypes difficult to interpret. In this sense, to better understand the mechanisms of resistance to fosfomycin in Klebsiella pneumoniae and to optimize the use of this antimicrobial, the following study was carried out. The objectives were to characterize the role of the genes uhpT, glpT, and fosA in resistance to fosfomycin in K. pneumoniae and to evaluate the use of phosphonoformate sodium (PPF) due to its ability to inhibit the FosA enzyme, in combination with fosfomycin. For this purpose, seven clinical isolates of K. pneumoniae and the reference strain (ATCC 700721) were used, and their genomes were sequenced. Mutants for transporters and fosA were constructed from two isolates of K. pneumoniae ATCC 700721. The susceptibility test to fosfomycin was performed using the gradient strip method. Synergy between fosfomycin and PPF was studied by checkerboard assay and analyzed with SynergyFinder. Spontaneous frequencies of occurrence of fosfomycin and PPF mutants, in vitro activity by growth curves with gradient concentrations of fosfomycin with and without PPF, and time-kill assays with and without PPF were also evaluated. The fosfomycin MICs of the clinical isolates ranged from 16 to 1,024mg/L. The addition of 0.623 mM PPF reduced the MIC by 2 to 8-times. Deletion of fosA gene led to a 32-fold decrease. Synergistic activities were observed with the combination of fosfomycin and PPF (most synergistic area at 0.623mM). The lowest frequencies of fosfomycin resistant mutants were found in ΔfosA mutants with frequency ranging from 1.69x10-1 to 1.60x10-5 for 64 mg/L fosfomycin. Finally, the growth monitoring and time-kill assays, fosfomycin showed bactericidal activity only against fosA mutants and not with the addition of PPF. The study concludes that inactivation of the fosA gene results in decreased resistance to fosfomycin in K. pneumoniae. The pharmacological approach using PPF did not achieve sufficient activity and the effect decreased with the presence of other fosfomycin resistant mutations. The second chapter of the Thesis follows the line of optimizing the use of fosfomycin with the addition of an adjuvant, and to better understand how the regulation of fosfomycin transporters may affect their activity. The main objective was to evaluate the role of glycerol at therapeutically relevant concentrations in combination with fosfomycin in Escherichia coli, since this molecule is clinically used as a treatment for example for elevated intracranial pressure and can induce glpT expression. For this purpose, a collection of isogenic mutants of fosfomycin-related genes was evaluated in E. coli strains. The induction of fosfomycin transporters was evaluated and susceptibility tests, interaction assays, and time-to-death assays were performed. Our results showed that glycerol allows the activation of the GlpT transporter, but this induction is delayed in time and is not homogeneous in all E. coli strains throughout the bacterial population, leading to contradictory results in terms of fosfomycin activity. 5 The susceptibility assays showed increased fosfomycin activity with glycerol in the disc diffusion assay, but not in the agar dilution or broth microdilution assays. Similarly, in time-kill assays, the effect of glycerol was absent because of the appearance of fosfomycin-resistant subpopulations. In conclusion, glycerol may not be a good candidate for use as an adjuvant to fosfomycin. Finally, to better understand physiological factors that affect fosfomycin transporters activity, the aim of third chapter was to evaluate the in vitro activity of fosfomycin under different physiological concentrations of inorganic phosphate (Pi). For this purpose, the wild-type strain BW25113, four isogenic mutants (∆glpT, ∆uhpT, ∆glpT-uhpT and ∆phoB) and six clinical isolates of E. coli with different fosfomycin susceptibilities were used. Susceptibility was assessed by agar dilution using Mueller-Hinton agar (Pi=1mM) and supplemented with Pi (13 and 42mM, minimum and maximum urinary concentrations of Pi) and/or glucose-6-phosphate (25mg/L). The promoter activity of the fosfomycin transporter was assessed by monitoring fluorescence accumulation using pUA66-PglpT::gfpmut2 or pUA66-PuhpT::gfpmut2 plasmids in standard Mueller-Hinton broth (MHB) supplemented with Pi (13 or 42mM) ± glucose-6-phosphate. Fosfomycin activity was quantified spectrophotometrically at 24 hours as before with glucose-6-phosphate, and fosfomycin ranged from 1 to 1024mg/L. The EC50 of fosfomycin was estimated and compared. Time-kill assays were performed with fosfomycin concentrations of 307 (plasma Cmax), 1053 and 4415mg/L (urinary Cmax range), using MHB with 28mM Pi (mean urinary concentration) +25mg/L glucose-6-phosphate. The results showed that all strains decreased fosfomycin susceptibility linked to increasing Pi concentrations: 1-4-log2 dilution differences from 1 to 13mM, and 1-8-log2 dilution differences at 42 mM Pi. Changes in phosphate concentration did not affect the expression of fosfomycin transporter promoters. Also, increasing Pi concentrations resulted in a higher bacterial viability EC50 of fosfomycin, except against the ΔglpT-uhpT mutant strain. Therefore, the present study concludes that Pi variations in physiological fluids may reduce the activity of fosfomycin against E. coli. Also, the elevated urinary Pi concentrations may explain the failure of oral fosfomycin in non-wild but fosfomycin-susceptible E. coli strains. 6 7 III. INTRODUCTION 1. Bacterial resistance The development of antimicrobials has been one of the greatest advances in the field of health since it has made it possible to cure numerous infectious processes1. After the discovery of penicillin and its possible uses in the 1940s by Alexander Fleming, he himself warned in 1945 that the availability of antimicrobials did not ensure therapeutic success, since microorganisms adapted, becoming resistant to treatment and, therefore, reducing their efficacy2. The emergence of bacterial resistance is a natural phenomenon that is inherent in microorganisms, and efforts have been made to overcome it with the introduction of new antimicrobials. Despite this, soon after their use, microorganisms with resistance mechanisms against them appear1,3. Figure 1. Timeline showing the introduction of antibiotics into clinical practice and the emergence of antimicrobial resistance. Most antibiotic scaffolds were discovered up to the 1970s. Then, most of them were chemically expanded by introducing new functional groups4. The beginning of the study of bacterial resistance occurred when the decrease in the efficacy of antimicrobial treatments began to be detected, although the existence of resistance mechanisms predates the use of these substances1,5,6. The existence of these factors, prior to the introduction of antimicrobial treatments, has been demonstrated in some studies and can be explained by the evolution of microorganisms coexisting in the same natural habitat, accumulating protective and metabolic mechanisms that promote their survival7–9. The set of genes involved in antimicrobial resistance that bacteria have been acquiring is called the "antimicrobial resistome", a term that began to be used in 2006 by Gerry Wright's group in an article on the determinants of resistance present in 8 soil10. Subsequently, the resistome was defined as the set of all genes conferring resistance to antimicrobials and precursors, in pathogenic and non-pathogenic bacteria. This includes all types of resistance genes, both acquired and intrinsic, their precursors, and some potential resistance mechanisms in microbial communities that require evolution or altered gene expression to confer resistance. The study of resistance since then, based on the resistome concept, has confirmed that; (i) antimicrobial resistance is ancient and ubiquitous in microbiomes, (ii) the resistome is complex and diverse, (iii) the environmental resistome is the origin and reservoir of antimicrobial resistance genes, (iv) the resistome is determined by microbial community structure in the natural environment, (v) human activities modify the environmental resistome, (vi) mobile genetic elements are responsible for the transmission of these genes, and (vii) antimicrobial resistance genes flow between humans, animals, and the environment. These findings have served as a basis for approaching a solution to the resistance problem from a "One-Health" point of view, that the study of antimicrobial resistance should be directed at all sectors: human, animal, and environmental, as all are affected by each other10,11. In general, antimicrobial resistance mechanisms can be summarized as: (i) genes captured by gene mobilization and gene transfer (conjugation, transduction, and transformation); (ii) chromosomal mutations, which usually have a biological cost, but can be accompanied by compensatory mutations; (iii) overexpression of intrinsic resistance mechanisms, such as enzymes that inactivate antimicrobial, or efflux pumps12–14. This natural phenomenon has been amplified by the extensive use of antimicrobials by humans, both in medicine and veterinary medicine, as well as their disposal as waste. This has helped the evolution and diffusion of resistance mechanisms both in clinical environments and in natural ecosystems. All this has led to the spread of bacteria resistant to different antimicrobials simultaneously (multidrug-resistant bacteria), especially in clinical settings, although they are increasingly distributed in community environments, making the treatment of infections caused by these pathogens more difficult, since they allow few therapeutic options15–17. These infections are associated with increased mortality and hospital stay, and therefore also increase the economic cost. Therefore, antimicrobial resistance has been identified as a threat to public health and the global economy18. 2. The public health problem of antimicrobial resistance. In the last decade, different studies on infections caused by microorganisms carrying antimicrobial resistance mechanisms have shown the increase in morbidity, mortality, and therefore the increase in the economic cost of infection19,20. 9 The massive use of antimicrobials, both in health care and in industry (mainly livestock) causes the selection and evolution of resistance mechanisms in microorganisms. On the other hand, globalization throughout the world has facilitated the mobility of these microorganisms, leading to an exponential increase in pathogens resistant to existing treatments21,22. Genes encoding resistance mechanisms can be inherited, but they can also be transmitted between bacterial species by mobile genetic elements such as integrons, transposons, or plasmids. In addition, there are what are known as "successful" or "high-risk" clones. These bacterial clones have great relevance in the dispersion of resistance mechanisms because they accumulate and exchange resistance mechanisms effectively, have a high pathogenic, invasive and colonizing capacity, and therefore also have high transmissibility between people.22,23. Figure 2. Global deaths (counts) attributable to and associated with bacterial antimicrobial resistance by infectious syndrome, 201924. All of this is of great epidemiological interest and has been the subject of study. The first review of the global health and economic consequences of bacterial resistance was published in 2014 by economist Jim O'Neil. This review estimates that 700,000 people die each year from infections caused by antimicrobial resistant pathogens and estimates that by 2050 there will be 10 million deaths, making it the leading cause of death worldwide, if the trend of increasing and expanding bacterial resistance continues18. Another major study conducted in 2019, which extracted data on 23 pathogens in 204 territories, showed results on AMR-associated deaths. Estimated, through statistical predictive modelling, there were 4.95 million (3.62-6.57) deaths associated with AMR, including 1.27 million (0.911-1.71) deaths attributable to AMR in 2019. The six leading pathogens for AMR-associated deaths 10 were Escherichia coli, followed by Staphylococcus aureus, Klebsiella pneumoniae, Streptococcus pneumoniae, Acinetobacter baumannii, and Pseudomonas aeruginosa. These pathogens were responsible for 929000 (660 000-1270 000) AMR-attributable deaths attributable to AMR and 3.57 million (2.62-4.78) deaths associated with AMR in 201925. The main cause of the increase in bacterial resistance is the consumption of antimicrobials26, which has been increasing globally, raising by 65% between 2000 and 2015; and if the trend continues, it will increase by around 200% in the next 10 years27. However, antimicrobial consumption and bacterial resistance are not evenly distributed throughout the world. An increase in antimicrobial resistance can be observed in countries with lower socioeconomic and health inequities levels27. This public health problem has managed to become one of the main points on the international agenda, in which various institutions such as the World Health Organization (WHO), the Food and Agriculture Organization of the United Nations, or better known as FAO, and the World Organization for Animal Health (Office International des Epizooties or OIE) have prepared documents proposing solutions to alleviate this threat to public health. In 2015, the WHO approved the Global Action Plan on Antimicrobial Resistance. This document recognizes the magnitude of the problem of antimicrobial resistance worldwide and proposes measures to be carried out, all from a One-Health perspective, i.e., medicine, veterinary medicine, the environment, agriculture, and society in general should be involved11,21. The general measures proposed are: (i) improving awareness of antimicrobial resistance and behavioural change among policy makers, farmers, veterinarians, and health personnel, the food industry and the general public through education, communication and training on the most efficient topic; (ii) strengthening the knowledge and evidence base through research and surveillance; (iii) reducing the incidence of infections in healthcare facilities, farms, and communities, as well as environmental contamination, through effective prevention; (iv) optimizing the use of antimicrobials in human and animal health; eliminating the use of animals for growth promotion; (v) increasing the development of R&D development on new drugs, diagnostics, vaccines, and other interventions related to priority pathogens21. The ultimate objectives of this plan are to maintain reduced levels and decrease the development of resistance in order to maintain the ability to treat infectious diseases with effective and safe drugs. All this would reduce the impact of infectious diseases in humans and animals at the health and economic levels. This is being monitored and evaluated at different levels; regionally, nationally, and globally. FAO, OIE and WHO are individually monitoring activities and results against each organization's plan and budget, collectively tracking progress and reporting on joint work plan activities. The Inter-Agency Coordination Group on Antimicrobial Resistance may recommend additional issues and indicators for monitoring28. 11 In the European framework, in the report published by the European Antimicrobial Resistance Surveillance Network (EARS-Net) in 202125, a network that collects and analyzes bacterial resistance in 29 countries of the European Economic Area, data were reported for eight pathogens: Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, Acinetobacter spp, Streptococcus pneumoniae, Staphylococcus aureus, Enterococcus faecalis, and Enterococcus faecium. The most reported bacterial species was E. coli (39.4%), followed by S. aureus (22.1%), K. pneumoniae (11.9%), E. faecalis (8.8%), E. faecium (6.2%), P. aeruginosa (6.1%), Acinetobacter spp (3.0%) and S. pneumoniae (2.5%). In 2021, the overall number of reported isolates increased compared to 2020 for all bacterial species. The reported status of antimicrobial resistance varied widely depending on bacterial species, antimicrobial group, and geographic region. In general, most bacterial species/antimicrobial combinations in this report showed a significantly decreasing trend or no significant trend during 2017-2021. The exceptions were resistance to carbapenem in E. coli and K. pneumoniae and resistance to vancomycin in E. faecium, for which there was a significant increase during this period. Importantly, in 2021, more than half of the reported E. coli strains and more than one-third of K. pneumoniae strains were resistant to at least one antimicrobial group, and multidrug resistance (combined resistance to several antimicrobial groups) was a frequent occurrence. Resistance rates were generally higher in K. pneumoniae than in E. coli, as in the case of resistance to carbapenems, being higher in K. pneumoniae, but not frequently found in E. coli25. Resistance patterns are not homogeneous; generally, a higher rate of resistance is observed in countries located further south and east of the European economic area25. If we focus on our country during 2020, antimicrobial sensitivity data corresponding to a total of 20,438 isolates of blood cultures and cerebrospinal fluids from the same number of patients have been analyzed, which represents an increase of 6.7% compared to 2019. In the report communicated by EARS-Net, several points can be highlighted. Resistance to all antimicrobials studied in E. coli has increased from 2001 to 2020, and a quarter of strains show resistance to three or bad families of antibiotics. In 10 years, resistance to third generation cephalosporins has increased from 10.2% to 30%, an increase that has been accompanied by other antimicrobials such as fluoroquinolones (30%) and aminoglycosides (17%). Resistance to carbapenems has reached the highest figures, reaching 0.2% for E. coli and 4.7% for K. pneumoniae. Methicillin-resistant strains of S. aureus remain stable and the trend of a slight decrease in resistance in general and beta-lactams in particular has been maintained29. Therefore, antimicrobial resistance is a public health problem that also affects global ecology. For this reason, different international organizations are proposing surveillance and analysis to promote different strategies to solve the problem. 12 3. Strategies for the fight against antimicrobial resistance. The discovery and use of antimicrobials for prophylaxis and treatment have led to great medical advances. But, at the same time, the problem of bacterial resistance was growing and spreading globally, threatening the usefulness of these drugs. For this reason, it became necessary to take measures to control the emergence and spread of resistance mechanisms. These measures must be taken with a One Health approach, that is, in a multidisciplinary way, working together to try to solve or alleviate this problem that concerns human, animal, and environmental health11,30. To this end, in recent years, different organizations have proposed national and international programs whose main measures would be the following31: 3.1. Bacterial resistance monitoring system. The control of antimicrobial resistance requires surveillance, monitoring, and communication of data obtained on resistant microorganisms at the global, international, and national levels. To this end, programs such as the Global Antimicrobial Resistance Surveillance System (GLASS), which collects, analyzes, and exchanges resistance data worldwide21, the aforementioned EARS-Net at the European level32, or the National Plan against Antimicrobial Resistance (PRAN) have been established. The PRAN was approved in Spain in 2014 in which all autonomous communities, different ministries, as well as scientific societies, scientific organizations, and expert collaborators work together33. These plans allow tracking, study of resistance distribution, and analysis of the data obtained to develop specific interventions31. 3.2. Infection prevention. The aim of infection prevention is to reduce the spread of microorganisms in order to avoid or reduce the spread of resistance mechanisms; at the same time, the reduction of infections would allow a reduction in the use of antimicrobials34. In the hospital setting, prevention measures have been shown to have a real impact on reducing bacterial transmission. Measures such as cleaning and decontamination of surfaces or hand washing; in addition to screening and isolation systems for patients carrying microorganisms with resistance mechanisms35. Education of the population in basic hygiene measures, in addition to a good urban waste management system, is fundamental to preventing the transmission of infectious diseases36,37. Another important point is environmental cleanliness, with waste treatment in general, and also with water control, with greater emphasis on hospital wastewater where both, microorganisms and antimicrobials, have been found that would facilitate the selection of resistance mechanisms38. 13 3.3. Optimizing the use of available antimicrobials. Inappropriate use of antimicrobials promotes an increase in resistance to these drugs. It was estimated that, in humans, more than half of the treatments are not appropriate39. To avoid unnecessary, inappropriate, or excessive treatment, it is necessary to know the microbiological epidemiology and up-to-date training to select the most appropriate antimicrobial treatment and dosage. To achieve the optimized use of antimicrobials, PROA programs are implemented40; coordinated by physicians specializing in infectious diseases, microbiologists, pharmacists, epidemiologists, nurses, and other specialists in the affected areas; they aim to optimize treatments, reduce costs, and avoid the emergence of bacterial resistance41. The implementation of these programs has improved the control of bacterial resistance, reduced healthcare-associated infections, and associated hospital costs42, and could save some 4.8 billion dollars a year and up to 1.6 million lives by the year 2050, according to estimates by the OECD (Organisation for Economic Cooperation and Development) together with the CDC (Centers for Disease Control and Prevention) 43. In Spain, experts from scientific societies have established national consensus on how to implement PROA programs at the hospital level40. In Andalusia specifically, in 2013, the Comprehensive Program for the Prevention and Control of Healthcare-Related Infections and the Appropriate Use of Antimicrobials, or also known as the PIRASOA program, was approved44. 