Title of Thesis

Regulatory Mechanisms Involved in the Pathoadaptation of Extraintestinal Pathogenic Escherichia coli

Nikola Zlatkov

The Laboratory for Molecular Infection Medicine Sweden /Department of Molecular Biology Umeå 2019 This work is protected by the Swedish Copyright Legislation (Act 1960:729) Dissertation for PhD ISBN: 978-91-7855-002-9 ISSN: 0346-6612 New Series Number 2008 Information about cover design / cover photo / composition Electronic version available at: http://umu.diva-portal.org/ Printed by: UmU Print Service Umeå, Sweden 2019

To the better child: My dear sister, keep on fighting for the survival of the goodness in this world!

“Every part is disposed to unite with the whole, that it may thereby escape from its own incompleteness.” Leonardo Da Vinci

Table of Contents:

Abstract ...... 3 Papers included in the