3.4. Reduction of the use of antimicrobials for non-therapeutic purposes. Antimicrobials have been used since their discovery in the veterinary field, and although there are antimicrobials exclusively for that use, most are the same or share similar targets as those used for humans45. In addition, in veterinary medicine, antimicrobials have been used for nontherapeutic or prophylactic purposes, since they were found to serve as a dietary supplement to promote growth46. This led to the expansion of the use of antimicrobials in the livestock industry for this purpose47. Numerous studies agree that the use of antimicrobials in food production increases the development of bacterial resistance, endangering its use as a therapeutic treatment, and this has a great ecological impact48. Furthermore, the transmission of the resistance generated and the effect on humans is because animals act as reservoirs for microorganisms that carry resistance mechanisms. Humans come into contact with food, animals, or environmental sources such as water and can acquire the microorganisms in all these ways49. The regulations on the use of antimicrobials as growth promoters varies among the different countries. In Europe, it was banned in 2006, decreasing the sales of antimicrobials for veterinary use by 32.5% in ten years. Additionally, it is intended to restrict the use of some antimicrobials for prophylaxis and medicated feed50. But, for example, China has one of the highest consumptions of antimicrobials 20 adenylate cyclase (AC), the enzyme that produces cAMP116,117. Therefore, PTS sugars cause catabolic repression by decreasing cAMP, negatively affecting the expression of glpT and uhpT115. ▪ Regulation of uhpT o Transporter structure, expression, and transporter activity Structurally, UhpT is a monomer with 12 α-transmembrane helices, matching the general structure of a large number of MFS transporters, such as GlpT. Arginine residues R46 and R275 are essential for the function of this transporter, as the others can be replaced by cysteine or lysine by targeted mutations. These residues (arginine) are highly conserved in this group of proteins118. Regulation of expression is mediated by a two-component system and a kinase119. Loss-of- function mutations in this transporter or in the genes of the regulatory system prevent bacterial growth in media whose only carbon source is G6P120. The twocomponent system is made up of UhpB and C, transmembrane proteins whose mission is to detect the presence of G6P in the periplasmic space (UhpC) and when this occurs, UhpB, which is phosphorylated, is able to yield its phosphate to UhpA, a kinase found in the cytosol120,121. The uhpT promoter has a binding site for phosphorylated UhpA, which acts as an inducer together with the cAMP-CRP complex 122. Different studies have found that phosphorylated UhpA acts as the essential element in the expression of uhpT, and that the cAMP-CRP complex acts as a cofactor of UhpA, increasing induction. Furthermore, there is another cofactor that can cause increased expression of this gene under anaerobic conditions, the FNR regulator also has a binding site to the uhpT promoter and, therefore, can also activate its expression. This regulator should not be the only factor that affects anaerobiosis, since even if fnr is mutated, uhpT still has a certain level of expression under this condition123,124. Reinforcing that the main inducer of uhpT is phosphorylated UhpA, a study shows that, in the absence of the two-component system, UhpBC, which detects the presence of G6P in the medium, UhpA can be activated independently in the presence of pyruvate. This is possible because UhpA can accept the phosphate group from acetyl phosphate, which is highly present during the growth on pyruvate via the Pta-AckA pathway125. However, the inorganic phosphate in the medium hinders transport126,127. Figure 7. UhpT topology. The black circles represent arginines, highlighting the essential ones (R46 and R275). White circles represent lysines 118. 21 Regardless of how it is regulated, the main function of this transporter is to introduce G6P, removing inorganic phosphate from the cytoplasm in an electrically neutral antiport reaction128. In addition, it can also incorporate other molecules: hexose phosphate and fosfomycin107. The integrity of this transporter is very important in the study of the susceptibility to fosfomycin. Early experiments with fosfomycin suggest that glucose-6-phosphate may be present in human tissues at sufficiently high concentrations to induce transcription of the uhpT transporter and thus increase the activity of the antimicrobial against organisms such as E. coli and K. pneumoniae128. Consequently, guidelines from committees such as CLSI and EUCAST state that the sensitivity of fosfomycin should be determined in the presence of 25mg/L glucose-6-phosphate to induce this transporter whenever it or its key regulators are present129 ▪ Regulation of glpT o Structure Structurally, the GlpT transporter in E. coli is a protein with 12 α-transmembrane helices that form 2 domains connected by a loop where the opening allows the translocation of the substrate to the cytoplasm is located. The 12 helices, grouped from 2 domains of 6 helices (N- and C-terminal domains) present a pseudo-symmetry, in agreement with the weak homology presented by the two halves of the protein, indicating a gene duplication event in the evolution of the transporter. The two halves are connected by a long cytoplasmic loop. There are no salt bridges, and there are few hydrogen bonds between the domains, but there are many Van der Waals interactions in the periplasmic part of the protein. Importantly, there is a pore between the N- and C-terminal domains on the cytoplasmic side of the protein, which narrows towards the centre of the molecule. This represents the substrate translocation pore. The electrostatic potential of the GlpT surface shows an electroneutral pore surface, except at the closed end of the pore in the centre of the membrane, where it is positive. This area of strong positive charge is attributed to two conserved arginine residues, Arg45 of helix 1 and Arg269 of helix 7. These two arginines have also been shown to be essential for the activity of the E. coli UhpT protein. These conserved residues appear to form the substrate binding site for binding to the phosphate moiety of a substrate130. Figure 8. Mechanism of GlpT transport. The positions of the arginines of the binding site are indicated. The inorganic phosphate is shown as a small circle, and G3P as a triangle 130. 22 The structure of GlpT and biochemical data suggest that this transporter acts with a single binding site, alternating access with rocker-switch-like movement of the N- and C-terminal domains. Therefore, it requires two main conformations with the substrate binding site facing opposite sides of the bacterial membrane. In one study, it is proposed that inorganic phosphate binds to the arginines at the binding site, bringing the arginines and consequently the N- and C-terminal domains closer together, causing the pore to narrow. This binding would destabilise the interface between the two domains on the periplasmic side, allowing the conformational change. In the periplasmic space, the lower affinity of the protein for inorganic phosphate allows its substitution by G3P, while in the cytoplasm, phosphate substitutes G3P due to concentration130,131. o Transporter expression The glpT gene is part of the glp regulon whose function is to regulate the entry and metabolism of glycerol and G3P. This regulon is made up of several operons (glpABC132, glpTQ133, glpEGR134, glpD135, and glpFKX136 in E. coli), and its main regulator is the GlpR repressor, which binds to the promoter region of the operons and blocks their transcription134. GlpT transcription is induced by its own substrate (G3P), which binds to the GlpR repressor, releasing from the promoter region, thus a null mutation in the glpR gene activates a constutive glpT transcription. In studies by Larson et al., it was shown that G3P binds GlpR with an affinity Kd between 20-50 µM affinity, and that other glycerol-phosphate analogues could also bind to GlpR, while glycerol cannot do it. GlpR blocks the promoters of the different operons but binds strongly to those of glpT and glpD than to glpFK. The end products of lipid biosynthesis and glycolysis may provide the second site for glpT inhibition. On the other hand, inorganic phosphate in the medium hinders transport126,127. Other inducers of the glpT transporter would be the cAMP-CRP complex or the FNR regulator. As explained before, the cAMP-CRP complex has a binding site on the glpT promoter, binding when uptake of secondary carbon sources is needed and inducing its transcription107. In anaerobiosis, as is also the case for uhpT, the FNR regulator can also activate its expression, as it has an fnr site on the promoter. Furthermore, it seems that the binding affinity in the promoter is higher in the case of glpT, compared to uhpT123,124. This higher affinity may be due to the fact that G3P, a substrate transported by GlpT, can be used by the bacteria as an electron donor in anaerobic respiration. Under this condition, G3P is oxidized to dihydroxyacetone phosphate (DHAP) by glycerol-3-phosphate dehydrogenase, which in anaerobiosis situations is encoded by glpABC, and transfers electrons to terminal reductases. This would compensate for the loss of fitness that occurs in anaerobiosis123. 23 o Glycerol-mediated induction Glycerol is a triol that can be used as a carbon source. This molecule can enter the cytosol by passive diffusion, but also has a specific transporter, GlpF. Once inside the bacterium, glycerol is phosphorylated by the glycerol kinase (GlpK), producing intracellular G3P. This endogenous G3P can bind to GlpR, decreasing its affinity for promoters of the glp regulon and, in particular, glpT, and inducing expression137,138. This metabolic pathway in which glycerol is transformed into G3P within the bacterium occurs not only in E. coli, but also in other species such as P. aeruginosa or P. putida. What has also been observed is that the consumption of this carbon source is slow, resulting in a late and prolonged induction139,140. Thus, glycerol can act as an intracellular inducer of glpT expression. Figure 9. Representation of the transcriptional regulation system of the GlpT transporter by indirect induction of glycerol. Figure adapted and modified from Castañeda et al107. o Transporter activity The main function of GlpT is the incorporation of G3P into the bacterium, removing inorganic phosphate. G3P is a molecule that can be used as a secondary carbon source and is essential in the biosynthesis of phospholipids. G3P is part of the backbone of all phospholipid molecules and the polar groups of phosphatidylglycerol and cardiolipin, which are indispensable in membrane formation141. The loss of function in this transporter means that bacteria suffering from it cannot grow on G3P as their sole carbon source142. In addition to G3P, GlpT allows the entry of other molecules such as phosphoenol-pyruvate or fosfomycin, although the affinity of the transporter for G3P is higher143. However, in contrast to activation of the UhpT transporter using G6P, the use of G3P also induces the expression of the GlpT transporter, but this molecule would compete with other substrates like fosfomycin for the occupancy of the transporter143. 24 ▪ Phosphate homeostasis The control and metabolism of the inorganic phosphate plays an essential role for the bacterial physiology. As example, they are found at the hydrophilic ends of amphipathic lipids that form part of membranes; they are also present together with the ribose or deoxyribose sugars, forming a structural part of DNA and RNA. In addition, the cell's energy is based on the hydrolysis of phosphoanhydride bonds between the phosphates that form ATP, and the regulation of many biological systems are mediated by the phosphorylation or dephosphorylation of specific proteins 144. All of this means that there are critical regulation systems to maintain optimal phosphate levels. In the case of E. coli, a range between 1 and 10mM has been observed with the participation of multiple transporter proteins with different expression patterns145,146. The main transporters that bring phosphate into the bacteria are PitA, PitB and PstSCAB systems; on the other hand, the transporters responsible for exporting inorganic phosphate are PitA, PitB, YjbB, GlpT, and UhpT147–150. This homeostasis is mainly controlled by the histidine kinase PhoR and the response regulator PhoB, which have been shown to regulate at least 31 genes, but this system is not directly involved in regulating glpT and uhpT145. Therefore, although the expression of the transporters is not regulated by the phosphate control system, the different concentrations to which the bacteria may be exposed can affect the activity of GlpT and UhpT, and can decrease transport under high phosphate conditions by going against the osmotic gradient. 4.5. Resistance mechanisms. Fosfomycin resistance mechanisms may be intrinsically present in some bacterial species, or they may be acquired mechanisms. In both cases, fundamental resistance mechanisms affect the target (MurA) or the ability of the antimicrobial to reach the cytosol of the bacterium, where the target is located. ▪ Intrinsic resistance There are bacteria that are intrinsically resistant to fosfomycin trough different mechanisms: o Mutation in the murA gene Some bacteria have mutations that produce an amino acid change at position 115, replacing the cysteine to which fosfomycin must bind with another amino acid. These mutations generate functional proteins, but lack the binding site for fosfomycin, leading to resistance to fosfomycin. This amino acid change has been described in bacteria such as Vibrio fischeri151, Mycobacterium tuberculosis152 and Chlamydia spp153. 25 o Peptidoglycan recycling pathways The other main mechanism of resistance to fosfomycin is intrinsic in the use of peptidoglycan recycling pathways. Some bacterial species can use alternative pathways of peptidoglycan synthesis, not requiring that it be created de novo by MurA. These alternative pathways have been described in Acinetobacter baumannii154, Pseudomonas aeruginosa and P. putida155,156. o Other intrinsic mechanisms Listeria monocytogenes is resistant to fosfomycin in in vitro assays, as it is not able to incorporate this antimicrobial through its transporters. However, in vivo, this bacterial species is susceptible, because a virulence factor (hypoxanthine phosphoribosyltransferase) with G6P permease activity (like UhpT) is produced and can be used by fosfomycin to penetrate into the bacterium157. The constitutive decrease in membrane permeability of some species, such as Acinetobacter baumannii, is also considered an intrinsic mechanism of resistance to fosfomycin. This species has a chromosomal gene (abrP), which encodes a peptidase C13 family, whose deletion was able to increase cell membrane permeability and showing slower cell growth rate. This gene can reduce susceptibility to fosfomycin and other antimicrobials such as tigecycline, tetracycline, and chloramphenicol158. ▪ Acquired resistance o Mutations in murA Mutations in murA are exceptional in clinical isolates of E. coli. While no mutations have been found at position 115, the binding site of fosfomycin to MurA, some substitutions have been described in other positions (aspartic acid for asparagine at position 369 and leucine for isoleucine at position 370). Although not at the active site, these may interfere with the binding of fosfomycin to its target159,160. Also, an overexpression of murA gene, has been associated with fosfomycin resistance, but a bacterial fitness cost is observed under this phenotype 161. Mutations in which a deletion at position 717 have also been found in S. aureus, decreasing the sensitivity to fosfomycin162. o Decreased transport of fosfomycin As discussed above, due to the location of the fosfomycin target (MurA), the antimicrobial must reach the bacterial cytoplasm 109. Therefore, resistance to fosfomycin mainly depend on a decrease in Figure 10: Simplified schematic of the peptidoglycan recycling pathway in P. aeuroginosa155. 26 the uptake of fosfomycin. This decrease may occur through the acquisition of mutations that lead to loss of function in the transporters, or it may occur in other genes involved in their regulation107. As mentioned above, the expression of both transporters depends on induction by the presence of their own substrates (G6P and G3P), and on the binding of the cAMP-CRP complex115. The expression of both transporters is also increased by the presence of FNR, a global bacterial regulator that is activated under conditions of anaerobiosis123,163. It is important to note that not all the bacteria species have both transporters. As example, the Enterobacterales E. coli or K. pneumoniae present both permeases, however in the genus Pseudomonas, P. aeruginosa do not have the uhpT gen164, while P. putida does not have the glpT gene139. The promoters of the glpT and uhpT genes present a binding site for the cAMP-CRP complex, which would act as an inducer of expression, so that a decrease in this complex would lead to a lower presence of the transporters and therefore a decrease in the entry of fosfomycin into bacteria causing fosfomycin resistance115. Loss-of-function mutations in the cyaA and ptsI genes, which are associated with decreased cAMP and thus decreased induction of fosfomycin transporters, increase resistance to fosfomycin and/or the frequency of emergence of fosfomycin-resistant mutants165. Because these mutations have a significant biological cost, they are rarely found in clinical isolates166. Also, loss-of-function mutations in fnr, one of the two main regulators in anaerobiosis, increase the resistance to fosfomycin 4-fold over the wild-type phenotype in anaerobiosis in in vitro experiments123. Mutations in uhpT or in the two-component system, uhpBC, or the transcriptional activator, uhpA, show a similar fosfomycin resistance phenotype120. On the other hand, while mutations in glpT do not increase the fosfomycin MIC value, the frequency fosfomycin resistant mutants increase. Furthermore, the combination of loss of function mutations in both transporters results in a greater increase in MIC than in the case of the uhpT mutation alone165. On the other hand, loss-of-function mutations in glpR, would lead to an increase in intracellular fosfomycin due to the resulting glpT overexpression133. ▪ Intrinsic and acquired resistance mechanisms o Antibiotic-modifying enzymes Increased resistance to fosfomycin may also be caused by the existence of enzymes that can modify the structure of the antimicrobial molecule, losing its antimicrobial properties. These enzymes 27 are present in the chromosome of certain microorganisms, in transposons, or plasmids. These can be divide into two groups: metalloenzymes and kinase-type enzymes167. The kinase-type enzymes are encoded by the fomA and fomB genes; their mechanism of action is phosphorylation of the fosfomycin molecule to inactivate it. FomA forms fosfomycin monophosphate and FomB uses this molecule as a substrate to obtain fosfomycin diphosphate. Both enzymes use ATP and Mg2+. They are specific to bacteria that produce fosfomycin, such as Streptomyces wedmorensis and Streptomyces fradiae168. On the other hand, the mechanism of action of metalloenzymes is to break the epoxide ring of the antimicrobial, requiring a metal as a cofactor. They are quite similar to each other, although they have some modifications that allow them to be classified169. - fosB: The fosB gene codes for a metalloenzyme that catalyzes the addition of L-cysteine to fosfomycin, using Mg2+ as a cofactor, resulting in the fosfomycin molecule without bactericidal activity. Regulation of the expression of this gene depends on the extracytoplasmic sigma factor, a factor involved in the induction of resistance mechanisms to other antimicrobials. The gene is present on the Bacillus subtillis chromosome and has been described in plasmids of Staphylococcus spp and Enterococcus spp170. - fosX: The fosX gene codes for a metalloenzyme with hydrolase activity. Its mechanism of action is the addition of a water molecule at the C1 position of fosfomycin. The catalyst in this case would be glutamic acid and requires Mn2+ as a cofactor171. o fosA The fosA gene codes for a glutathione-S-transferase, whose metal cofactor is manganese (Mn2+). Its mechanism of action involves the nucleophilic addition of glutathione and K+ to the epoxide ring of fosfomycin. Therefore, the presence of Mn2+ and K+ is necessary for its action172. Figure 11. Main enzyme mechanisms of fosfomycin resistance244. 28 It was the first fos-type enzyme described and was found in a plasmid in S. marcescens in a plasmid. This gene has been described in the chromosome of bacteria such as Klebsiella spp., Enterobacter spp., S. marcescens and P. aeruginosa. Plasmid-encoded or plasmid-transformed chromosomal fosA has been shown to confer a high level of resistance to fosfomycin. Although it has significant variations in amino acid sequence, the residues forming part of the active site are highly conserved. In Enterobacterales, there are different plasmid encoded fosA subtypes, of which fosA3 is the most frequently detected in clinical isolates. It appears to be originated from the Kluyvera georgiana chromosome and is associated with plasmid transmission. Other important acquired subtypes are fosA4, fosA5, fosA6, fosA8, fosA9 and fosA10173. The IS26 elements, which can mediate the integration of transposition units, are associated with fosA3, along with other resistance genes. This co-location can contribute as a reservoir of the spread of fosfomycin resistance and other resistance genes to clinically relevant pathogens174. Chromosomal fosA is part of Klebsiella spp. genome, being detectable in >99% of K. pneumoniae isolates. Furthermore, strains of K. pneumoniae and K. variicola have been found that, in addition to the chromosomal fosA, a plasmid-encoded fosA have been observed172. Different subtypes of fosA, both plasmid and chromosomal, confer different levels of resistance to fosfomycin, but all contribute to a lower fosfomycin susceptibility175. For example, most strains of K. pneumoniae strains have a MIC considered susceptible but borderline and depend on whether the use of the fosfomycin treatment is for urinary tract infection. Therefore, clinical use of fosfomycin may be limited against these species172. - Sodium phosphonoformate The enzymatic activity of FosA is reduced by sodium phosphonoformate (PPF). This compound is a small molecule (126 Da), pyrophosphate analogue commonly used as an antiviral for the treatment of CMV retinitis in patients with acquired immunodeficiency syndrome (AIDS) and for treatment of acyclovirresistant mucocutaneous HSV infections in immunocompromised patients. Its antiviral activity relies on the inhibition of the pyrophosphate binding site on virus-specific DNA polymerases at concentrations that do not affect cellular DNA polymerases176. PPF was first synthesized by Nylén in 1924 by alkaline hydrolysis of triethyl phosphonoformate although, the synthesis of trisodium phosphono [14C] formate from [14C] phosgene has been reported by Gawell in 1983. It has been found Figure 12. Molecular structure of phosphonoformate 29 that phosphonoformate can be synthesized in vivo by Streptomyces hygroscopicus mutants154 in a series of reactions starting with phosphoenolpyruvic acid. Attempts have been made to test the ability to reduce the enzymatic activity of FosA with other phosphonates derivatives, but only PPF has shown remarkable activity. This is because it is the only one capable of forming a complex with the active site of the metalloenzyme, thus avoiding the addition of glutathione that inactivates fosfomycin177. This inhibitory capacity has been used to show the presence of plasmid encoded fosA in E. coli. This test consists in the addition of 1mg of PPF to the commercial (fosfomycin-G6P) antimicrobial susceptibility discs. If the isolate studied increases the inhibition zone compared to the same disc, but without the addition of PPF, it can be said that fosA is present178. 4.6. Spectrum of action and epidemiology Fosfomycin is considered as a broad-spectrum antimicrobial due to its activity against different bacterial species, including Gram-negative and Gram-positive bacteria. At the same time, it is a molecule whose efficacy is maintained against multidrug-resistant microorganisms108. ▪ Gram-positive spectrum of action Among Gram-positive bacteria, S. aureus shows fosfomycin susceptibility with variability between studies with respect to the percentages of susceptibility to fosfomycin of this microorganism, with values ranging from 33.2% to 100% of the isolates. Also, different species of Enterococcus spp also present a good response to this antimicrobial, but again the results vary according to the different studies108. In a study carried out on a total of 1,847 Gram-positive isolates from a focus other than the urinary tract, a general percentage of resistance to fosfomycin of 69.1% was observed. Analysing these data by the different bacterial species, 99.3% of S. aureus isolates, including MRSA, were susceptible to fosfomycin; however, lower susceptibility percentages were found in coagulase-negative staphylococci (77.5%), Streptococcus pneumoniae (61.9%), Streptococcus pyogenes (40.6%) and other Gram-positive (48.4%)179. Among Gram-positive microorganisms, the bacterial species Staphylococcus saprophyticus, Staphylococcus capitis, and Corynebacterium spp. are considered intrinsically resistant to fosfomycin180. Figure 13. Disk potentiation assay for detection of glutathione S-Transferase (FosA) production in Escherichia coli isolates178. 36 - Resistant (R): a microorganism is classified as resistant when there is a high probability of therapeutic failure, even in a situation of increased exposure to the antimicrobial agent. ▪ Epidemiologic Cut-Off: ECOFF The epidemiological cut-off values, or ECOFF, was established by EUCAST and is determined for a bacterial species against a specific antimicrobial. The ECOFF corresponds to the highest MIC value of isolates of the same species that do not present acquired resistance mechanisms against the antimicrobial and are therefore considered as wild-type strains208. They are determined by studying the MIC distribution curve of values evaluated in a large number of isolates for an antimicrobial by its reference method209. The clinical response to a treatment does not depend on the defined category by the epidemiological cut-off value, but they serve to monitor and detect the appearance of new resistance mechanisms. Mechanisms that may occur despite MIC values below the clinical breakpoints. ECOFFs, like clinical cut-off points, should be reviewed periodically. For example, in 2020, the ECOFF for fosfomycin against E. coli changed from 8 to 4 mg/L and in 2022 the ECOFF for K. pneumoniae, was set at 128 mg/L. However, it is not always possible to establish a cut-off value that separates the population without acquired resistance mechanisms from those with acquired resistance mechanisms210. Figure 20. Distribution of fosfomycin MIC in 2351 E. coli isolates. ECOFF: 4 mg/L with confidence interval of 1-4210. 37 Figure 21. Distribution of fosfomycin MIC in 1396 K. pneumoniae isolates. ECOFF: 128 mg/L with confidence interval of 64-256210. In addition, it is important to keep in mind that the efficacy of an antimicrobial agent depends on the relationship between the MIC of the microorganism and the exposure of the microorganism to the agent in the patient. Thus, the clinical outcome depends on the triangular relationship between MIC, exposure, and efficacy. And in turn, the exposure of the microorganism to the antimicrobial agent in the patient depends on the dose and pharmacokinetic properties of the drug211. ▪ Fosfomycin susceptibility breakpoints According to CLSI and EUCAST, the agar dilution method supplemented with G6P at a concentration of 25 mg/L is the reference method for fosfomycin susceptibility studies and both recommend the addition of 50 μg of G6P for the sensitivity study with disc-diffusion of 200 μg of fosfomycin. However, they differ in the application and interpretation of the susceptibility breakpoints for this antimicrobial against Enterobacterales. EUCAST establishes that the study of susceptibility to fosfomycin by the disc diffusion method and/or techniques that provide MIC values is applicable to E. coli, while only the susceptibility breakpoints according to MIC can be applied to the rest of Enterobacterales. However, CLSI set the fosfomycin susceptibility breakpoints, both obtained by discdiffusion and MIC, exclusively to E. coli of urinary origin200,201. The clinical susceptibility breakpoints for Enterobacterales, EUCAST establishes as susceptible those MIC values ≤8 mg/L and as resistant if >8 mg/L for an oral use of fosfomycin for the treatment of 38 uncomplicated urinary tract infection caused by E. coli. However, it establishes in susceptible category those MIC values ≤32 mg/L and as resistant if the concentration is >32 mg/L for the intravenous use of fosfomycin. On the other hand, CLSI set the categories of susceptible if the MIC is ≤64 mg/L, intermediate for a MIC of 128 mg/L and resistant when it is ≥256 mg/L. Similarly, the susceptibility breakpoints for disc-diffusion techniques differ between committees. According to the CLSI, those growth inhibition halos ≥16 mm are considered as susceptible, intermediate if the diameter is between 13-15 mm and resistant when the size is ≤12 mm. For EUCAST, those isolates with a growth inhibition halo ≥24 mm is considered as susceptible. EUCAST differentiates between the susceptibility breakpoints applied for the intravenous fosfomycin formulation from those applied to oral fosfomycin for the exclusive treatment in uncomplicated UTIs200,201. However, these breakpoints are under revision and will be soon harmonised212. Since 2017, in the breakpoint tables for interpretation of MICs and zone diameters of EUCAST published a section exclusively dedicated to the interpretation of inhibition zones in the fosfomycin disc-diffusion assays. These colonies that appear inside the inhibition zones can be considered as bacterial subpopulations with reduced susceptibility with respect to the rest of the bacterial population, showing a heteroresistance phenotype. However, the document does not mention the possible clinical implication or relevance of this phenomenon. Figure 22. Examples to correct analysis and interpretation of inhibition zones in fosfomycin disc-diffusion assays against E. coli200. Contrary, CLSI has no interpretation for the inhibition zones in the presence of a heteroresistance phenotype for the study of fosfomycin susceptibility. CLSI simply states in general terms that in the presence of bacterial heteroresistance to an antimicrobial, the study with molecular techniques "may" be more appropriate than the phenotypic sensitivity study213. 39 4.9. Subpopulations The usual view of homogeneity in the metabolism of prokaryotic populations has been increasingly challenged in recent years, especially since the advent of single-cell technologies214. These methodologies have revealed the existence of a complete repertoire of responses to specific environmental conditions in individual microorganisms within the same population215. The diversification of metabolic pathways in single cells within otherwise clonal populations can be considered as a particular case of phenotypic variation, where different regulatory or epigenetic traits lead to the stochastic manifestation of alternative traits in isogenic individuals216,217. This allows phenomena such as persistence, i.e., the appearance of a fraction of living but nongrowing cells in a bacterial pool under a given circumstance, to occur, which is one of the most intriguing cases of phenotypic variation. Persistence ensures the survival of cells exposed to agents that act on growing bacteria, e.g., some antibiotics. Once the selective pressure is stopped, persistent bacteria can resume growth and completely rebuild the original population218. Also, the existence of subpopulations allows the phenomenon of heteroresistance219. Regardless of the mechanisms underlying these behaviours, the question that arises is whether these population differences are an adaptive trait or simply a chance occurrence that is beneficial for bacteria that are susceptible to the antimicrobial agents139. ▪ Heteroresistance The generalized definition of heteroresistance is the presence of a heterogeneous population of bacteria with one or more subpopulations with a decreased susceptibility to the antibiotic compared to the main population219. The genetic mechanisms leading to heteroresistance are multiple, and not all of them are well known. It should be noted that this definition encompasses a multitude of different underlying factors that bring different connotations to the heteroresistant bacterial population such as: the clonal origin of the bacterial population, the level of resistance of the subpopulations compared to that of the main population, the frequency of resistant subpopulations, and the stability of the heteroresistant phenotype (Figure 23)220. Depending on the clonal origin of the bacterial population, heteroresistance may be polyclonal or monoclonal. A bacterial clone is defined as those bacterial cells that arise from a single isolated CFU. Thus, polyclonal heteroresistance is defined when the heterogeneity in resistance of the bacterial population is due to the presence of different bacterial clones, genetically different, one susceptible, and the other resistant. In polyclonal heteroresistance, when the different bacterial clones are separated, isolated and purified in CFU, the heteroresistance phenomenon disappears. However, in monoclonal heteroresistance, the heterogeneity of resistance is generated from a single bacterial clone that differentiates into two different populations, one susceptible and the other resistant in the 40 absence of antibiotic pressure. The difference from the previous case is that the frequency with which resistant mutant populations spontaneously appear from the general susceptible population is very high. Therefore, a CFU contains millions of individual bacterial cells that has been formed from the division of a single bacterial cell, from a single clone, will in turn be composed of a population variety of resistant and susceptible cells. This heteroresistance phenotype will be maintained even when pure clones are analyzed as opposed to polyclonal220. Figure 23. Factors to consider when defining heteroresistance220. Considering the level of resistance of the subpopulations with respect to the majority population, it can be said that most bacterial isolates can present subpopulations that grow at antibiotic concentrations close to their MIC. But there is an absence of criterium to define heteroresistance phenotype. Some studies define bacteria as heteroresistant when the MIC of the subpopulations are higher than 2 to 8 (base 2 serial dilutions) above the MIC of the majority population. Other definitions 41 include the presence of populations in which the resistant subpopulation exceeds the clinical susceptibility breakpoints established for the antibiotic220. The proportion or frequency of resistant subpopulations within the majority population is also a factor to be consider when defining an isolate as heteroresistant. This frequency indicates the number of resistant bacterial cells (or cells with decreased sensitivity) in the population. In the monoclonal heteroresistance, both the susceptibility and concentrations of the antibiotic used to study the resistant subpopulations and the minimum proportion established to consider these subpopulations as significant, will influence the classification of the isolate as heteroresistant. In the latest review on heteroresistance by Andersson et al., they propose that the frequency of bacterial subpopulations should be reported at a concentration of antibiotic 8 times higher than the MIC of the main population220. The stability of the resistant subpopulations is another factor to be considered when defining heteroresistance, since this phenomenon can be stable or unstable and can affect the ability to be detect. Unstable heteroresistance is defined when resistant subpopulations can become susceptible again in the absence of antibiotic pressure after 50 generations in the absence of antibiotic220. The frequency of heteroresistance in clinical isolates depends on the definition used or the detection method; therefore, a great variability of heteroresistance rates is observed, ranging from 0% to 100% for the same combination of antibiotics and bacterial species, depending on the study. The frequency of heteroresistance to fosfomycin in clinical isolates has been studied in different bacterial species, and the same heterogeneous results have been observed194,221,222,223,224. 42 IV. HYPOTHESES AND OBJECTIVES The problem of antimicrobial resistance is limiting the treatability of infections caused by bacteria. Numerous strategies have been proposed to solve this problem, including the reintroduction of antimicrobials in disuse and the optimization of treatments. Therefore, more knowledge is needed on older antimicrobials, antibiotics to which bacteria remain sensitive, such as fosfomycin. These studies would allow an optimization of their use, in single treatment and in combination with other antimicrobials or adjuvants. Therefore, the hypotheses and objectives of this work are the following: 1. Hypothesis I. Fosfomycin resistance in Klebsiella pneumoniae is explained by the presence of the fosfomycin inactivating enzyme chromosomal FosA together with inactivation of the fosfomycin transporter GlpT and UhpT. II. Fosfomycin susceptibility in K. pneumoniae can be restored using sodium phosphonoformate as FosA inhibitor. III. Fosfomycin activity can be enhanced activating glpT transcription increasing glycerol-3- phosphate intracellular concentrations with glycerol in E. coli. IV. Inorganic phosphate extracellular concentrations modify E. coli fosfomycin susceptibility at human physiological concentrations. 2. Objectives I. To study the interaction between the main determinants of fosfomycin resistance in K. pneumoniae. II. To evaluate the activity of sodium phosphonoformate (PPF) as a FosA inhibitor in K. pneumoniae. III. To determine the effect of glycerol as an inducer of the glpT transporter at therapeutic concentrations and to understand the effect of glycerol on the activity of fosfomycin at therapeutic concentrations against E. coli. IV. To study whether the physiological concentrations of phosphate can affect the transport of fosfomycin in E. coli. 43 V. PUBLICATIONS 44 1. Interplay among Different Fosfomycin Resistance Mechanisms in Klebsiella pneumoniae. Infections with extended-spectrum beta-lactamase (ESBL)-producing Klebsiella pneumoniae are of particular concern. Fosfomycin is an antimicrobial currently approved in several countries for the treatment of uncomplicated urinary tract infections caused by Enterobacteriaceae, with activity against multidrug-resistant strains of K. pneumoniae. However, the study of fosfomycin resistance determinants in K. pneumoniae has not been sufficiently studied. Therefore, the first article of the work aimed to determine the in vitro interaction between the main determinants of fosfomycin resistance and fosfomycin resistance in K. pneumoniae. In addition, we also evaluated the possibility of using PPF as an adjuvant of fosfomycin in the treatment against K. pneumoniae. The results obtained in this work are the following: Fosfomycin resistance-related gene sequences. The results for fosfomycin resistance-related proteins (GlpT, UhpT, FosA, CyaA, Crp, PtsI, MurA, UhpA, UhpB, and UhpC) after de novo sequencing of the 12 K. pneumoniae isolates are shown in Table 1 (Art.1). No amino acid changes in CyaA, Crp, MurA, or UhpA were found in these clinical isolates compared with K. pneumoniae ATCC 700721. Apart from polymorphisms with activity whose significance is unknown, GlpT insertions were found in Kp58, Kp88 and Kp108, and deletions were detected in Kp86. The clinical isolate Kp88 also showed a truncated UhpT protein due to a premature stop codon. Fosfomycin sensitivity. The fosfomycin MICs for the isogenic collection and clinical isolates are shown in Table 1 (Art.1). The modal MIC of the clinical isolates was 32 mg/L, and the MICs of fosfomycin ranged from 16 to 1,024 mg/L. For isogenic mutants, similar results were obtained for K. pneumoniae ATCC 700721, Kp12 and Kp142. The DglpT mutant showed no increase in MIC relative to that of the wild-type (WT) strain (32 mg/L), whereas inactivation of uhpT led to a 32-fold increase. Deletion of fosA led to a change in MIC 32-fold decrease relative to wild-type (1 mg/L). Addition of 0.623mM PPF reduced the MIC values of fosfomycin against all clinical isolates by 2- to 8-fold compared to no PPF. ΔuhpT and ΔfosA strains from the isogenic collection showed no decrease in fosfomycin MIC with the addition of PPF. Furthermore, the MICs of fosfomycin against Kp12 and Kp142 ΔfosA strains were 4-fold and 8-fold higher, respectively. Synergy assays. The combination of fosfomycin and PPF showed synergistic activity against K. pneumoniae ATCC 700721, Kp12 and Kp142 strains (Fig. 1, Art.1). According to the synergy score (ZIP), the mean (and 45 maximum) bacterial response ratios attributable to the drug interaction were 15.46 (42.96), 26.88 (65.15) and 30.7 (65.77) for K. pneumoniae ATCC 700721, Kp12 and Kp142, respectively. This maximum synergistic zone occurred at fosfomycin ranges of 2 to 32 mg/L, and at PPF ranges of 0.3 to 2.5mM. Mutant frequencies. The results of mutant frequency estimates are shown in Fig. 2 (Art.1). The lowest frequencies of fosfomycin-resistant mutants were found among the ΔfosA mutants with a decrease in frequency from 1.69 x 10-1 to 1.60 x 10-5 at 64 mg/L and from 5.38 x 10-3 to 2.13 x 10-5 at 512 mg/L relative to their wild-type strains. In terms of fosfomycin mutant selection, little or no differences were observed between ΔglpT and wild-type strains. The ΔuhpT mutation showed a frequency close to 1, confirming the existence of a uniform population with an MIC above the selection conditions of fosfomycin. The addition of PPF had no effect on mutant frequencies (Fig. S1, Art.1). Bacterial growth monitoring. Figure 3 and Fig. S2 (Art.1) show the results of the 24-h growth monitoring assays. With respect to the isogenic collection of mutants (Fig. 3, Art.1), the ΔfosA strains showed the highest susceptibility after exposure to fosfomycin, with no growth at concentrations of 32, 8 and 4 against K. pneumoniae ATCC 700721, Kp12 and Kp142, respectively. High viability of the ΔglpT and ΔuhpT mutants was observed across the range of fosfomycin concentrations tested. Finally, the addition of PPF did not reduce the viability of any mutant except the Kp12 ΔglpT mutant. With respect to clinical isolates, in susceptible strains without known fosfomycin resistance-related mutations, viability was not observed at 256, 256 and 128 mg/L, respectively (Fig. S2, Art.1). The addition of PPF showed an increase in fosfomycin activity, reducing by 3 log2 dilutions the concentration of fosfomycin capable of killing viable bacteria. Although the isolates were susceptible but had mutations in GlpT, none of the tested fosfomycin concentrations were able to eradicate bacterial growth. Similar results were obtained with the highly resistant strains. No differences in viability were observed with the addition of PPF against isolates with mutations or resistant. Time-kill assays. K. pneumoniae ATCC 700721, Kp12, Kp142, and their derivatives at two concentrations of fosfomycin (64 and 307 mg/L), with and without 0.623mM PPF. The results of the assays are shown in Table 2 and Figs. S3, S4 and S5 (Art.1). Heterogeneous bactericidal effect of fosfomycin was observed at both concentrations studied, with a bactericidal effect or decrease in bacterial load during the first hours of the study, in all strains except for the uhpT gene mutant. In all cases (except Kp142 at 307mg/L and mutants for the fosA gene) there is regrowth at the end of the assay. The bactericidal effect was not increased by the addition of PPF. The present study evaluated the role of specific fosfomycin resistance determinants present in K. pneumoniae and their impact on the emergence of highly resistant strains during in vitro fosfomycin exposure. It also investigated a potential therapeutic approach using the FosA inhibitor sodium phosphonoformate, aimed at increasing fosfomycin susceptibility in K. pneumoniae clinical isolates. In our isogenic K. pneumoniae collection of fosfomycin resistance-related genes, a baseline fosfomycin MIC of 32 mg/liter was observed for ATCC 700721, Kp12, and Kp142. Inactivation of glpT did not increase fosfomycin resistance relative to that of the wild-type strains. This phenotype has previously been observed in E. coli due to the use of glucose-6-phosphate in susceptibility assays (1, 17). Under these conditions, uhpT expression masks other fosfomycin resistance-related mutations, especially those related to the glycerol-3-phosphate transporter. On the other hand, inactivation of uhpT significantly increased the fosfomycin MIC ($1,024 mg/liter). The likely explanation for the reduced susceptibility is reduced uptake together with fosfomycin inactivation by the chromosomally mediated FosA. Similar susceptibility results were found among fosA mutants. The inactivation of this chromosomal resistance determinant produced a 32-fold reduction in the fosfomycin MIC with respect to that of the wild-type strains. This change in susceptibility matches the modal MIC observed in the EUCAST fosfomycin MIC distribution found in E. coli (1 mg/liter and epidemiological cutoff [ECOFF] of 4 mg/liter), which lacks the chromosomal fosA gene (9, 18). The addition of PPF increased fosfomycin susceptibility in a variable way, although no activity was found in the uhpT mutant. This effect would not be explained by an increase in mutant frequency caused by PPF. Nevertheless, no differences in selection of rifampicin-resistant mutants were observed. Another plausible explanation is that PPF may compete with fosfomycin for transportation via the UhpT transporter, and this effect would be more evident in a strain lacking fosA, although this hypothesis was not studied. With respect to the pharmacological approach for inhibiting FosA activity, a synergy study was performed combining fosfomycin and PPF. The maximum synergistic area occurred within the fosfomycin susceptibility range (2 to 8 mg/liter for K. pneumoniae ATCC 700721 and 8 to 32 mg/liter of fosfomycin for Kp12 and Kp142), using EUCAST susceptibility breakpoints for E. coli (19). This synergistic activity was also promoted with the addition of PPF in a range from 0.3 to 2.5 mM, which corresponds to human blood concentrations after administration of 90 mg/kg of body weight every 12 h (q12h) (mean 6standard deviation [SD] for steady-state maximum concentration of drug in serum [C max ], 0.623 60.132 mM) (20). Similar results were obtained by Ito et al. using agar dilution assays (12), in which they found an apparent lack of activity of PPF at concentrations of up to 0.667 mM against some clinical strains or transformants with high fosfomycin MIC values. The growth monitoring assays showed increased fosfomycin activity when fosA was absent relative to the wild-type strain. Despite showing better fosfomycin activity with the addition of PPF, these results did not improve those obtained with DfosA strains. These discrepancies could be explained as the inefficient activity of PPF blocking FosA or the degradation or inactivation of PPF during the assay. With respect to the DglpT mutants, fosfomycin activity was lower than for the wildtype strains, despite having similar MICs. These results were similar to those previously obtained by our group in E. coli, whereby highly resistant mutants emerged from DglpT strains following the selection of additional mutations in loci associated with fosfomycin resistance (1). These results also fit the higher fosfomycin mutant frequencies observed in DglpT strains than in wild-type strains. No improved effects were observed against DuhpT mutants when the assays were supplemented with PPF. Comparable results were obtained with the other clinical isolates. The effect of PPF was observed in strains with low MICs and in the absence of fosfomycin resistancerelated mutations, whereas a lack of PPF efficacy was observed when fosfomycin MIC was high or DglpT was present. Overall, the time-kill assays showed results similar to those found in the growth assays. Inactivation of fosA greatly improved fosfomycin efficacy with bacterial Fosfomycin Resistance in Klebsiella pneumoniae Antimicrobial Agents and Chemotherapy March 2021 Volume 65 Issue 3 e01911-20 aac.asm.org 7 on March 5, 2021 at USE/BCTA.GEN UNIVERSITARIAhttp://aac.asm.org/Downloaded from reductions of between 1 and 5log 10 CFU/ml. The addition of PPF to the time-kill assays improved fosfomycin activity only in the first hours, but the emergence of highly resistant mutants occurred in all assays. Our study differs slightly from the results obtained by Ito et al. (12). In that study, using a single K. pneumoniae strain with fosfomycin MIC of 256 mg/liter, the authors found a bactericidal effect at 8 and 24 h at concentrations of 256 and 512 mg/liter but bacterial regrowth at 24 h with 128 mg/liter. The addition of PPF to this assay showed a bactericidal effect at all concentrations of fosfomycin tested, including 128 mg/liter. In conclusion, strategies aimed at inactivating fosA activity through gene editing (21) or using pharmacological approaches are promising for increasing fosfomycin activity against K. pneumoniae strains. In this respect, the pharmacological route is the most plausible one, but further PK/PD models should be carried out to better assess the activity of FosA-inactivating compounds. MATERIALS AND METHODS Bacterial strains. Seven clinical multidrug-resistant K. pneumoniae strains from the Andalusian Reference Laboratory for Molecular Typing of Nosocomial Pathogens (PIRASOA program) categorized as susceptible or resistant to fosfomycin by MicroScan WalkAway (Beckman Coulter, Indianapolis, IN) were used. K. pneumoniae ATCC 700721 was used as the reference strain. The two clinical isolates (Kp12 and Kp142) with the lowest fosfomycin MIC values and the reference strain K. pneumoniae ATCC 700721 were selected for isogenic mutant construction. Whole-genome sequencing. The 7 clinical isolates were subjected to whole-genome sequencing on the MiSeq platform (Illumina, San Diego, CA), and sequencing was performed in-house. For this, the libraries were prepared with the Nextera XT DNA library preparation kit, and sequencing was done with a V3 600-cycle reagent cartridge. Sequencing was achieved with at least 30average coverage. Illumina sequences were assembled de novo using the CLC genomics Workbench (Qiagen, Netherlands). The genomes were annotated with Rapid Annotation using Subsystem Technology (RAST) (22). The fosfomycin resistance-related proteins (MurA, GlpT, UhpT, FosA, CyaA, Crp, PtsI, UhpA, UhpB, and UhpC) were compared with K. pneumoniae ATCC 700721 using the NCBI BLAST online application. Isogenic mutant construction. DfosA,DglpT, and DuhpT mutants were constructed using the l -red gene replacement method, described by Huang et al. (23). All the required primers and plasmids are listed in Table S1. Briefly, the genes of interest (fosA,glpT, and uhpT) were removed by homologous recombination using an apramycin resistance cassette flanked by two 60-bp homologous sequences of the desired gene using the L-arabinose (Sigma-Aldrich, Madrid, Spain)-inducible recombinase of the pACBSR-Hyg plasmid. FLP recombination target (FRT) sites were also included in the construction to remove the apramycin resistance cassette (using pFLP-Hyg) from the chromosome once the mutants had been constructed. All the mutants were confirmed by PCR with sequence-specific primers for each region. Susceptibility testing. Fosfomycin in vitro activity was determined by the gradient strip method (Liofilchem Diagnostici, Italy) using Mueller-Hinton agar (MHA). EUCAST fosfomycin susceptibility breakpoints for Enterobacterales were used (19). In addition, fosfomycin susceptibility testing was performed in Mueller-Hinton agar containing 0.623 mM PPF (Clinigen Healthcare, Staffordshire, UK). This concentration was selected on the basis of the maximum synergy results obtained in the synergy assay and corresponding to the C max values for dosing of 90 mg/kg q12h (20). Assays were performed in duplicate. Synergy assay. Synergies between fosfomycin and PPF were studied against Kp12, Kp142, and K. pneumoniae ATCC 700721 using the checkerboard assay. Briefly, the synergy assay was performed with an inoculum of 5 10 5 CFU/ml in 96-well plates with a final volume of 100 m l per well. Assays were conducted using Mueller-Hinton broth (MHB) supplemented with 25 mg/liter of glucose-6-phosphate. Fosfomycin concentrations ranged from 0.5 to 512 mg/liter and PPF concentrations from 0.16 to 10 mM. Wells without fosfomycin or PPF were used as single-drug assays or growth controls. Bacterial densities were quantified spectrophotometrically by measuring optical density (OD) at 595 nm using an Infinite200 Pro plate reader at 24 h. Bacterial viability was calculated as the ratio of final bacterial OD to the final bacterial OD of the control growth well (without drug). A four-parameter log logistic model was fitted to the data to generate dose-response curves for fosfomycin and PPF. Drug combination responses were also plotted as heat maps to assess the therapeutic significance of the combination by identifying the concentrations at which the drug combination had maximum effect on bacterial growth inhibition. The degree of drug synergy over the full dose-response matrix was analyzed using the response surface model, zero interaction potency (ZIP) with the SynergyFinder package for R (24). Synergy assays were performed in triplicate. The summary synergy represents the average excess response due to drug interactions. A synergy score of less than 210 was considered antagonistic, a range from 210 to 110 as additive, and greater than 110 as synergistic (25). Mutant frequencies. Fosfomycin-resistant mutant frequencies were assessed for K. pneumoniae ATCC 700721, Kp12, Kp142, and their isogenic mutants. As PPF is a pyrophosphate analogue, its role in interfering with the replication process was assessed by analyzing the frequency of rifampicin-resistant Ortiz-Padilla et al. Antimicrobial Agents and Chemotherapy March 2021 Volume 65 Issue 3 e01911-20 aac.asm.org 8 on March 5, 2021 at USE/BCTA.GEN UNIVERSITARIAhttp://aac.asm.org/Downloaded from spontaneous mutants in wild-type K. pneumoniae and E. coli MG1655 strains (as a control due to the absence of chromosomal fosA). Mutant frequencies were determined in triplicate as previously described (1). Fosfomycin-resistant mutants were recovered in MHA plates supplemented with 25 mg/liter of glucose-6-phosphate (Sigma- Aldrich) and two concentrations of fosfomycin (64 and 512 mg/liter; Sigma-Aldrich). Rifampicin-resistant mutants were recovered after overnight incubation of the bacterial culture with and without PPF (0.623 mM) and then plated on MHA with 100 mg/liter of rifampicin (Sigma-Aldrich). Mutant frequencies were the ratio of mutant to total CFU. Bacterial growth monitoring. Bacterial growth curves were performed with 5 10 5 CFU/ml as the starting inoculum, using 96-well plates (Nunclon Delta Surface; Thermo Scientific, MA) with a 200- m l volume per well. Assays were conducted using MHB supplemented with 25 mg/liter glucose-6-phosphate, with or without 0.623 mM PPF. Fosfomycin concentrations ranged from 0.25 to 256 mg/liter. Bacterial growth was quantified spectrophotometrically (595 nm) every hour, for 24 h, with an Infinite200 Pro plate reader (Tecan Group AG, Männedorf, Switzerland). Assays were performed in triplicate. Time-kill assays. Time-kill assays were conducted in duplicate using fosfomycin concentrations of 0 (as a growth control), 64, and 307 mg/liter (maximal concentration of fosfomycin in plasma after intravenous administration of 8 g q8h) (26) with 25 mg/liter of glucose-6-phosphate, with or without 0.623 mM PPF. Briefly, single overnight colonies of each strain were used to prepare the preinoculum in Mueller- Hinton II broth and incubated overnight with shaking at 37°C. The starting inoculum was set at 5 10 5 CFU/ml in a final volume of 20 ml, and bacterial cultures were incubated at 37°C with shaking. The number of viable CFU was determined at 0, 2, 4, 8, and 24 h by serial dilution, followed by plating on MHA plates with or without 64 mg/liter fosfomycin and 25 mg/liter of glucose-6-phosphate. The number of colonies was counted after 24 h of incubation. When bacterial regrowth was observed after 24 h, up to 3 colonies were selected to assess fosfomycin MIC gradient strips. The survivors were serially passaged three times on fosfomycin-free plates to assess the stability of the phenotype and the fosfomycin MIC was determined, as previously described. SUPPLEMENTAL MATERIAL Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 0.8 MB. ACKNOWLEDGMENTS Miriam Ortiz-Padilla is supported by a PFIS fellowship from Instituto de Salud Carlos III. This study was supported by the Plan Nacional de I1D1i 2013-2016 and Instituto de Salud Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Economía, Industria y Competitividad, Spanish Network for Research in Infectious Diseases (PI16/01824, REIPI RD12/0015/0010, and REIPI RD16/0016/0001), cofinanced by European Development Regional Fund A Way To Achieve Europe, Operative Program Intelligent Growth 2014-2020, and the Consejería de Igualdad, Salud y Políticas Sociales, Junta de Andalucía (PI-0044 to 2013). The funders had no role in the design, collection of data, analysis and writing of the manuscript, or the decision to publish. J.R.-B. has been scientific coordinator of a research project unrelated to the project funded by AstraZeneca and a speaker at accredited educational activities funded by Merck through unrestricted grants. J.R.-B. and A.P. received funding in kind for research from COMBACTE-NET (grant agreement 115523), COMBACTE-CARE (grant agreement 115620), and COMBACTE-MAGNET (grant agreement 115737) projects under the Innovative Medicines Initiative (IMI), the European Union and EFPIA companies. REFERENCES 1. Ballestero-Téllez M, Docobo-Pérez F, Portillo-Calderón I, Rodríguez-Martínez JM, Racero L, Ramos-Guelfo MS, Blázquez J, Rodríguez-Baño J, Pascual A. 2017. Molecular insights into fosfomycin resistance in Escherichia coli. J Antimicrob Chemother 72:1303–1309. https://doi.org/10.1093/jac/dkw573. 2. Prestinaci F, Pezzotti P, Pantosti A. 2015. Antimicrobial resistance: a global multifaceted phenomenon. Pathog Glob Health 109:309–318. https://doi .org/10.1179/2047773215Y.0000000030. 3. Docobo-Pérez F, Drusano GL, Johnson A, Goodwin J, Whalley S, Ramos- Martín V, Ballestero-Tellez M, Rodriguez-Martinez JM, Conejo MC, Van Guilder M, Rodríguez-Baño J, Pascual A, Hope WW. 2015. Pharmacodynamics of fosfomycin: insights into clinical use for antimicrobial resistance. Antimicrob Agents Chemother 59:5602–5610. https://doi.org/10 .1128/AAC.00752-15. 4. Falagas ME, Maraki S, Karageorgopoulos DE, Kastoris AC, Mavromanolakis E, Samonis G. 2010. Antimicrobial susceptibility of multidrug-resistant (MDR) and extensively drug-resistant (XDR) Enterobacteriaceae isolates to fosfomycin. Int J Antimicrob Agents 35:240–243. https://doi.org/10.1016/j .ijantimicag.2009.10.019. 5. Pitout JD, Laupland KB. 2008. Extended-spectrum b -lactamase-producing Enterobacteriaceae: an emerging public-health concern. Lancet Infect Dis 8:159–166. https://doi.org/10.1016/S1473-3099(08)70041-0. 6. Veve MP, Wagner JL, Kenney RM, Grunwald JL, Davis SL. 2016. Comparison of fosfomycin to ertapenem for outpatient or step-down therapy of extended-spectrum b -lactamase urinary tract infections. Int J Antimicrob Agents 48:56–60. https://doi.org/10.1016/j.ijantimicag.2016.04.014. 7. Castañeda-García A, Blázquez J, Rodríguez-Rojas A. 2013. Molecular Fosfomycin Resistance in Klebsiella pneumoniae Antimicrobial Agents and Chemotherapy March 2021 Volume 65 Issue 3 e01911-20 aac.asm.org 9 on March 5, 2021 at USE/BCTA.GEN UNIVERSITARIAhttp://aac.asm.org/Downloaded from mechanisms and clinical impact of acquired and intrinsic fosfomycin resistance. Antibiotics (Basel) 2:217–236. https://doi.org/10.3390/ antibiotics2020217. 8. Marger MD, Saier MH. 1993. A major superfamily of transmembrane facilitators that catalyse uniport, symport and antiport. Trends Biochem Sci 18:13–20. https://doi.org/10.1016/0968-0004(93)90081-W. 9. Ito R, Mustapha MM, Tomich AD, Callaghan JD, McElheny CL, Mettus RT, Shanks RMQ, Sluis-Cremer N, Doi Y. 2017. Widespread fosfomycin resistance in gram-negative bacteria attributable to the chromosomal fosA gene. mBio 8:e00749-17. https://doi.org/10.1128/mBio.00749-17. 10. Li YY, Zheng B, Li YY, Zhu S, Xue F, Liu J. 2015. Antimicrobial susceptibility and molecular mechanisms of fosfomycin resistance in clinical Escherichia coli isolates in mainland China. PLoS One 10:e0135269. https://doi.org/10 .1371/journal.pone.0135269. 11. Nakamura G, Wachino JI, Sato N, Kimura K, Yamada K, Jin W, Shibayama K, Yagi T, Kawamura K, Arakawa Y. 2014. Practical agar-based disk potentiation test for detection of fosfomycin-nonsusceptible Escherichia coli clinical isolates producing glutathione S-transferases. J Clin Microbiol 52:3175–3179. https://doi.org/10.1128/JCM.01094-14. 12. Ito R, Tomich AD, McElheny CL, Mettus RT, Sluis-Cremer N, Doi Y. 2017. Inhibition of fosfomycin resistance protein FosA by phosphonoformate (foscarnet) in multidrug-resistant gram-negative pathogens. Antimicrob Agents Chemother 61. https://doi.org/10.1128/AAC.01424-17. 13. Abbott IJ, Meletiadis J, Belghanch I, Wijma RA, Kanioura L, Roberts JA, Peleg AY, Mouton JW. 2018. Fosfomycin efficacy and emergence of resistance among Enterobacteriaceae in an in vitro dynamic bladder infection model. J Antimicrob Chemother 73:709–719. https://doi.org/10.1093/jac/ dkx441. 14. Elliott ZS, Barry KE, Cox HL, Stoesser N, Carroll J, Vegesana K, Kotay S, Sheppard AE, Wailan A, Crook DW, Parikh H, Mathers AJ. 2019. The role of fosA in challenges with fosfomycin susceptibility testing of multispecies Klebsiella pneumoniae carbapenemase-producing clinical isolates. J Clin Microbiol 57:e00634-19. https://doi.org/10.1128/JCM.00634-19. 15. Thirumal Kumar D, Lavanya P, George Priya Doss C, Tayubi IA, Naveen Kumar DR, Francis Yesurajan I, Siva R, Balaji V. 2017. A molecular docking and dynamics approach to screen potent inhibitors against fosfomycin resistant enzyme in clinical Klebsiella pneumoniae. J Cell Biochem 118:4088–4094. https://doi.org/10.1002/jcb.26064. 16. Tomich AD, Klontz EH, Deredge D, Barnard JP, McElheny CL, Eshbach ML, Weisz OA, Wintrode P, Doi Y, Sundberg EJ, Sluis-Cremer N. 2019. Smallmolecule inhibitor of FosA expands fosfomycin activity to multidrug-re- sistant Gram-negative pathogens. Antimicrob Agents Chemother 63: e01524-18. https://doi.org/10.1128/AAC.01524-18. 17. Portillo-Calderón I, Ortiz-Padilla M, Rodríguez-Martínez JM, de Gregorio- Iaria B, Blázquez J, Rodríguez-Baño J, Pascual A, Docobo-Pérez F. 2020. Contribution of hypermutation to fosfomycin heteroresistance in Escherichia coli. J Antimicrob Chemother 75:2066–2075. https://doi.org/10 .1093/jac/dkaa131. 18. European Committee on Antimicrobial Susceptibility Testing. 2020. Antimicrobial wild type distributions of microorganisms. https://mic.eucast .org/Eucast2/. 19. The European Committee on Antimicrobial Susceptibility Testing. 2020. Breakpoint tables for interpretation of MICs and zone diameters. Version 10.0. http://www.eucast.org. 20. Spanish Agency of Medicines and Medical Devices (Agencia española del medicamento y productos sanitarios, AEMPS). 2020. Ficha técnica: foscavir 24 mg/ml solución para perfusión. Agencia Española de Medicamentos y Productos Sanitarios (AEMPS), Madrid, Spain. https://cima.aemps.es/ cima/pdfs/es/ft/59712/FT_59712.pdf. 21. Kim J-S, Cho D-H, Park M, Chung W-J, Shin D, Ko KS, Kweon D-H. 2016. CRISPR/Cas9-mediated re-sensitization of antibiotic-resistant Escherichia coli harboring extended-spectrum b -lactamases. J Microbiol Biotechnol 26:394–401. https://doi.org/10.4014/jmb.1508.08080. 22. Aziz RK, Bartels D, Best AA, DeJongh M, Disz T, Edwards RA, Formsma K, Gerdes S, Glass EM, Kubal M, Meyer F, Olsen GJ, Olson R, Osterman AL, Overbeek RA, McNeil LK, Paarmann D, Paczian T, Parrello B, Pusch GD, Reich C, Stevens R, Vassieva O, Vonstein V, Wilke A, Zagnitko O. 2008. The RAST Server: Rapid Annotations using Subsystems Technology. BMC Genomics 9:75. https://doi.org/10.1186/1471-2164-9-75. 23. Huang TW, Lam I, Chang HY, Tsai SF, Palsson BO, Charusanti P. 2014. Capsule deletion via a l -Red knockout system perturbs biofilm formation and fimbriae expression in Klebsiella pneumoniae MGH 78578. BMC Res Notes 7:13. https://doi.org/10.1186/1756-0500-7-13. 24. Yadav B, Wennerberg K, Aittokallio T, Tang J. 2015. Searching for drug synergy in complex dose-response landscapes using an interaction potency model. Comput Struct Biotechnol J 13:504–513. https://doi.org/10 .1016/j.csbj.2015.09.001. 25. Ianevski A, Giri AK, Aittokallio T. 2020. SynergyFinder 2.0: visual analytics of multi-drug combination synergies. Nucleic Acids Res 48:W488–W493. https://doi.org/10.1093/nar/gkaa216. 26. Pfausler B, Spiss H, Dittrich P, Zeitlinger M, Schmutzhard E, Joukhadar C. 2004. Concentrations of fosfomycin in the cerebrospinal fluid of neurointensive care patients with ventriculostomy-associated ventriculitis. J Antimicrob Chemother 53:848–852. https://doi.org/10.1093/jac/dkh158. Ortiz-Padilla et al. Antimicrobial Agents and Chemotherapy March 2021 Volume 65 Issue 3 e01911-20 aac.asm.org 10 on March 5, 2021 at USE/BCTA.GEN UNIVERSITARIAhttp://aac.asm.org/Downloaded from 56 Supplementary Material of Interplay among Different Fosfomycin Resistance Mechanisms in Klebsiella pneumoniae. Table S1. Primers and plasmids used to construct and check the mutant isogenic collection. 57 Table S2. Median (IQR) frequencies of spontaneous fosfomycin (64 and 512 mg/L) and rifampicin resistant mutants (100 mg/L), with and without the addition of 0.623mM of PPF. Figure S1. Frequencies of spontaneous rifampicin resistant mutants of EcMG1655, KpATCC, Kp12 and Kp142 wild-type strains after incubation, with and without PPF. Black lines represent the median values of the three replicates. 58 Figure S2. Viability of clinical isolates against fosfomycin concentrations from 0-256 mg/L, with and without 0.623 mM PPF after 24h. Lines stand for mean values of measured viability. Symbols represent single results of the replicates. 59 Figure S3. Time-kill results for KpATCC 700721, wild-type (WT), and isogenic mutant strains using 64 and 307 mg/L fosfomycin, alone and in combination with 0.623 mM PPF. Dashed lines represent bacterial concentrations in fosfomycin-containing plates (64 mg/L). 60 Figure S4. Time-kill results for Kp12, wild-type (WT), and isogenic mutant strains using 64 and 307 mg/L fosfomycin, alone and in combination with 0.623 mM PPF. Dashed lines represent bacterial concentrations in fosfomycin-containing plates (64 mg/L) 61 Figure S5. Time-kill results for Kp142, wild-type (WT), and isogenic mutant strains using 64 and 307 mg/L fosfomycin, alone and in combination with 0.623 mM PPF. Dashed lines represent bacterial concentrations in fosfomycin-containing plates (64 mg/L) Antibiotics 2022,11, 1612 4 of 12 2.2. Susceptibility Testing The fosfomycin MICs, performed with the reference method, for the isogenic collection are shown in Table 1, and the results of disk diffusion are shown in Figure 3. Table 1. Fosfomycin MIC (mg/L) results by agar dilution without and with 7 mg/mL of glycerol and 25 mg/L of glucose-6-phosphate (G6P). Fosfomycin MIC (mg/L) No Glycerol Glycerol 7 mg/L No G6P G6P 25 mg/L No G6P G6P 25 mg/L E. coli ATCC25922 >64 2 64 2 E. coli MG1655 Wild-type 64 2 32 2 ∆glpT >64 2 >64 2 ∆uhpT >64 >64 >64 >64 ∆cyaA >64 4 >64 2 ∆glpR 4121 ∆glpK >64 4 32 4 ∆glpT-uhpT >64 >64 >64 >64 Antibiotics 2022, 11, x FOR PEER REVIEW 5 of 13 Figure 3. Results of fosfomycin susceptibility by disk diffusion assay with the addition of 0.45, 1.8, and 7 mg/L of glycerol and/or 25 mg/L of glucose-6-phosphate (G6P). The dark-colored columns show the diameter of the main inhibition zone (mean, mm), and the light-colored columns show the diameter of the reduced inhibition zone (mean, mm). Error bars represent the standard deviation. * denotes p < 0.05 with respect to their respective control. Regarding the results observed by the agar dilution method, the addition of glycerol did not improve fosfomycin activity (MIC drop of ≥2 log2 dilution) with respect to fosfomycin alone or in combination with G6P. Regarding the disk diffusion assays, considering the main inhibition zone, the results showed that the wild-type strains significantly increased their fosfomycin susceptibility with the addition of glycerol. This increase occurred with and without the addition of G6P for both wild-type strains, except for the E. coli ATCC25922 strain with glycerol (0.45 mg/mL) with G6P. The use of fosfomycin without G6P generated a reduced inhibition zone for the E. coli ATCC25922 but not for the E. coli MG1655 strains. It should be noted that the addition of glycerol also generated this reduced inhibition zone under certain conditions. The addition of glycerol did not modify the size of the reduced inhibition zone for the E. coli ATCC25922 strain but produced a reduced inhibition zone with the addition of G6P, also observed for the E. coli MG1655. With respect to the ∆glpT mutant, a significant increase in susceptibility was observed with the addition of 1.8 and 7 mg/mL of glycerol. This increase did not occur with the addition of G6P. Regarding the reduced inhibition zone, no changes were observed with the addition of the different concentrations of glycerol, with or without G6P. Figure 3. Results of fosfomycin susceptibility by disk diffusion assay with the addition of 0.45, 1.8, and 7 mg/L of glycerol and/or 25 mg/L of glucose-6-phosphate (G6P). The dark-colored columns show the diameter of the main inhibition zone (mean, mm), and the light-colored columns show the diameter of the reduced inhibition zone (mean, mm). Error bars represent the standard deviation. * denotes p< 0.05 with respect to their respective control. Antibiotics 2022,11, 1612 5 of 12 With the agar dilution assay, the fosfomycin MIC for E. coli ATCC25922 was 2 mg/L with the addition of G6P. The absence of G6P increased the MIC to >64 mg/L, and the addition of 7 mg/mL glycerol did not restore fosfomycin susceptibility. With respect to the E. coli MG1655 strain, similar results were observed with a MIC of 2 mg/L with G6P, 64 mg/L without G6P, and 32 mg/L without G6P but with the addition of glycerol. The fosfomycin MIC for the single uhpT and the double-gene mutant for both transporters ( ∆ glpT-uhpT) were >64 mg/L for all the assayed conditions. For the remaining mutant strains, the MIC ranged from 1–4 mg/L with the addition of G6P, ≥ 32 mg/L in the absence of G6P, regardless of the addition of glycerol, except the mutant for ∆ glpR gene (2 and 4 mg/L). Regarding the results observed by the agar dilution method, the addition of glycerol did not improve fosfomycin activity (MIC drop of ≥ 2 log 2 dilution) with respect to fosfomycin alone or in combination with G6P. Regarding the disk diffusion assays, considering the main inhibition zone, the results showed that the wild-type strains significantly increased their fosfomycin susceptibility with the addition of glycerol. This increase occurred with and without the addition of G6P for both wild-type strains, except for the E. coli ATCC25922 strain with glycerol (0.45 mg/mL) with G6P. The use of fosfomycin without G6P generated a reduced inhibition zone for the E. coli ATCC25922 but not for the E. coli MG1655 strains. It should be noted that the addition of glycerol also generated this reduced inhibition zone under certain conditions. The addition of glycerol did not modify the size of the reduced inhibition zone for the E. coli ATCC25922 strain but produced a reduced inhibition zone with the addition of G6P, also observed for the E. coli MG1655. With respect to the ∆ glpT mutant, a significant increase in susceptibility was observed with the addition of 1.8 and 7 mg/mL of glycerol. This increase did not occur with the addition of G6P. Regarding the reduced inhibition zone, no changes were observed with the addition of the different concentrations of glycerol, with or without G6P. The E. coli MG1655 ∆ uhpT mutant showed, in the reduced inhibition zone, a significant increase in susceptibility with 1.8 and 7 mg/mL of glycerol, with and without G6P. Furthermore, the addition of any concentration of glycerol significantly increased the main inhibition zone, with or without G6P. E. coli MG1655 ∆ cyaA strain showed a significant increase in susceptibility with 0.45 and 1.8 mg/mL of glycerol. The addition of 7 mg/mL of glycerol did not produce any effect, regardless of the addition of G6P. E. coli MG1655 ∆ glpT-uhpT, ∆ glpR, and ∆ glpK strains did not show any significant increase in susceptibility in any of the conditions evaluated. 2.3. Fosfomycin and Glycerol Interaction Assay The result of the interaction assay is shown in Table 2and Figure S2. The combination of fosfomycin and glycerol showed synergistic activity (ZIP score > +10) G6P for the wildtype strains (E. coli ATCC25922: 16.3 ± 1.2 and E. coli MG1655: 16.5 ± 1.9), and for E. coli MG1655 ∆ uhpT strain with G6P (12.3 ± 3.6) and without (11.3 ± 2.9). The most synergistic area occurred between 7–28 mg/mL of glycerol and 0.5–2 mg/L of fosfomycin for the E. coli ATCC25922 strain. Similar results were observed for the E. coli MG1655 strain, but at 0.25–1 mg/L of fosfomycin. In the case of E. coli MG1655 ∆ uhpT, the greatest synergy was observed at the same concentration as that observed for the E. coli ATCC25922 strain with and without G6P. Antibiotics 2022,11, 1612 6 of 12 Table 2. ZIP synergy scores for the combination of fosfomycin with glycerol, with and without glucose- 6-phosphate (G6P). Green and red colors show synergistic and antagonistic results, respectively. ZIP SYNERGY SCORE ATCC25922 MG1655 Wild-Type ∆glpT ∆uhpT ∆glpT-uhpT ∆cyaA ∆glpR ∆glpK Fosfomycin + Glycerol 16.4 ±1.2 16.5 ±1.9 2.5 ±4.8 11.3 ±2.9 −12.8 ±5.4 −0.7 ±6.4 2.7 ±1.6 1.7 ±1.9 Fosfomycin + Glycerol + G6P 6.8 ±1.9 5.8 ±2.6 6.9 ±4.3 12.4 ±3.6 −21 ±3.8 1.8 ±6.6 6.1 ±1.6 2.7 ±3.7 Antagonism (ZIP score < − 10) was observed against the E. coli MG1655 ∆ glpT-uhpT strain under any condition ( − 12.8 ± 5.4 and − 21 ± 3.8). The addition of glycerol did not interact with the rest of the evaluated strains. 2.4. Fosfomycin and Glycerol Time-Kill Assays E. coli E. coli ATCC25922, E. coli MG1655, and their mutant derivatives were evaluated at two concentrations of fosfomycin (64 and 307 mg/L), without glycerol, and with two concentrations (0.45 and 7 mg/mL). The results are shown in Figure 4and Figure S3 (for E. coli ∆glpT-uhpT,∆cyaA,∆glpR, and ∆glpK strains). Antibiotics 2022, 11, x FOR PEER REVIEW 7 of 13 Figure 4. Time-kill assays of fosfomycin alone and in combination with glycerol (0.45 and 7 mg/L) and/or glucose-6-phosphate (G6P, 25 mg/L) against E. coli wild-type E. coli ATCC25922 and E. coli MG1655 wild-type and mutant derivative strains (∆glpT and ∆uhpT), at 0, 2, 4, 8, and 24 h. Bacterial concentrations (CFU/mL) are represented as symbols for mean and range. The addition of glycerol and G6P did not modify growth under control conditions (without fosfomycin, data not shown). Fosfomycin alone at 64 mg/L showed a bactericidal effect (decreased the bacterial burden ≥3 log10 CFU/mL) within the first 4 h against the wild-type and the E. coli MG1655 ∆uhpT mutant strains, but bacterial regrowths were observed at 8 and 24 h. The addition of G6P also showed a bactericidal effect within the first 4–8 h in all strains except E. coli MG1655 ∆uhpT and ∆glpT-uhpT and prevented the bacterial regrowth for the wild-type strains and the E. coli MG1655 ∆glpR and ∆glpK mutant strains. The addition of glycerol at both tested concentrations did not improve the bactericidal effect against any of the evaluated strains. Except for E. coli MG1655 ∆glpT and ∆glpT-uhpT strains, fosfomycin alone at 307 mg/L showed a bactericidal effect against all the strains within the first hours. However, bacterial regrowth was observed except for the wild-type E. coli ATCC and MG1655 ∆glpR strains. The addition of G6P improved the bactericidal effect and prevented the regrowth in all the strains except for E. coli MG1655 ∆uhpT and ∆glpT-uhpT. The addition of glycerol at both concentrations did not improve the bactericidal effect in any of the evaluated strains. Only against E. coli MG1655 ∆uhpT strain fosfomycin combined with 0.45 mg/mL of glycerol showed an initial improved activity; however, bacterial regrowth was observed after 24 h. 3. Discussion The present study evaluates the role of glycerol in enhancing fosfomycin activity against wild-type strains of E. coli and strains that harbor specific determinants of resistance to fosfomycin. Figure 4. Time-kill assays of fosfomycin alone and in combination with glycerol (0.45 and 7 mg/L) and/or glucose-6-phosphate (G6P, 25 mg/L) against E. coli wild-type E. coli ATCC25922 and E. coli MG1655 wild-type and mutant derivative strains ( ∆ glpT and ∆ uhpT), at 0, 2, 4, 8, and 24 h. Bacterial concentrations (CFU/mL) are represented as symbols for mean and range. The addition of glycerol and G6P did not modify growth under control conditions (without fosfomycin, data not shown). Fosfomycin alone at 64 mg/L showed a bactericidal Antibiotics 2022,11, 1612 7 of 12 effect (decreased the bacterial burden ≥ 3 log 10 CFU/mL) within the first 4 h against the wild-type and the E. coli MG1655 ∆ uhpT mutant strains, but bacterial regrowths were observed at 8 and 24 h. The addition of G6P also showed a bactericidal effect within the first 4–8 h in all strains except E. coli MG1655 ∆ uhpT and ∆ glpT-uhpT and prevented the bacterial regrowth for the wild-type strains and the E. coli MG1655 ∆ glpR and ∆ glpK mutant strains. The addition of glycerol at both tested concentrations did not improve the bactericidal effect against any of the evaluated strains. Except for E. coli MG1655 ∆ glpT and ∆ glpT-uhpT strains,fosfomycin alone at 307 mg/L showed a bactericidal effect against all the strains within the first hours. However, bacterial regrowth was observed except for the wild-type E. coli ATCC and MG1655 ∆ glpR strains. The addition of G6P improved the bactericidal effect and prevented the regrowth in all the strains except for E. coli MG1655 ∆ uhpT and ∆ glpT-uhpT. The addition of glycerol at both concentrations did not improve the bactericidal effect in any of the evaluated strains. Only against E. coli MG1655 ∆ uhpT strain fosfomycin combined with 0.45 mg/mL of glycerol showed an initial improved activity; however, bacterial regrowth was observed after 24 h. 3. Discussion The present study evaluates the role of glycerol in enhancing fosfomycin activity against wild-type strains of E. coli and strains that harbor specific determinants of resistance to fosfomycin. In general, the activation of the GlpT transporter using glycerol as an internal activator showed contradictory results on the activity of fosfomycin. The present study shows that glycerol, at physiological concentrations, can activate the glpT promoter, increasing glpT expression [ 13 , 16 ]. This result agrees with previous works showing the activation of genes (glpACB,glpD glpFKX, glpTQ, etc.) involved in the glycerol catabolism and regulated by the glpR repressor in E. coli,P. aeruginosa or P. putida [ 14 , 15 , 17 ]. However, this glycerol induction of glpT shows a delayed initiation in contrast to the rapid activation of the hexoses-6-phosphate transporter (uhpT) with the addition of G6P. These effects have been observed using GFP promoter fusions in real-time fluorescence monitoring and flow cytometry assays for glpT and uhpT genes. A possible explanation for this behavior was given by Nikel et al. [ 14 ], who observed a protracted lag phase in cultures of P. putida KT2440 growing in glycerol. The regulatory network of the glp genes needs the product of the first biochemical reaction (sn-glycerol-3-P) to derepress gene expression, which is otherwise inhibited by GlpR. However, the genes that encode the glycerol transporter and the kinase that produces G3P from glycerol are repressed by the GlpR protein. Thus, to get the transcription started is the low-probability effector-independent stochastic lifting of the repression. While the derepression process is taking place, this transcriptional architecture translates into different levels of metabolic activity (representing, in this context, the ability of the cells to catabolize glycerol). It is important to note that while the glpT gene is absent in P. putida, the regulation network is conserved in E. coli and P. aeruginosa. [1,15,18] With regard to fosfomycin activity with the addition of glycerol, the present study shows distinct results depending on the characteristics of the assay. In the disk diffusion assays, secondary and reduced inhibition zones were observed, suggesting the presence of subpopulations with different fosfomycin susceptibility, even with the use of glycerol and G6P. The maximum fosfomycin activity was observed against the glpR repressor mutant because the glpT transporter is fully derepressed and insensitive to the addition of glycerol. Additionally, the addition of glycerol did not show any effect in glpK mutant due to the inability to transform glycerol into glycerol-3-phosphate [16]. However, fosfomycin activity was observed, similar to that observed against the E. coli MG1655 wild-type strain without glycerol, and this could be partially explained by the intracellular biosynthesis of sn-glycerol-3-phosphate as a precursor of phospholipid synthesis [ 19 ]. The absence of activity was observed in the double-gen mutant ∆ glpT-uhpT Antibiotics 2022,11, 1612 8 of 12 strain, indicating that the activation of the fosfomycin transporters is the main control factor of fosfomycin activity, as previously observed by Ballesteros et al. [1]. It is also important to note that the reduced inhibition zones remained unchanged, irrespective of the glycerol concentration or the addition of G6P, suggesting the presence of baseline defects in the complex regulatory networks of the fosfomycin transporters glpT and uhpT [6,20]. In the present study, discrepancies between disk diffusion and agar dilution susceptibility assays have been observed. These results agree with our previous studies in which fosfomycin susceptibility showed inconsistent results between broth and agar dilutions and agar diffusion techniques in collections of clinical isolates and isogenic mutants related to fosfomycin resistance, including heteroresistant strains [ 1 , 21 , 22 ]. Although there were no differences in the MIC of fosfomycin when glycerol was added to the assay, it must be noted that lower bacterial densities were observed (data not shown), indicating a mild synergistic effect not measurable with the assay. Regarding the interaction assays between fosfomycin and glycerol, only the wild-type strains and E. coli MG1655 ∆ uhpT mutant showed synergistic activity. However, the time-kill assays did not show sensitization with the combination of fosfomycin and glycerol. It is important to note that in the time-kill assays, divergent results were observed between the replicates for wild-type strains and for the E. coli MG1655 ∆ uhpT. For these strains, one replicate of the time-kill assay showed a total bacterial clearance and the other a bacterial regrowth after 24 h, which partially agrees with the results observed in the interaction assay. This divergence found between these replicates could be explained considering the rapid bactericidal activity of fosfomycin in contrast to our previous results showing the delayed activation of the glpT transporter, increasing the probability of the emergence of fosfomycin-resistant subpopulations as commonly observed in previous studies [1,23]. In conclusion, glycerol showed increased expression of the fosfomycin transporter glpT and a synergistic effect with fosfomycin in the interaction and disk diffusion assays. However, this molecule does not appear to be a good candidate as an adjuvant to fosfomycin therapy since the delay in the induction of glpT allows the selection of resistant subpopulations. It would therefore be necessary to perform further in vitro and in vivo studies aimed at overcoming these problems. 4. Materials and Methods 4.1. Bacterial Strains E. coli ATCC25922 and E. coli MG1655 were used as the reference strain, and six isogenic mutants from E. coli MG1655 mutants ( ∆ glpT, ∆ uhpT, ∆ glpT-uhpT, ∆ cyaA, ∆ glpR, and ∆ glpK) were used in the assays. Isogenic mutants were generated from the KEIO collection [ 24 ] using phage P1vir transduction (Coli Genetic Stock Center [CGSC], Yale University, New Haven, CT, USA) as previously described [25]. 4.2. Activation Kinetic of Fosfomycin Transporters Promoters The activity of the fosfomycin transporter promoters glpT and uhpT was evaluated by monitoring the fluorescence accumulation in E. coli ATCC25922 and E. coli MG1655 carrying pMS201-PglpT::gfpmut2 or pMS201-PuhpT::gfpmut2 reporters, as described by Zaslaver et al. [ 26 ]. Bacterial culture fluorescence was determined after transferring overnight cultures in MHB to 96-well plates (Nunclon Delta Surface, Thermo Scientific, Waltham, MA, USA) with 200 µ L of MHB per well. Starting bacterial concentrations were adjusted to 5×105CFU/mL. The assays were carried out in Mueller Hinton Broth II (MHB) alone or supplemented with 0.45 mg/mL, 1.8 mg/mL, and 7 mg/mL of glycerol with and without G6P (Sigma-Aldrich, Madrid, Spain). Low, medium, and high glycerol physiological concentrations were assayed in combination with fosfomycin. The low glycerol concentration of 0.45 mg/mL corresponded to a target effective concentration to reduce the intracranial pressure [ 12 ]. The high glycerol concentration of 7 mg/mL corresponded to steady-state serum concentrations in patients with normal hepatic and renal function following constant Antibiotics 2022,11, 1612 9 of 12 intravenous infusion of 0.87 g/kg/h [ 27 ]. Finally, the medium glycerol concentration of 1.8 mg/mL was selected, corresponding to an intermediate concentration between 0.45 and 7 mg/mL in a log 2 dilution scale. Green fluorescence (excited at 485 nm and measured at 540 nm) and bacterial growth (measured at 595 nm) were monitored each hour for 24 h with an Infinite200 Pro plate reader (Tecan Group AG, Männedorf, Switzerland). The assays were performed in duplicate. The OD:fluorescence ratio of the promoterless construction (pMS201-Ø::gfpmut2) was used as background for all experiments under the different growth conditions. Fluorescence was normalized to the OD, and the background was subtracted. The results were compared with ANOVA and Tukey’s multiple comparison test. 4.3. Population Analysis of Fosfomycin Transporters Promoters Induction Flow cytometry assays were conducted to assess the population distribution of glpT and uhpT expressions (pMS201-PglpT::gfpmut2 or pMS201-PuhpT::gfpmut2 reporters) in E. coli ATCC25922 and E. coli MG1655. Furthermore, a negative expression control (pMS201- Ø::gfpmut2) and positive expression control (pMS201-PcyaA::gfpmut2) were used as described. Overnight cultures were diluted in 20 mL adjusting a bacterial concentration of 5 × 10 5 CFU/mL. Bacterial growths were carried out in MHB, MHB supplemented with 25 mg/L of G6P, and MHB supplemented with glycerol: 0.45 mg/mL, 1.8 mg/mL, and 7 mg/mL. A milliliter sample of each condition was taken, and fluorescence was monitored at 4 and 24 h. The Beckman Coulter FC500 cytometer (Beckman Coulter, United States) was used for size (FSC) and complexity (SSC) measurements to define the bacterial population. Green fluorescence was excited using a blue laser (488 nm) and measured at 530/30 nm. The promoterless construction (pMS201-Ø::gfpmut2) was used as a negative control, and pMS201-PcyaA::gfpmut2 was used as a positive control. 4.4. Susceptibility Testing Fosfomycin MIC was performed using the reference method, the agar dilution assay, following EUCAST standards [ 28 ]. Mueller Hinton II agar (MHA) plates (Sigma-Aldrich, Madrid, Spain), with and without 25 mg/L of G6P, with and without 7 mg/mL of glycerol. The fosfomycin (Sigma-Aldrich) concentration ranged from 0.5 to 64 mg/L. Plates were dried and incubated for 20 h at 35 ◦C. The assays were performed in duplicate. Fosfomycin susceptibility was also determined using the disk diffusion method using blank antimicrobial disks loaded with 200 µ g of fosfomycin alone or with 50 µ g of G6P, following EUCAST recommendations [ 28 ]. Additionally, fosfomycin activity was assayed by supplementing MHA plates with the addition of 0.45 mg/mL, 1.8 mg/mL, and 7 mg/mL of glycerol. The diameter of the main and any other reduced inhibition zone was measured. The addition of glycerol was compared with respect to the negative control (MHA with or without G6P). The results were compared with ANOVA and Tukey’s multiple comparison test. The assays were performed in triplicate. 4.5. Fosfomycin and Glycerol Interaction Assay The interaction between fosfomycin and glycerol was studied using the checkerboard assay in duplicate. Briefly, the interaction assay was performed with an inoculum of 5 × 10 5 CFU/mL in 96-well plates with a final volume of 200 µ L per well. Assays were performed with MHB with and without 25 mg/L G6P. Fosfomycin concentrations ranged from 0.125 to 128 mg/L, and glycerol concentrations from 0.45 to28.8 mg/mL. Wells without fosfomycin or glycerol were used as single-drug assays or growth controls. Bacterial densities were quantified spectrophotometrically by measuring optical density at 595 nm using an Infinite200 Pro plate reader. Bacterial viability was calculated as the ratio of the final bacterial OD to the final bacterial OD of the control growth well (without drug). A four-parameter log-logistic model was fitted to the data to generate dose–response curves for fosfomycin and glycerol. The degree of drug synergy across the entire dose– response matrix was analyzed using the response surface model, zero interaction potential Antibiotics 2022,11, 1612 10 of 12 (ZIP) [ 29 ]. The ZIP model assumes that two noninteracting drugs are expected to incur minimal changes in their dose–response curves. A delta score was calculated to quantify the deviation from the expectation of ZIP for a given dose pair and used the average delta over a dose–response matrix as a summary interaction score for a drug combination. Model construction and synergy studies were performed with the Synergyfinder package for R [ 29 ]. A synergy score of < − 10 was considered antagonistic, a range from − 10 to 10 as additive, and >10 as synergistic. 4.6. Fosfomycin and Glycerol Time-Kill Assays Time-kill assays were performed in duplicate using fosfomycin concentrations of 0 (as growth control), 64, and 307 mg/L with and without 25 mg/L G6P, with and without 0.45 and 7 mg/mL of glycerol. The activity of fosfomycin alone at concentrations of 64 mg/L (lowest concentrations of fosfomycin in the non-susceptible category according to EUCAST breakpoints) and at 307 mg/L (mean maximum plasma concentrations in humans at steady-state after a dose of fosfomycin 8 g/Q8h), respectively, was determined [28,30]. Briefly, isolated colonies overnight of each strain were used to prepare the preinoculum in MHB and incubated overnight with shaking at 37 ◦ C. The starting inoculum was set at 5 × 10 5 CFU/mL in a final volume of 20 mL, and bacterial cultures were incubated at 37 ◦ C with shaking. The number of viable CFUs was determined at 0, 2, 4, 8, and 24 h by serial dilution, followed by plating on MH agar plates with or without 64 mg/L fosfomycin and 25 mg/L G6P. The number of colonies was counted after 24 h of incubation. Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/antibiotics11111612/s1, Figure S1: Assay of promoter activities in response to glycerol (GLY) and glucose-6-phosphate (G6P). Time-course quantification of GFP expression as a measure of induction of the glpT and uhpT genes in the ATCC25922 and MG1655 strains after 4, 12, and 24 h of exposure. The data were normalized to promoterless strains. Error bars represent standard deviations. Concentrations of glycerol (Gly) of 0.45, 1.8, and 7 mg/mL and glucose-6-phosphate (G6P) of 25 mg/L were used as inductors; Figure S2: Interaction assay of fosfomycin in combination with glycerol against Escherichia coli ATCC25922 and MG1655 strains, represented as heat maps. The red and green areas represent synergy and antagonism, respectively. The white rectangles show the maximum synergistic area. The concentration–response curves for fosfomycin and glycerol alone are found on the left side of each heatmap. Figure S3: Time-kill assays of fosfomycin alone and in combination with glycerol (0.45 and 7 mg/L) and/or glucose-6-phosphate (G6P, 25 mg/L) against E. coli wild-type E. coli MG1655 mutant derivative strains ( ∆ glpT-uhpT, ∆ cyaA, ∆ glpR, and ∆ glpK), at 0, 2, 4, 8, and 24 h. Bacterial concentrations (CFU/mL) are represented as symbols for mean and range. Author Contributions: Conceptualization: M.O.-P., I.P.-C. and F.D.-P.; methodology: M.O.-P., J.M.R.-M. and F.D.-P.; validation: M.O.-P., I.P.-C. and F.D.-P.; formal analysis: M.O.-P. and F.D.-P.; investigation: M.O.-P. and A.V.-E.; resources: Á.P., J.M.R.-M. and F.D.-P.; writing-original draft preparation: M.O.-P. and F.D.-P.; writing—review and editing: M.O.-P., I.P.-C., J.R.-B. and F.D.-P.; visualization: M.O.-P. and F.D.-P.; supervision: Á.P., J.M.R.-M. and F.D.-P.; project administration: J.M.R.-M. and F.D.-P.; funding acquisition: J.R.-B., Á.P. and F.D.-P. All authors have read and agreed to the published version of the manuscript. Funding: This study has been funded by Instituto de Salud Carlos III through the projects PI19/01645 and REIPI RD16/0016/0001 (co-funded by European Regional Development Fund “Away to Make Europe”). Miriam Ortiz-Padilla is supported by a PFIS grant from the Instituto de Salud Carlos III (FI17/00235). Inés Portillo-Calderón is supported by a Rio Hortega grant from the Instituto de Salud Carlos III (CM20/0092). (Co-funded by European Social Fund “Investing in your future”). The funders had no role in the design, collection of data, analysis, and writing of the manuscript or the decision to publish. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Antibiotics 2022,11, 1612 11 of 12 Data Availability Statement: The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. Acknowledgments: This study was presented in part at ECCMID 2022, Lisbon, Portugal (Poster Presentation P0794). Conflicts of Interest: The authors declare no conflict of interest. References 1. Ballestero-Téllez, M.; Docobo-Pérez, F.; Portillo-Calderón, I.; Rodríguez-Martínez, J.M.; Racero, L.; Ramos-Guelfo, M.S.; Blázquez, J.; Rodríguez-Baño, J.; Pascual, A. Molecular insights into fosfomycin resistance in Escherichia coli.J. Antimicrob. Chemother. 2017 , 72, 1303–1309. [CrossRef] 2. Prestinaci, F.; Pezzotti, P.; Pantosti, A. Antimicrobial resistance: A global multifaceted phenomenon. Pathog Glob Health 2015 ,109, 309–318. [CrossRef] [PubMed] 3. Cassini, A.; Högberg, L.D.; Plachouras, D.; Quattrocchi, A.; Hoxha, A.; Simonsen, G.S.; Colomb-Cotinat, M.; Kretzschmar, M.E.; Devleesschauwer, B.; Cecchini, M.; et al. Attributable deaths and disability-adjusted life-years caused by infections with antibiotic-resistant bacteria in the EU and the European Economic Area in 2015: A population-level modelling analysis. Lancet Infect. Dis. 2019,19, 56–66. [CrossRef] 4. Docobo-Pérez, F.; Drusano, G.L.; Johnson, A.; Goodwin, J.; Whalley, S.; Ramos-Martín, V.; Ballestero-Tellez, M.; Rodriguez- Martinez, J.M.; Conejo, M.C.; van Guilder, M.; et al. Pharmacodynamics of fosfomycin: Insights into clinical use for antimicrobial resistance. Antimicrob. Agents Chemother. 2015,59, 5602–5610. [CrossRef] [PubMed] 5. Falagas, M.E.; Maraki, S.; Karageorgopoulos, D.E.; Kastoris, A.C.; Mavromanolakis, E.; Samonis, G. Antimicrobial susceptibility of multidrug-resistant (MDR) and extensively drug-resistant (XDR) Enterobacteriaceae isolates to fosfomycin. Int. J. Antimicrob. Agents 2010,35, 240–243. [CrossRef] 6. Marger, M.D.; Saier, M.H.J. A major superfamily of transmembrane facilitators that catalyse uniport, symport and antiport. Trends Biochem. Sci. 1993,18, 13–20. [CrossRef] 7. Franchini, A.G.; Ihssen, J.; Egli, T. Effect of Global Regulators RpoS and Cyclic-AMP/CRP on the Catabolome and Transcriptome of Escherichia coli K12 during Carbon- and Energy-Limited Growth. PLoS ONE 2015,10, e0133793. [CrossRef] 8. Sakamoto, Y.; Furukawa, S.; Ogihara, H.; Yamasaki, M. Fosmidomycin resistance in adenylate cyclase deficient (cya) mutants of Escherichia coli.Biosci. Biotechnol. Biochem. 2003,67, 2030–2033. [CrossRef] 9. Kadner, R.J.; Shattuck-Eidens, D.M. Genetic control of the hexose phosphate transport system of Escherichia coli: Mapping of deletion and insertion mutations in the uhp region. J. Bacteriol. 1983,155, 1052–1061. [CrossRef] [PubMed] 10. Yang, B.; Gerhardt, S.G.; Larson, T.J. Action at a distance for glp repressor control of glpTQ transcription in Escherichia coli K-12. Mol. Microbiol. 1997,24, 511–521. [CrossRef] 11. Santoro, A.; Cappello, A.R.; Madeo, M.; Martello, E.; Iacopetta, D.; Dolce, V. Interaction of fosfomycin with the Glycerol 3-phosphate Transporter of Escherichia coli.Biochim. Biophys. Acta (BBA)-Gen. Subj. 2011,1810, 1323–1329. [CrossRef] [PubMed] 12. Frank, M.S.; Nahata, M.C.; Hilty, M.D. Glycerol: A review of its pharmacology, pharmacokinetics, adverse reactions, and clinical use. Pharmacotherapy 1981,1, 147–160. [CrossRef] [PubMed] 13. Lin, E.C. Glycerol dissimilation and its regulation in bacteria. Annu. Rev. Microbiol. 1976,30, 535–578. [CrossRef] 14. Nikel, P.I.; Romero-Campero, F.J.; Zeidman, J.A.; Goñi-Moreno, Á.; de Lorenzo, V. The glycerol-dependent metabolic persistence of Pseudomonas putida KT2440 reflects the regulatory logic of the GlpR repressor. MBio 2015 ,6, e00340-15. [CrossRef] [PubMed] 15. Koch, J.P.; Hayashi, S.; Lin, E.C. The Control of Dissimilation of Glycerol and L- α -Glycerophosphate in Escherichia coli.J. Biol. Chem. 1964,239, 3106–3108. [CrossRef] 16. Martínez-Gómez, K.; Flores, N.; Castañeda, H.M.; Martínez-Batallar, G.; Hernández-Chávez, G.; Ramírez, O.T.; Gosset, G.; Encarnación, S.; Bolivar, F. New insights into Escherichia coli metabolism: Carbon scavenging, acetate metabolism and carbon recycling responses during growth on glycerol. Microb. Cell Factories 2012,11, 46. [CrossRef] 17. Nikel, P.I.; Kim, J.; de Lorenzo, V. Metabolic and regulatory rearrangements underlying glycerol metabolism in Pseudomonas putidaKT2440. Environ. Microbiol. 2014,16, 239–254. [CrossRef] 18. Castañeda-García, A.; Rodríguez-Rojas, A.; Guelfo, J.R.; Blázquez, J. The Glycerol-3-Phosphate Permease GlpT Is the Only Fosfomycin Transporter in Pseudomonas aeruginosa. J. Bacteriol. 2009,191, 6968–6974. [CrossRef] 19. Edgar, J.R.; Bell, R.M. Biosynthesis in Escherichia coli of sn-glycerol-3-phosphate, a precursor of phospholipid. Further kinetic characterization of wild type and feedback-resistant forms of the biosynthetic sn-glycerol-3-phosphate dehydrogenase. J. Biol. Chem. 1980,255, 3492–3497. [CrossRef] 20. Castañeda-García, A.; Blázquez, J.; Rodríguez-Rojas, A. Molecular mechanisms and clinical impact of acquired and intrinsic fosfomycin resistance. Antibiotics 2013,2, 217–236. [CrossRef] 21. Ballestero-Téllez, M.; Docobo-Pérez, F.; Rodríguez-Martínez, J.; Conejo, M.; Ramos-Guelfo, M.; Blázquez, J.; Rodríguez-Baño, J.; Pascual, A. Role of inoculum and mutant frequency on fosfomycin MIC discrepancies by agar dilution and broth microdilution methods in Enterobacteriaceae. Clin. Microbiol. Infect. 2017,23, 325–331. [CrossRef] Antibiotics 2022,11, 1612 12 of 12 22. Portillo-Calderón, I.; Ortiz-Padilla, M.; Rodríguez-Martínez, J.M.; De Gregorio-Iaria, B.; Blázquez, J.; Rodríguez-Baño, J.; Pascual, A.; Docobo-Pérez, F. Contribution of hypermutation to fosfomycin heteroresistance in Escherichia coli.J. Antimicrob. Chemother. 2020,75, 2066–2075. [CrossRef] 23. Portillo-Calderón, I.; Ortiz-Padilla, M.; de Gregorio-Iaria, B.; Merino-Bohorquez, V.; Blázquez, J.; Rodríguez-Baño, J.; Rodríguez- Martínez, J.M.; Pascual, A.; Docobo-Pérez, F. Activity of fosfomycin and amikacin against fosfomycin-heteroresistant Escherichia coli strains in a hollow-fiber infection model. Antimicrob. Agents Chemother. 2021,65, e02213-20. [CrossRef] 24. Baba, T.; Ara, T.; Hasegawa, M.; Takai, Y.; Okumura, Y.; Baba, M.; Datsenko, K.A.; Tomita, M.; Wanner, B.L.; Mori, H. Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: The Keio collection. Mol. Syst. Biol. 2006 ,2, 2006.0008. [CrossRef] [PubMed] 25. Thomason, L.C.; Costantino, N.; Court, D.L. E. coli genome manipulation by P1 transduction. Curr. Protoc. Mol. Biol. 2007 ,79, 1.17.1–1.17.8. [CrossRef] 26. Zaslaver, A.; Bren, A.; Ronen, M.; Itzkovitz, S.; Kikoin, I.; Shavit, S.; Liebermeister, W.; Surette, M.G.; Alon, U. A comprehensive library of fluorescent transcriptional reporters for Escherichia coli.Nat. Methods 2006,3, 623–628. [CrossRef] [PubMed] 27. Snyder, J.; Powner, D.; Grenvik, A. Neurologic intensive care. In Anesthesia and Neurosurgery; Cottrell, J., Turndorf, H., Eds.; Mosby: Maryland Heights, MI, USA, 1980; pp. 322–360. 28. The European Committee on Antimicrobial Susceptibility Testing. Breakpoint Tables for Interpretation of MICs and Zone Diameters. Version 12.0. 2022. Available online: http://www.eucast.org (accessed on 21 May 2022). 29. Yadav, B.; Wennerberg, K.; Aittokallio, T.; Tang, J. Searching for Drug Synergy in Complex Dose–Response Landscapes Using an Interaction Potency Model. Comput. Struct. Biotechnol. J. 2015,13, 504–513. [CrossRef] [PubMed] 30. Pfausler, B. Concentrations of fosfomycin in the cerebrospinal fluid of neurointensive care patients with ventriculostomyassociated ventriculitis. J. Antimicrob. Chemother. 2004,53, 848–852. [CrossRef] [PubMed] 77 Supplementary Material of Effect of Glycerol on Fosfomycin Activity against Escherichia coli. Figure S1. Assay of promoter activities in response to glycerol (GLY) and glucose-6-phosphate (G6P). Time-course quantification of GFP expression as a measure of induction of the glpT and uhpT genes in the ATCC25922 and MG1655 strains after 4, 12, and 24 h of exposure. The data were normalized to promoterless strains. Error bars represent standard deviations. Concentrations of glycerol (Gly) of 0.45, 1.8, and 7 mg/mL and glucose-6-phosphate (G6P) of 25 mg/L were used as inductors. (P i ) from the cell [2]. This intracellular P i is maintained within the range 1e10 mM for Escherichia coli [3]byP i importers (PitA, PitB and PstSCAB) and exporters (PitA, PitB, GlpT, UhpT), and controlled by the histidine kinase PhoR and the response regulator PhoB [3]. Human P i levels vary among the different tissues and fluids, being maintained within a narrow range through a complex interplay between intestinal absorption, exchange with intracellular and bone storage pools, and renal tubular reabsorption [4]. Thus, the aim of the present study was to evaluate the in vitro activity of fosfomycin mimicking P i physiological concentrations against a collection of isogenic E. coli mutants and clinical isolates. Material and methods Bacterial strains and plasmids The wild-type E. coli BW25113 strain and isogenic mutants ( D glpT, D uhpT and D phoB) from the KEIO collection and the double mutant D glpT-uhpT, generated by phage P1vir transduction [5], were used. Six urinary clinical E. coli isolates (Ec20190430, Ec20190556, Ec20190736, Ec2019800, Ec20200150 and Ec20200178) with different fosfomycin resistance from the Andalusian reference laboratory for molecular typing of nosocomial pathogens (PIRASOA programme) were included. Bacterial whole genome sequencing was performed as described in the supplementary material. Escherichia coli ATCC 25922 was used as control strain for the susceptibility tests. pUA66-PglpT::gfpmut2 and pUA66-PuhpT::gfpmut2 promoter fusions were constructed as described by Zaslaver et al. [6]. Bacterial growth medium MuellereHinton Broth (MHB) and MuellereHinton Agar (MHA) were used with different concentrations of P i by adjusting standard MuellereHinton medium (P i ¼1 mM, similar to plasma concentration (0.8e1.4 mM)) with Na 2 HPO 4 to 13 and 42 mM P i , corresponding to the minimum and maximum concentrations of P i in urine, adjusted for a 1-L volume (13e42 mmol/24 hr) or 28 mM (average concentration of P i in urine) [4]. Susceptibility testing Fosfomycin MICs were determined by agar dilution [7], in standard MHA (reference) or supplemented with P i (13 and 42 mM P i ). P i concentrations in MuellereHinton were quantified as described in the supplementary material. EUCAST breakpoints were used for intravenous fosfomycin performed at 1 mM of P i (i.e. R64 mg/L) and for oral fosfomycin at 13 and 42 mM of P i (i.e. R16 mg/L) according to the P i concentrations in blood and urine, respectively. Fosfomycin activity Fosfomycin transporter expressions were assayed by monitoring the fluorescence accumulation in E. coli MG1655 carrying pUA66- PglpT::gfpmut2 or pUA66-PuhpT::gfpmut2 reporters as described in supplementary data. Bacterial growth curves were performed in triplicate using 510 5 CFU/mL in 96-well plates with 200 m L of volume of standard MHB or supplemented with P i (13 and 42 mM) with 25 mg/L of glucose-6-phosphate. Fosfomycin concentrations ranged from 1 to 1024 mg/L, and controls without drug were used. Bacterial viability was quantified spectrophotometrically (595 nm) at 24 hr with an Infinite 200 Pro plate reader. Fosfomycin concentration that reduces bacterial viability to 50% (EC 50 ) was estimated by fitting the bacterial viability to the equation: Viabilityð%Þ¼ 100÷1þ fosfomycin concentration EC 50 . The mean fosfomycin EC 50 values were compared using the ANOVA and Tukey and Bonferroni post hoc tests. Timeekill assays were conducted in duplicate, using fosfomycin concentrations of 0 (control), 307 mg/L (plasma C max concentration after 8 g/every 8 hr intravenously) [8], 1053 and 4415 mg/L (lower and upper limit for the mean C max observed in urine after a single dose of 3 g of oral fosfomycin tromethamine) [9], as fully described in the supplementary material. No ethical considerations were necessary for the present study. Results Susceptibility testing The fosfomycin MICs are shown in Table 1. Increasing MHA P i concentrations from 1 to 13 mM, 1e4 log 2 dilution differences were observed. Clinical category changes occurred for D glpT and Ec20190736, Ec20200150 and Ec20200178. The susceptible strains changed to resistant using 42 mM P i (1e8 log 2 dilution differences). The D uhpT and D glpT-uhpT mutants, Ec20190430 ( D glpT-uhpB) and Ec20190556 were resistant irrespective of the P i concentration. The isogenic mutant D phoB and the wild-type strain showed similar MICs at all P i concentrations tested. Similar results were observed with D glpT with a 1 log 2 and 2 log 2 increase at P i 13 and 42 mM, respectively, with respect to the wild-type strain. Fosfomycin activity The increased P i concentration did not cause significant changes in glpT or uhpT transcription (Fig. S1). Fosfomycin EC 50 under increasing P i concentrations are shown in Table 2 and Fig. S2. Concentration of 13 mM P i raised the fosfomycin EC 50 for the wild-type (1.42 vs.0.34 mg/L) and D glpT (1.62 vs. 0.32 mg/L) with respect to 1 mM P i . Concentrations of 42 mM P i also increased fosfomycin EC 50 with respect to 1 and 13 mM P i for the wild-type (8.9 vs.0.34 and 1.42 mg/L), D glpT (81.28 vs. 0.32 and 1.62 mg/L) and D uhpT (65.15 vs.11.75 and 24.79 mg/L). P i increased from 1 to 13 mM, showed a 2 log 2 increase in fosfomycin concentration Table 1 Fosfomycin MICs (mg/L) of the wild-type BW25113, single-gene and double-gene mutants and clinical isolates, tested by agar dilution in MHA with increasing P i concentrations Strains P i concentration 1mM (standard) 13mM 42mM BW25113 wild type 4 (S) 8 (S) 64 (R) BW25113 D glpT 4 (S) 16 (R) 256 (R) BW25113 D uhpT 128 (R) 256 (R) 512 (R) BW25113 D glpT-uhpT 512 (R) 512 (R) 1024 (R) BW25113 D phoB 4 (S) 8 (S) 128 (R) Ec20190430 512 (R) 1024 (R) 1024 (R) Ec20190556 64 (R) 128 (R) 1024 (R) Ec20190736 1 (S) 16 (R) 256 (R) Ec2019800 2 (S) 8 (S) 32 (R) Ec20200150 8 (S) 16 (R) 64 (R) Ec20200178 4 (S) 16 (R) 64 (R) Clinical categories for intravenous fosfomycin at 1 mM of P i (R 64 mg/L) and oral fosfomycin at P i 13 and 42 mM (R 16mg/L) are shown as susceptible (S) or resistant (R). Bold indicates changes in the clinical category. M. Ortiz-Padilla et al. / Clinical Microbiology and Infection 28 (2022) 302.e1e302.e4 302.e2 able to inhibit bacterial viability, a 4 log 2 dilution increase at 42 mM P i in the wild-type and D uhpT strains, and 8 log 2 increase in the D glpT strain. None of these changes were observed for D glpT-uhpT. The timeekill assay results (described in the supplementary material) showed that all fosfomycin concentrations exhibited bactericidal activity (>3 log CFU/mL decrease) against the wild-type strain, but fosfomycin activity was more affected in mutants, showing bacteriostatic activity or even bacterial regrowth at higher P i concentrations. Discussion The present study showed a reduction of fosfomycin in vitro activity as the P i concentration increased, focusing on P i and fosfomycin blood and urine concentrations where should exert its antimicrobial activity. This phenomenon was observed in susceptibility testing, viability and timeekill assays in both the isogenic and clinical isolate collections. Environmental variables and the presence of different carbon sources have a great impact on fosfomycin activity [10e13], and the importance of phosphate homeostasis in fosfomycin activity was recently highlighted by Turner et al., showing a strong selection of fosfomycin mutants for the phosphonate uptake and catabolism operon and the phosphate transporter [14]. In a previous study aimed at understanding how urinary tract conditions affected fosfomycin activity against E. coli strains harbouring chromosomal mutations involved in fosfomycin uptake, urine at pH 7 reduced fosfomycin activity, especially against the null mutants D glpT and D ptsI [12]. Among other factors, the different concentrations of P i could explain, in part, these observations. Recently, a dynamic bladder infection model simulating oral therapy using MHB, and human and synthetic urine, showed a lack of correlation between fosfomycin MICs determined using agar dilution or broth microdilution with MHB supplemented with glucose-6-phosphate and fosfomycin activity [15], underlining underlines the gap between fosfomycin susceptibility tests and outcome possibly by the relationship between bacterial metabolism and fosfomycin activity. The main limitation of our in vitro study is that MuellereHinton composition may have conditioned fosfomycin activity due to the lack of other components present in blood or urine. Also, the limited number of clinical strains with specific genotypes may have biased the conclusions. In conclusion, P i content modifies fosfomycin activity at concentrations achieved using standard oral or parenteral dosages. Furthermore, MuellereHinton media are suitable for fosfomycin susceptibility testing due to their low P i content, but may overestimate in vitro activity with respect to infection sites with higher P i levels. Both observations should be validated with a larger number of clinical isolates, together with their clinical outcomes especially in cases of P i homeostasis disorders, to establish the clinical significance. Transparency declaration This study has been funded by Instituto de Salud Carlos III through the projects PI16/01824, PI19/01645 and REIPI RD16/0016/ 0001 (co-funded by European Regional Development Fund “Away to make Europe”). Miriam Ortiz-Padilla is supported by a PFIS grant from the Instituto de Salud Carlos III (FI17/00235). In es Portillo- Calder on is supported by a Rio Hortega grant from the Instituto de Salud Carlos III (CM20/0092). (Co-funded by European Social Fund “Investing in your future”). The funders had no role in the design, collection of data, analysis and writing of the manuscript or the decision to publish. This study was presented in part at ECCMID 2021 (Abstract number: 00868). Author contributions F.D.P., J.M.R.M. and A.P.H. conceived the idea and experimental design. M.O.P., I.P.C. and B.G.I. conducted the phenotypic experiments. M.O.P., I.P.C. N.M.A. and B.G.I. conducted the molecular experiments. V.M.B. quantified P i concentrations in MuellereHinton. F.D.P. and A.P.H. supervised the project. M.O.P., I.P.C., J.M.R.M., J.R.B. and F.D.P. performed the analysis and interpreted the results. F.D.P., A.P.H. and M.O.P. wrote the manuscript. All authors contributed to manuscript revision and read and approved the submitted version. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.cmi.2021.09.037. References [1] Dijkmans AC, Zacarías NVO, Burggraaf J, Mouton JW, Wilms EB, van Nieuwkoop C, et al. Fosfomycin: pharmacological, clinical and future perspectives. Antibiotics 2017;6:24. [2] Keseler IM, Mackie A, Santos-Zavaleta A, Billington R, Bonavides-Martínez C, Caspi R, et al. The EcoCyc database: reflecting new knowledge about Escherichia coli K-12. Nucleic Acids Res 2017;45:D543e50. [3] McCleary WR. Molecular mechanisms of phosphate homeostasis in Escherichia coli. In: Escherichia coli erecent advances on physiology, pathogenesis and biotechnology applications. InTech; 2017. https://doi.org/10.5772/67283. [4] Goldman L, Schafer A. GoldmaneCecil Medicine. 26th ed. New York: Elsevier; 2019. [5] Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, et al. Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2006;2. 2006.0008. [6] Zaslaver A, Bren A, Ronen M, Itzkovitz S, Kikoin I, Shavit S, et al. A comprehensive library of fluorescent transcriptional reporters for Escherichia coli. Nat Methods 2006;3:623e8. Table 2 Summary of EC 50 values calculated from fosfomycin doseeresponse experiments generated by measuring bacterial viability over 24 hr at different P i concentrations P i concentration Fosfomycin EC 50 (95%CI) Wild type D glpT D uhpT D glpT-uhpT 1 mM 0.34 (0.26, 0.42) 0.32 (0.18, 0.48) 11.75 (6.7, 20.74) 260.4 (188.9, 360.5) 13 mM 1.42 a (1.1, 1.82) 1.62 a (1.15, 2.25) 24.79 (16.82, 36.41) 334.4 (240.2, 469.3) 42 mM 8.9 a,b (6.24, 12.56) 81.28 a,b (57.63, 114.1) 65.15 a,b (45.51, 92.62) 367.1 (248.4, 549.5) a p<0.05 with respect to 1 mM P i . b p<0.05 with respect to 13 mM P i . M. Ortiz-Padilla et al. / Clinical Microbiology and Infection 28 (2022) 302.e1e302.e4302.e3 [7] The European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters. Version 11.0. 2021. http://www.eucast.org. [8] Pfausler B. Concentrations of fosfomycin in the cerebrospinal fluid of neurointensive care patients with ventriculostomy-associated ventriculitis. J Antimicrob Chemother 2004;53:848e52. [9] Patel SS, Balfour JA, Bryson HM. Fosfomycin tromethamine. A review of its antibacterial activity, pharmacokinetic properties and therapeutic efficacy as a single-dose oral treatment for acute uncomplicated lower urinary tract infections. Drugs 1997;53:637e56. [10] Kahan FM, Kahan JS, Cassidy PJ, Kropp H. The mechanism of action of fosfomycin (phosphonomycin). Ann N Y Acad Sci 1974;235:364e86. [11] Kurabayashi K, Tanimoto K, Fueki S, Tomita H, Hirakawa H. Elevated expression of GlpT and UhpT via FNR activation contributes to increased fosfomycin susceptibility in Escherichia coli under anaerobic conditions. Antimicrob Agents Chemother 2015;59:6352e60. [12] Martin-Guti errez G,Docobo-P erez F, Rodriguez-Belt rn J, Rodriguez-Martinez JM, Aznar J, Pascual A, et al. Urinary tractconditions affect fosfomycin activity against Escherichia coli strains harboring chromosomal mutations involved in fosfomycin uptake. Antimicrob Agents Chemother 2017;62. e01899-17. [13] Xavier KB, Kossmann M, Santos H, Boos W. Kinetic analysis by in vivo 31P nuclear magnetic resonance of internal P i during the uptake of sn-glycerol-3- phosphate by the pho regulon-dependent Ugp system and the glp regulondependent GlpT system. J Bacteriol 1995;177:699e704. [14] Turner AK, Yasir M, Bastkowski S, Telatin A, Page AJ, Charles IG, et al. A genome-wide analysis of Escherichia coli responses to fosfomycin using TraDIS-Xpress reveals novel roles for phosphonate degradation and phosphate transport systems. J Antimicrob Chemother 2020;75:3144e51. [15] Abbott IJ, van Gorp E, Wijma RA, Meletiadis J, Mouton JW, Peleg AY. Evaluation of pooled human urine and synthetic alternatives in a dynamic bladder infection in vitro model simulating oral fosfomycin therapy. J Microbiol Methods 2020;171:105861. M. Ortiz-Padilla et al. / Clinical Microbiology and Infection 28 (2022) 302.e1e302.e4 302.e4 87 Supplementary Material of Role of inorganic phosphate concentrations in in vitro activity of fosfomycin. MATERIAL AND METHODS Bacterial growth medium Ion chromatography, using Metrohm 883 basic IC plus equipment with a Metrosep A supp 5 column and precolumn (Metrohm Hispania, Madrid, Spain), was carried out to determine Pi concentrations in standard Mueller Hinton media. Pi was identified by comparing the retention time established in the conditions of analysis with a certified standard, and quantified using a 7- point calibration line with R2=0.999 and average relative standard deviation of 1.943%, ≤ 5% in any case. Mueller-Hinton Broth (MHB) and Mueller-Hinton Agar (MHA) were used with different concentrations of Pi by supplementing standard Mueller-Hinton medium (Pi=1mM) with Na2HPO4 until the desired concentrations of 13 and 42mM of Pi were reached, corresponding to the minimum and maximum concentrations of Pi in urine, adjusted for a 1L volume (13-42mmol/24h) (1), and 28mM, corresponding to the average concentration of Pi in urine. Fosfomycin-transporters induction Fosfomycin transporter expressions were assayed by monitoring the fluorescence accumulation in E. coli MG1655 carrying pUA66-PglpT::gfpmut2 or pUA66-PuhpT::gfpmut2 reporters. Bacterial growth curves were performed using 5x105CFU/mL as starting inoculum in 96-well plates with 200μL of volume per well. MHB 1mM Pi (standard), MHB 13Mm or 42Mm Pi were used. Glucose- 6-phosphate (25mg/L) was added as control for uhpT gene expression. Green fluorescence (excited at 485nm and measured at 540nm) and bacterial growth (measured at 595nm) were monitored at 0 and 24 hours with an Infinite200 Pro plate reader (Tecan Group AG, Männedorf, Switzerland). Assays were performed in quintuplicate. Fluorescence was normalized to OD and the background was subtracted. The OD:fluorescence ratio of the promoterless construction (pUA66-Ø::gfpmut2) was used as background for all experiments under the different growth conditions. The results were compared with ANOVA and Tukey and Bonferroni post-hoc tests. Time-kill assays. Time-kill assays were conducted in duplicate, using fosfomycin concentrations of 0 (control), 307mg/L (Cmax concentration in plasma after an intravenous dose of 8g/Q8h) (2), 1053 and 4415mg/L (lower and upper limit for the mean Cmax observed in urine after a single dose of 3g of oral fosfomycin tromethamine) (3). The assay was done in standard MHB or supplemented with Pi (28mM, mean Pi urine concentration), with 25mg/L glucose-6-phosphate. Bacterial inoculum was 88 5x105CFU/mL in 20mL. The number of viable CFUs, after incubation with shaking at 37ºC, was determined at 0, 2, 4, 8, and 24h by serial dilution followed by plating on MHA plates. RESULTS Whole-genome sequencing Analyses of translated nucleotide sequences of genes associated with fosfomycin resistance or regulation (Table S1) showed no mutations in Crp and UhpA proteins in our collection of clinical isolates. No deletions or insertions were detected in the rest of the amino acid sequences associated with fosfomycin resistance, although many different polymorphisms were found, some of them conserved in most clinical isolates, except for the Ec20190430 isolate (fosfomycin MIC of 512 mg/L), which showed deletions in GlpT and UhpB, and Ec20190178 (fosfomycin MIC of 4 mg/L), which presented a deletion causing a frameshift mutation and a premature stop codon. With respect to proteins involved in the bacterial phosphate regulatory system (Table S2), only PitA and PhoB showed WT sequences. The rest of the amino acid sequences likewise showed many polymorphisms, some of them conserved in the clinical isolates. Nevertheless, the biological significance of many of the observed polymorphisms is unclear. Fosfomycin transport induction Peak fosfomycin transporter promoter activities (at 4h), evaluated at Pi concentrations of 1, 13 and 42mM ±25mg/L of glucose-6-phosphate, are shown in figure S1. The increased Pi concentration did not cause significant changes in glpT or uhpT transcription. Only the addition of glucose-6- phosphate significantly increased uhpT transcription. Time-kill assay results All fosfomycin concentrations showed bactericidal activity (>3log CFU/mL decrease) against the wild-type strain. It should be noted that 307mg/L of fosfomycin was unable to clear the bacterial culture at 28mM Pi. Similar results were observed in the ΔglpT strain at the low Pi concentration. At 28mM Pi, however, bacterial regrowth occurred at fosfomycin concentrations of 307 and 1053mg/L. With respect to the ΔuhpT strain, the bacterial counts dropped in the first 4h at all fosfomycin concentrations, although regrowth was observed at 307mg/L of fosfomycin after 8h with both Pi concentrations, and at 1053mg/L of fosfomycin with 28mM Pi. Finally, against the double mutant (ΔglpT-uhpT), not even 4015mg/L of fosfomycin was able to sterilise the bacterial culture with 28mM Pi after 24h. However, early fosfomycin activity was observed at 1053 and 4015mg/L with 89 both Pi concentrations, although bacterial regrowth was observed with the lowest fosfomycin concentration and a static effect with the highest. Table S1. Amino acid modifications in fosfomycin-related resistance proteins relative to reference strain BW25113. del = deletion. 90 Table S2. Amino acid modifications in genes related to phosphate uptake relative to reference strain BW25113. del = deletion. 91 Figure S1. Promoter activity of fosfomycin transporters glpT and uhpT at different concentrations of Pi. Glucose-6-phosphate (G6P, 25mg/L) was used as positive control for uhpT expression. Bars and whiskers indicate mean and standard deviation. a P<0.05 with respect to the absence of G6P. Figure S2. Viability after 24h of wild-type E. coli BW25113 and isogenic mutant strains (ΔglpT , ΔuhpT and ΔglpT-uhpT) against fosfomycin concentrations ranging from 0-1024 mg/L in standard MHB (Pi=1 mM) and supplemented with Pi (final concentration of 13 and 42mM). Lines and symbols indicate mean values (±SD) of viability measurements. 92 Figure S3. 24-hour time-kill results for wild-type E. coli BW25113 and isogenic mutant strains (ΔglpT , ΔuhpT and ΔglpT-uhpT) using 307, 1053 and 4015 mg/L fosfomycin in standard MHB (Pi=1 mM, continuous lines) and supplemented with Pi (final Pi concentration of 28mM, dashed lines) . REFERENCES 1. Goldman L (Columbia U, Schafer A (Weill CMC. Goldman-Cecil Medicine 26th Edition. Elsevier. New York; 2019. 2944 p. 2. Pfausler B. Concentrations of fosfomycin in the cerebrospinal fluid of neurointensive care patients with ventriculostomy-associated ventriculitis. J Antimicrob Chemother [Internet]. 2004 Mar 24;53(5):848–52. Available from: http://www.jac.oupjournals.org/cgi/doi/10.1093/jac/dkh158 3. Patel SS, Balfour JA, Bryson HM. Fosfomycin tromethamine. A review of its antibacterial activity, pharmacokinetic properties and therapeutic efficacy as a single-dose oral treatment for acute uncomplicated lower urinary tract infections. Drugs. 1997 Apr;53(4):637–56. 93 VI. DISCUSSION The progressive increase in antimicrobial resistance is a worldwide public health problem that science is trying to find a solution under different ways18. The current global interest in the recovery of old antimicrobials and the optimization of their use for the treatment of infections caused by multidrugresistant microorganisms is the result of the scarcity of therapeutic alternatives. This strategy has been proposed by international organizations such as the WHO or the ECDC, focusing in certain antimicrobials 21,80,225. These old antimicrobials have remained active and in some cases with low resistance profiles over the years due to their low use, motivated by the emergence of more promising drugs80. In this sense, fosfomycin is an antimicrobial discovered more than 50 years ago and is still active against bacteria with multiple resistance mechanisms194. Thus, the systemic use of this antibiotic belongs to the options in the therapeutic arsenal used against infections caused by multidrug resistant bacteria according to the EMA83. Despite the availability of this drug for several decades, its use as intravenous formulation has been limited until the last decade. One of the restricting factors in the use of this drug has been the peculiarities and controversies in the development and interpretation of the susceptibility studies, a key factor for the prediction of the therapeutic success or failure of an antimicrobial treatment226,227. In this sense, the CLSI and EUCAST committees establish that the reference method for determiningthe susceptibility or resistance to fosfomycin is the dilution in agar supplemented with G6P (25 mg/L)201,228. However, this procedure is unfeasible for its development in the daily routine of the clinical microbiology laboratory, for this reason the use of better adapted, easier, and faster alternative methodologies (disc-diffusion, broth microdilution, or gradient strips) have been adopted by the laboratories. These adaptations from the reference method differ and originates discrepant susceptibility data depending on the methodology used227,229. This produces uncertainty in the extrapolation of the results into the clinic. This may have contributed to some extent to the generation of some mistrust in the use of this drug as a first-line systemic treatment. Considering the potential relevance of this antimicrobial as a potential alternative use for systemic infections caused by multidrug resistant bacteria, it is essential to understand the factors that can alter its mechanism of action, like how fosfomycin is transported into the bacteria, how they are regulated (a main step obtain its full activity), or whether there are molecules that can optimize and improve fosfomycin activity. All this, aimed to be able to predict the efficacy of the drug in a more adequate and individualized way and to increase the chance of therapeutic success. 100 VII. CONCLUSIONS 1. The inactivation of the fosA gene show decreased fosfomycin resistance in K. pneumoniae clinical isolates, leading to a susceptibility level similar to that found in E. coli. 2. The pharmacological approach with sodium phosphonoformate did not achieve enough activity and the effect decreased with the presence of fosfomycin resistance mutations. 3. Glycerol induces transcription of the GlpT transporter, has a synergistic effect in combination with fosfomycin and increases fosfomycin susceptibility in wild-type and mutant strains such as in the UhpT transporter gene at therapeutically relevant concentrations. 4. Glycerol does not appear to be a suitable molecule for use as an adjuvant together with fosfomycin, as the induction of the glpT transporter is slow, resulting in emergence of resistant subpopulations. 5. Extracellular high inorganic phosphate concentration increases fosfomycin resistance. 6. This increase is not mediated by direct regulation of GlpT and UhpT transporters and occurs at physiological inorganic phosphate concentrations. 101 VIII. BIBLIOGRAPHY 1. Aminov, R. I. A brief history of the antibiotic era: lessons learned and challenges for the future. Front. Microbiol. 1, (2010). 2. Uso y abuso de los antibióticos. ¿Dónde estamos y adonde queremos llegar? (2006). 3. CDC. 2019 Antibiotic Resistance Threats Report | CDC. https://www.cdc.gov/drugresistance/biggest-threats.html (2019). 4. Das, B., Chaudhuri, S., Srivastava, R., Nair, G. B. & Ramamurthy, T. Fostering research into antimicrobial resistance in India. BMJ 358, j3535 (2017). 5. Clemente, J. C. et al. The microbiome of uncontacted Amerindians. Sci. Adv. 1, (2015). 6. Bhullar, K. et al. Antibiotic resistance is prevalent in an isolated cave microbiome. PLoS One 7, (2012). 7. Davies, J., Spiegelman, G. B. & Yim, G. The world of subinhibitory antibiotic concentrations. Curr. Opin. Microbiol. 9, 445–453 (2006). 8. Fajardo, A. & Martínez, J. L. Antibiotics as signals that trigger specific bacterial responses. Curr. Opin. Microbiol. 11, 161–167 (2008). 9. Aminov, R. I. The role of antibiotics and antibiotic resistance in nature. Environ. Microbiol. 11, 2970–2988 (2009). 10. Perry, J. A., Westman, E. L. & Wright, G. D. The antibiotic resistome: what’s new? Curr. Opin. Microbiol. 21, 45–50 (2014). 11. Kim, D. W. & Cha, C. J. Antibiotic resistome from the One-Health perspective: understanding and controlling antimicrobial resistance transmission. 53, 301–309 (2021). 12. Partridge, S. R., Kwong, S. M., Firth, N. & Jensen, S. O. Mobile Genetic Elements Associated with Antimicrobial Resistance. Clin. Microbiol. Rev. 31, (2018). 13. Cox, G. & Wright, G. D. Intrinsic antibiotic resistance: mechanisms, origins, challenges and solutions. Int. J. Med. Microbiol. 303, 287–292 (2013). 14. Frère, J.-M., Sauvage, E. & Kerff, F. From ‘An Enzyme Able to Destroy Penicillin’ to Carbapenemases: 70 Years of Beta-lactamase Misbehaviour. Curr. Drug Targets 17, 974–982 (2016). 15. J, D. & D, D. Origins and evolution of antibiotic resistance. Microbiol. Mol. Biol. Rev. 74, 9–16 (2010). 16. Hiltunen, T., Virta, M. & Anna-Liisa, L. Antibiotic resistance in the wild: an eco-evolutionary perspective. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 372, (2017). 17. Baquero, F., Tedim, A. P. & Coque, T. M. Antibiotic resistance shaping multi-level population biology of bacteria. Front. Microbiol. 4, (2013). 18. O’Neil, J. Tackling drug-resistant infection globally: final report and recommendations the review on antimicrobial resistant chaired by Jim O’Neil. (2016). 19. Cassini, A. et al. Attributable deaths and disability-adjusted life-years caused by infections with antibiotic-resistant bacteria in the EU and the European Economic Area in 2015: a populationlevel modelling analysis. 19, 56–66 (2019). 102 20. Falagas, M. E., Tansarli, G. S., Karageorgopoulos, D. E. & Vardakas, K. Z. Deaths attributable to carbapenem-resistant Enterobacteriaceae infections. Emerg. Infect. Dis. 20, 1170–1175 (2014). 21. WHO. WHO Library Cataloguing-in-Publication Data Global Action Plan on Antimicrobial Resistance. Microbe Mag. 10, 354–355 (2015). 22. Mathers, A. J., Peirano, G. & Pitout, J. D. D. The role of epidemic resistance plasmids and international high-risk clones in the spread of multidrug-resistant Enterobacteriaceae. Clin. Microbiol. Rev. 28, 565–591 (2015). 23. Carattoli, A. Plasmids and the spread of resistance. Int. J. Med. Microbiol. 303, 298–304 (2013). 24. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet (London, England) 399, 629–655 (2022). 25. European Centre for Disease Prevention and Control. Antimicrobial resistance in the EU/EEA (EARS-Net) - Annual Epidemiological Report 2021. Stockholm: ECDC; 2022. 26. Daneman, N. et al. Variability in Antibiotic Use Across Nursing Homes and the Risk of Antibiotic- Related Adverse Outcomes for Individual Residents. JAMA Intern. Med. 175, 1331–1339 (2015). 27. Klein, E. Y. et al. Global increase and geographic convergence in antibiotic consumption between 2000 and 2015. Proc. Natl. Acad. Sci. U. S. A. 115, E3463–E3470 (2018). 28. WHO. Monitoring and evaluation of the global action plan on antimicrobial resistance: framework and recommended indicators. https://apps.who.int/iris/handle/10665/325006 (2021). 29. ISCIII (Instituto de Salud Carlos III). Memoria 2020 de la Red de vigilancia EARS-Net. (2020). 30. Robinson, T. P. et al. Antibiotic resistance is the quintessential One Health issue. Trans. R. Soc. Trop. Med. Hyg. 110, 377–380 (2016). 31. WHO. The evolving threat of antimicrobial resistance: options for action. https://apps.who.int/iris/handle/10665/44812 (2014). 32. ECDC, E. C. for D. P. and C. Antimicrobial resistance in the EU/EEA (EARS-Net) - Annual Epidemiological Report for 2020. https://www.ecdc.europa.eu/en/publicationsdata/antimicrobial-resistance-eueea-ears-net-annual-epidemiological-report-2020 (2020). 33. AEMPS. Plan estratégico y de acción para reducir el riesgo de selección y diseminación de la resistencia a los antibioticos. (2014). 34. Storr, J. et al. Core components for effective infection prevention and control programmes: new WHO evidence-based recommendations. Antimicrob. Resist. Infect. Control 6, (2017). 35. Tacconelli, E. et al. ESCMID guidelines for the management of the infection control measures to reduce transmission of multidrug-resistant Gram-negative bacteria in hospitalized patients. Clin. Microbiol. Infect. 20 Suppl 1, 1–55 (2014). 36. Aiello, A. E., Coulborn, R. M., Perez, V. & Larson, E. L. Effect of hand hygiene on infectious disease risk in the community setting: a meta-analysis. Am. J. Public Health 98, 1372–1381 (2008). 37. Vikesland, P. et al. Differential Drivers of Antimicrobial Resistance across the World. Acc. Chem. Res. 52, 916–924 (2019). 38. Bürgmann, H. et al. Water and sanitation: an essential battlefront in the war on antimicrobial resistance. FEMS Microbiol. Ecol. 94, (2018). 103 39. Dellit, T. H. et al. Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin. Infect. Dis. 44, 159–177 (2007). 40. Rodríguez-Baño, J. et al. Programas de optimización de uso de antimicrobianos (PROA) en hospitales españoles: documento de consenso GEIH-SEIMC, SEFH y SEMPSPH. Enferm. Infecc. Microbiol. Clin. 30, 22.e1-22.e23 (2012). 41. Barlam, T. F. The state of antibiotic stewardship programs in 2021: The perspective of an experienced steward. Antimicrob. Steward. Healthc. Epidemiol. ASHE 1, (2021). 42. Schuts, E. C. et al. Current evidence on hospital antimicrobial stewardship objectives: a systematic review and meta-analysis. Lancet. Infect. Dis. 16, 847–856 (2016). 43. OECD. Stemming the Superbug Tide. https://www.oecd-ilibrary.org/social-issues-migration- health/stemming-the-superbug-tide_9789264307599-en (2018) doi:10.1787/9789264307599- EN. 44. Servicio Andaluz de Salud. Programa integral de prevención, control de las infecciones relacionadas con la asistencia sanitaria, y uso apropiado de los antimicrobianos. https://www.juntadeandalucia.es/servicioandaluzdesalud/ (2014). 45. Cantas, L. et al. A brief multi-disciplinary review on antimicrobial resistance in medicine and its linkage to the global environmental microbiota. Front. Microbiol. 4, (2013). 46. MOORE, P. R. & EVENSON, A. Use of sulfasuxidine, streptothricin, and streptomycin in nutritional studies with the chick. J. Biol. Chem. 165, 437–441 (1946). 47. Anderson, A. D., Nelson, J. M., Rossiter, S. & Angulo, F. J. Public health consequences of use of antimicrobial agents in food animals in the United States. Microb. Drug Resist. 9, 373–379 (2003). 48. Góchez, D. et al. OIE Annual Report on Antimicrobial Agents Intended for Use in Animals: Methods Used. Front. Vet. Sci. 6, (2019). 49. Muloi, D. et al. Are Food Animals Responsible for Transfer of Antimicrobial-Resistant Escherichia coli or Their Resistance Determinants to Human Populations? A Systematic Review. Foodborne Pathog. Dis. 15, 467–474 (2018). 50. (EMA), E. medicines A. Sales of veterinary antimicrobial agents in 31 European countries in 2019 and 2020. (2021) doi:10.2809/167341. 51. Walsh, T. R. & Wu, Y. China bans colistin as a feed additive for animals. Lancet. Infect. Dis. 16, 1102–1103 (2016). 52. Liu, Y. Y. et al. Emergence of plasmid-mediated colistin resistance mechanism MCR-1 in animals and human beings in China: a microbiological and molecular biological study. Lancet. Infect. Dis. 16, 161–168 (2016). 53. Van Boeckel, T. P. et al. Reducing antimicrobial use in food animals. Science 357, 1350–1352 (2017). 54. World Health Organisation (WHO). Critically Important Antimicrobials for Human Medicine 6th Re.vision 2018. Ranking of medically important antimicrobials for risk management of antimicrobial resistance due to non-human use. Https://Apps.Who.Int/Iris/Bitstream/Handle/10665/312266/9789241515528-Eng.Pdf?Ua=1 ISBN 978-92-4-151552-8 (2018). 104 55. (FAO), F. and agriculture organization. The FAO Action Plan on Antimicrobial Resistance 2016- 2020. https://www.fao.org/fsnforum/resources/reports-and-briefs/fao-action-plan- antimicrobial-resistance-2016-2020 (2021). 56. Lewis, K. Platforms for antibiotic discovery. Nat. Rev. Drug Discov. 2013 125 12, 371–387 (2013). 57. Singer, A. C., Kirchhelle, C. & Roberts, A. P. (Inter)nationalising the antibiotic research and development pipeline. Lancet. Infect. Dis. 20, e54–e62 (2020). 58. Tacconelli, E. et al. Discovery, research, and development of new antibiotics: the WHO priority list of antibiotic-resistant bacteria and tuberculosis. Lancet. Infect. Dis. 18, 318–327 (2018). 59. WHO. 2019 antibacterial agents in clinical development: an analysis of the antibacterial clinical development pipeline. https://www.who.int/publications/i/item/9789240000193 (2019). 60. The Pew Charitable Trust. Tracking the Global Pipeline of Antibiotics in Development, March 2021 | The Pew Charitable Trusts. https://www.pewtrusts.org/en/research-and-analysis/issuebriefs/2021/03/tracking-the-global-pipeline-of-antibiotics-in-development (2021). 61. Niu, G. & Li, W. Next-Generation Drug Discovery to Combat Antimicrobial Resistance. Trends Biochem. Sci. 44, 961–972 (2019). 62. Czaplewski, L. et al. Alternatives to antibiotics-a pipeline portfolio review. Lancet. Infect. Dis. 16, 239–251 (2016). 63. Yang, J. H., Bening, S. C. & Collins, J. J. Antibiotic efficacy-context matters. Curr. Opin. Microbiol. 39, 73–80 (2017). 64. Bush, K. & Bradford, P. A. β-Lactams and β-Lactamase Inhibitors: An Overview. Cold Spring Harb. Perspect. Med. 6, (2016). 65. DIckey, S. W., Cheung, G. Y. C. & Otto, M. Different drugs for bad bugs: antivirulence strategies in the age of antibiotic resistance. Nat. Rev. Drug Discov. 16, 457–471 (2017). 66. Kaufmann, S. H. E., Dorhoi, A., Hotchkiss, R. S. & Bartenschlager, R. Host-directed therapies for bacterial and viral infections. Nat. Rev. Drug Discov. 17, 35–56 (2018). 67. Lee, H. H. & Collins, J. J. Microbial environments confound antibiotic efficacy. Nat. Chem. Biol. 8, 6–9 (2011). 68. Farha, M. A. & Brown, E. D. Drug repurposing for antimicrobial discovery. Nat. Microbiol. 4, 565– 577 (2019). 69. Santacroce, L., Charitos, I. A. & Bottalico, L. A successful history: probiotics and their potential as antimicrobials. Expert Rev. Anti. Infect. Ther. 17, 635–645 (2019). 70. Golkar, Z., Bagasra, O. & Gene Pace, D. Bacteriophage therapy: a potential solution for the antibiotic resistance crisis. J. Infect. Dev. Ctries. 8, 129–136 (2014). 71. Vázquez-Ucha, J. C., Arca-Suárez, J., Bou, G. & Beceiro, A. New Carbapenemase Inhibitors: Clearing the Way for the β-Lactams. Int. J. Mol. Sci. 21, 1–32 (2020). 72. Tomich, A. D. et al. Small-Molecule Inhibitor of FosA Expands Fosfomycin Activity to Multidrug- Resistant Gram-Negative Pathogens. Antimicrob. Agents Chemother. 63, (2019). 73. Savage, P. B. Multidrug-resistant bacteria: overcoming antibiotic permeability barriers of gramnegative bacteria. Ann. Med. 33, 167–171 (2001). 74. Blanco, P., Sanz-García, F., Hernando-Amado, S., Martínez, J. L. & Alcalde-Rico, M. The 105 development of efflux pump inhibitors to treat Gram-negative infections. Expert Opin. Drug Discov. 13, 919–931 (2018). 75. Antunes, L. C. M., Ferreira, R. B. R., Buckner, M. M. C. & Finlay, B. B. Quorum sensing in bacterial virulence. Microbiology 156, 2271–2282 (2010). 76. Stokes, J. M., Lopatkin, A. J., Lobritz, M. A. & Collins, J. J. Bacterial Metabolism and Antibiotic Efficacy. Cell Metab. 30, 251–259 (2019). 77. Liu, Y., Li, R., Xiao, X. & Wang, Z. Bacterial metabolism-inspired molecules to modulate antibiotic efficacy. J. Antimicrob. Chemother. 74, 3409–3417 (2019). 78. de la Fuente-Núñez, C., Reffuveille, F., Haney, E. F., Straus, S. K. & Hancock, R. E. W. Broadspectrum anti-biofilm peptide that targets a cellular stress response. PLoS Pathog. 10, (2014). 79. Kottur, J. & Nair, D. T. Reactive Oxygen Species Play an Important Role in the Bactericidal Activity of Quinolone Antibiotics. Angew. Chem. Int. Ed. Engl. 55, 2397–2400 (2016). 80. Zayyad, H., Eliakim-Raz, N., Leibovici, L. & Paul, M. Revival of old antibiotics: needs, the state of evidence and expectations. Int. J. Antimicrob. Agents 49, 536–541 (2017). 81. Theuretzbacher, U. et al. Reviving old antibiotics. J. Antimicrob. Chemother. 70, 2177–2181 (2015). 82. Mouton, J. W. et al. Conserving antibiotics for the future: new ways to use old and new drugs from a pharmacokinetic and pharmacodynamic perspective. Drug Resist. Updat. 14, 107–117 (2011). 83. Karaiskos, I., Lagou, S., Pontikis, K., Rapti, V. & Poulakou, G. The ‘Old’ and the ‘New’ Antibiotics for MDR Gram-Negative Pathogens: For Whom, When, and How. Front. public Heal. 7, (2019). 84. Sheu, C. C., Chang, Y. T., Lin, S. Y., Chen, Y. H. & Hsueh, P. R. Infections Caused by Carbapenem- Resistant Enterobacteriaceae: An Update on Therapeutic Options. Front. Microbiol. 10, (2019). 85. López-Montesinos, I. & Horcajada, J. P. Oral and intravenous fosfomycin in complicated urinary tract infections. Rev. Esp. Quimioter. 32 Suppl 1, 37–44 (2019). 86. Muller, A. E., Verhaegh, E. M., Harbarth, S., Mouton, J. W. & Huttner, A. Nitrofurantoin’s efficacy and safety as prophylaxis for urinary tract infections: a systematic review of the literature and meta-analysis of controlled trials. Clin. Microbiol. Infect. 23, 355–362 (2017). 87. Hendlin, D. et al. Phosphonomycin, a new antibiotic produced by strains of streptomyces. Science 166, 122–123 (1969). 88. Shoji, J. et al. Production of fosfomycin (phosphonomycin) by Pseudomonas syringae. J. Antibiot. (Tokyo). 39, 1011–1012 (1986). 89. Cameron, S., McLuskey, K., Chamberlayne, R., Hallyburton, I. & Hunter, W. N. Initiating a crystallographic analysis of recombinant (S)-2-hydroxypropylphosphonic acid epoxidase from Streptomyces wedmorensis. Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun. 61, 534 (2005). 90. Kim, S. Y. et al. Different Biosynthetic Pathways to Fosfomycin in Pseudomonas syringae and Streptomyces Species. Antimicrob. Agents Chemother. 56, 4175 (2012). 91. Cao, Y., Peng, Q., Li, S., Deng, Z. & Gao, J. The intriguing biology and chemistry of fosfomycin: the only marketed phosphonate antibiotic. RSC Adv. 9, 42204 (2019). 92. Kahan, F. M., Kahan, J. S., Cassidy, P. J. & Kropp, H. The mechanism of action of fosfomycin (phosphonomycin). Ann. N. Y. Acad. Sci. 235, 364–386 (1974). 106 93. Andrews, J. M. et al. International collaborative study on standardization of bacterial sensitivity to fosfomycin. J. Antimicrob. Chemother. 12, 357–361 (1983). 94. Grimm, H. In vitro investigations with fosfomycin on Mueller-Hinton agar with and without glucose-6-phosphate. Infection 7, 256–259 (1979). 95. Bergan, T. Degree of absorption, pharmacokinetics of fosfomycin trometamol and duration of urinary antibacterial activity. Infection 18 Suppl 2, (1990). 96. Gupta, K. et al. International clinical practice guidelines for the treatment of acute uncomplicated cystitis and pyelonephritis in women: A 2010 update by the Infectious Diseases Society of America and the European Society for Microbiology and Infectious Diseases. Clin. Infect. Dis. 52, (2011). 97. EMA. Recommendations to restrict use of fosfomycin antibiotics | European Medicines Agency. https://www.ema.europa.eu/en/news/recommendations-restrict-use-fosfomycin-antibiotics (2020). 98. Nabriva Therapeutics Resubmits New Drug Application for Intravenous CONTEPOTM (fosfomycin) for Injection | Nabriva Therapeutics. https://investors.nabriva.com/news-releases/newsrelease-details/nabriva-therapeutics-resubmits-new-drug-application-intravenous (2019). 99. Soledad Pérez, D., Pérez, D., Tapia, M. & Soraci, A. Fosfomycin: Uses and potentialities in veterinary medicine. Open Vet. J. 4, 26 (2014). 100. Gobernado, M. A.-Sociedad Española de Quimioterapia Revisión Fosfomicina. Marzo 16, 15–40 (2003). 101. Cao, Y., Peng, Q., Li, S., Deng, Z. & Gao, J. The intriguing biology and chemistry of fosfomycin: the only marketed phosphonate antibiotic. RSC Adv. 9, 42204–42218 (2019). 102. Kahan, F. M., Kahan, J. S., Cassidy, P. J. & Kropp, H. The mechanism of action of fosfomycin (phosphonomycin). Ann. N. Y. Acad. Sci. 235, 364–386 (1974). 103. Falagas, M. E., Vouloumanou, E. K., Samonis, G. & Vardakas, K. Z. Fosfomycin. Clin. Microbiol. Rev. 29, 321 LP – 347 (2016). 104. Brown, E. D., Vivas, E. I., Walsh, C. T. & Kolter, R. MurA (MurZ), the enzyme that catalyzes the first committed step in peptidoglycan biosynthesis, is essential in Escherichia coli. J. Bacteriol. 177, 4194–4197 (1995). 105. Skarzynski, T. et al. Structure of UDP-N-acetylglucosamine enolpyruvyl transferase, an enzyme essential for the synthesis of bacterial peptidoglycan, complexed with substrate UDP-N- acetylglucosamine and the drug fosfomycin. Structure 4, 1465–1474 (1996). 106. Eschenburg, S., Priestman, M. & Schönbrunn, E. Evidence that the fosfomycin target Cys115 in UDP-N-acetylglucosamine enolpyruvyl transferase (MurA) is essential for product release. J. Biol. Chem. 280, 3757–3763 (2005). 107. Castañeda-García, A., Blázquez, J. & Rodríguez-Rojas, A. Molecular mechanisms and clinical impact of acquired and intrinsic fosfomycin resistance. Antibiotics 2, 217–236 (2013). 108. Vardakas, K. Z., Legakis, N. J., Triarides, N. & Falagas, M. E. Susceptibility of contemporary isolates to fosfomycin: a systematic review of the literature. Int. J. Antimicrob. Agents 47, 269– 285 (2016). 109. Marger, M. D. & Saier, M. H. A major superfamily of transmembrane facilitators that catalyse uniport, symport and antiport. Trends Biochem. Sci. 18, 13–20 (1993). 107 110. Ambudkar, S. V., Larson, T. J. & Maloney, P. C. Reconstitution of sugar phosphate transport systems of Escherichia coli. J. Biol. Chem. 261, 9083–9086 (1986). 111. Grainger, D. C., Hurd, D., Harrison, M., Holdstock, J. & Busby, S. J. W. Studies of the distribution of Escherichia coli cAMP-receptor protein and RNA polymerase along the E. coli chromosome. Proc. Natl. Acad. Sci. U. S. A. 102, 17693 (2005). 112. Zubay, G., Schwartz, D. & Beckwith, J. Mechanism of activation of catabolite-sensitive genes: a positive control system. Proc. Natl. Acad. Sci. U. S. A. 66, 104–110 (1970). 113. Franchini, A. G., Ihssen, J. & Egli, T. Effect of Global Regulators RpoS and Cyclic-AMP/CRP on the Catabolome and Transcriptome of Escherichia coli K12 during Carbon- and Energy-Limited Growth. PLoS One 10, (2015). 114. Balsalobre, C., Johansson, J. & Uhlin, B. E. Cyclic AMP-dependent osmoregulation of crp gene expression in Escherichia coli. J. Bacteriol. 188, 5935–5944 (2006). 115. Sakamoto, Y., Furukawa, S., Ogihara, H. & Yamasaki, M. Fosmidomycin resistance in adenylate cyclase deficient (cya) mutants of Escherichia coli. Biosci. Biotechnol. Biochem. 67, 2030–2033 (2003). 116. Görke, B. & Stülke, J. Carbon catabolite repression in bacteria: many ways to make the most out of nutrients. Nat. Rev. Microbiol. 6, 613–624 (2008). 117. Deutscher, J. The mechanisms of carbon catabolite repression in bacteria. Curr. Opin. Microbiol. 11, 87–93 (2008). 118. Fann, M. C. et al. Identification of two essential arginine residues in UhpT, the sugar phosphate antiporter of Escherichia coli. J. Membr. Biol. 164, 187–195 (1998). 119. Lloyd, A. D. & Kadner, R. J. Topology of the Escherichia coli uhpT sugar-phosphate transporter analyzed by using TnphoA fusions. J. Bacteriol. 172, 1688–1693 (1990). 120. Kadner, R. J. & Shattuck Eidens, D. M. Genetic control of the hexose phosphate transport system of Escherichia coli: mapping of deletion and insertion mutations in the uhp region. J. Bacteriol. 155, 1052–1061 (1983). 121. Island, M. D. & Kadner, R. J. Interplay between the membrane-associated UhpB and UhpC regulatory proteins. J. Bacteriol. 175, 5028–5034 (1993). 122. Chen, Q. & Kadner, R. J. Effect of altered spacing between uhpT promoter elements on transcription activation. J. Bacteriol. 182, 4430–4436 (2000). 123. Kurabayashi, K., Tanimoto, K., Fueki, S., Tomita, H. & Hirakawa, H. Elevated expression of GlpT and UhpT via FNR activation contributes to increased fosfomycin susceptibility in Escherichia coli under anaerobic conditions. Antimicrob. Agents Chemother. 59, 6352–6360 (2015). 124. Kurabayashi, K., Hirakawa, Y., Tanimoto, K., Tomita, H. & Hirakawa, H. Role of the CpxAR twocomponent signal transduction system in control of fosfomycin resistance and carbon substrate uptake. J. Bacteriol. 196, 248–256 (2014). 125. Verhamme, D. T., Postma, P. W., Crielaard, W. & Hellingwerf, K. J. Cooperativity in signal transfer through the Uhp system of Escherichia coli. J. Bacteriol. 184, 4205–4210 (2002). 126. Xavier, K. B., Kossmann, M., Santos, H. & Boos, W. Kinetic analysis by in vivo 31P nuclear magnetic resonance of internal Pi during the uptake of sn-glycerol-3-phosphate by the pho regulon-dependent Ugp system and the glp regulon-dependent GlpT system. J. Bacteriol. 177, 699–704 (1995). 108 127. Ortiz-Padilla, M. et al. Role of inorganic phosphate concentrations in in vitro activity of fosfomycin. Clin. Microbiol. Infect. 28, 302.e1-302.e4 (2022). 128. Fann, M. C. & Maloney, P. C. Functional symmetry of UhpT, the sugar phosphate transporter of Escherichia coli. J. Biol. Chem. 273, 33735–33740 (1998). 129. Grimm, H. In vitro investigations with fosfomycin on Mueller-Hinton agar with and without glucose-6-phosphate. Infection 7, 256–259 (1979). 130. Huang, Y., Lemieux, M. J., Song, J., Auer, M. & Wang, D. N. Structure and mechanism of the glycerol-3-phosphate transporter from Escherichia coli. Science 301, 616–620 (2003). 131. Lemieux, M. J., Huang, Y. & Wang, D. N. Crystal structure and mechanism of GlpT, the glycerol- 3-phosphate transporter from E. coli. J. Electron Microsc. (Tokyo). 54, i43–i46 (2005). 132. Cole, S. T. et al. Nucleotide sequence and gene-polypeptide relationships of the glpABC operon encoding the anaerobic sn-glycerol-3-phosphate dehydrogenase of Escherichia coli K-12. J. Bacteriol. 170, 2448–2456 (1988). 133. Yang, B., Gerhardt, S. G. & Larson, T. J. Action at a distance for glp repressor control of glpTQ transcription in Escherichia coli K-12. Mol. Microbiol. 24, 511–521 (1997). 134. Yang, B. & Larson, T. J. Multiple promoters are responsible for transcription of the glpEGR operon of Escherichia coli K-12. Biochim. Biophys. Acta 1396, 114–126 (1998). 135. Iuchi, S., Cole, S. T. & Lin, E. C. C. Multiple regulatory elements for the glpA operon encoding anaerobic glycerol-3-phosphate dehydrogenase and the glpD operon encoding aerobic glycerol- 3-phosphate dehydrogenase in Escherichia coli: further characterization of respiratory control. J. Bacteriol. 172, 179–184 (1990). 136. Truniger, V., Boos, W. & Sweet, G. Molecular analysis of the glpFKX regions of Escherichia coli and Shigella flexneri. J. Bacteriol. 174, 6981–6991 (1992). 137. Lin, E. C. Glycerol dissimilation and its regulation in bacteria. Annu. Rev. Microbiol. 30, 535–78 (1976). 138. Koch, J. P., Hayashi, S. & Lin, E. C. C. The Control of Dissimilation of Glycerol and l-α- Glycerophosphate in Escherichia coli. J. Biol. Chem. 239, 3106–3108 (1964). 139. Nikel, P. I., Romero-Campero, F. J., Zeidman, J. A., Goñi-Moreno, Á. & de Lorenzo, V. The glyceroldependent metabolic persistence of Pseudomonas putida KT2440 reflects the regulatory logic of the GlpR repressor. MBio 6, (2015). 140. Nikel, P. I., Kim, J. & de Lorenzo, V. Metabolic and regulatory rearrangements underlying glycerol metabolism in Pseudomonas putidaKT2440. Environ. Microbiol. 16, 239–254 (2014). 141. Freedberg, W. B. & Lin, E. C. C. Three kinds of controls affecting the expression of the glp regulon in Escherichia coli. J. Bacteriol. 115, 816–823 (1973). 142. Takahata, S. et al. Molecular mechanisms of fosfomycin resistance in clinical isolates of Escherichia coli. Int. J. Antimicrob. Agents 35, 333–337 (2010). 143. Santoro, A. et al. Interaction of fosfomycin with the Glycerol 3-phosphate Transporter of Escherichia coli. Biochim. Biophys. Acta - Gen. Subj. 1810, 1323–1329 (2011). 144. McCleary, W. R. Molecular Mechanisms of Phosphate Homeostasis in Escherichia coli. in Escherichia coli - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications (InTech, 2017). doi:10.5772/67283. 109 145. Hsieh, Y.-J. & Wanner, B. L. Global regulation by the seven-component Pi signaling system. Curr. Opin. Microbiol. 13, 198–203 (2010). 146. Gardner, S. G. & McCleary, W. R. Control of the phoBR Regulon in Escherichia coli. EcoSal Plus 8, (2019). 147. Harris, R. M., Webb, D. C., Howitt, S. M. & Cox, G. B. Characterization of PitA and PitB from<em>Escherichia coli</em> J. Bacteriol. 183, 5008 LP – 5014 (2001). 148. Elvin, C. M., Hardy, C. M. & Rosenberg, H. P(i) exchange mediated by the GlpT-dependent snglycerol-3-phosphate transport system in Escherichia coli. J. Bacteriol. 161, 1054–1058 (1985). 149. Ambudkar, S. V., Anantharam, V. & Maloney, P. C. UhpT, the sugar phosphate antiporter of Escherichia coli, functions as a monomer. J. Biol. Chem. 265, 12287–12292 (1990). 150. Motomura, K. et al. Overproduction of YjbB reduces the level of polyphosphate in Escherichia coli: A hypothetical role of YjbB in phosphate export and polyphosphate accumulation. FEMS Microbiol. Lett. 320, 25–32 (2011). 151. Kumar, S., Parvathi, A., Hernandez, R. L., Cadle, K. M. & Varela, M. F. Identification of a novel UDP-N-acetylglucosamine enolpyruvyl transferase (MurA) from Vibrio fischeri that confers high fosfomycin resistance in Escherichia coli. Arch. Microbiol. 191, 425–429 (2009). 152. De Smet, K. A. L., Kempsell, K. E., Gallagher, A., Duncan, K. & Young, D. B. Alteration of a single amino acid residue reverses fosfomycin resistance of recombinant MurA from Mycobacterium tuberculosis. Microbiology 145 ( Pt 11), 3177–3184 (1999). 153. McCoy, A. J., Sandlin, R. C. & Maurelli, A. T. In vitro and in vivo functional activity of Chlamydia MurA, a UDP-N-acetylglucosamine enolpyruvyl transferase involved in peptidoglycan synthesis and fosfomycin resistance. J. Bacteriol. 185, 1218–1228 (2003). 154. Gil-Marqués, M. L. et al. Peptidoglycan recycling contributes to intrinsic resistance to fosfomycin in Acinetobacter baumannii. J. Antimicrob. Chemother. 73, 2960–2968 (2018). 155. Borisova, M., Gisin, J. & Mayer, C. Blocking peptidoglycan recycling in Pseudomonas aeruginosa attenuates intrinsic resistance to fosfomycin. Microb. Drug Resist. 20, 231–237 (2014). 156. Gisin, J., Schneider, A., Nägele, B., Borisova, M. & Mayer, C. A cell wall recycling shortcut that bypasses peptidoglycan de novo biosynthesis. Nat. Chem. Biol. 9, 491–493 (2013). 157. Scortti, M. et al. Coexpression of virulence and fosfomycin susceptibility in Listeria: molecular basis of an antimicrobial in vitro-in vivo paradox. Nat. Med. 12, 515–517 (2006). 158. Li, X. et al. Abrp, a new gene, confers reduced susceptibility to tetracycline, glycylcine, chloramphenicol and fosfomycin classes in Acinetobacter baumannii. Eur. J. Clin. Microbiol. Infect. Dis. 35, 1371–1375 (2016). 159. Venkateswaran, P. S. & Wu, H. C. Isolation and characterization of a phosphonomycin-resistant mutant of Escherichia coli K-12. J. Bacteriol. 110, 935–944 (1972). 160. Ohkoshi, Y. et al. Mechanism of reduced susceptibility to fosfomycin in Escherichia coli clinical isolates. Biomed Res. Int. 2017, (2017). 161. Couce, A. et al. Genomewide overexpression screen for fosfomycin resistance in Escherichia coli: MurA confers clinical resistance at low fitness cost. Antimicrob. Agents Chemother. 56, 2767– 2769 (2012). 162. Lee, Y. C., Chen, P. Y., Wang, J. T. & Chang, S. C. Prevalence of fosfomycin resistance and gene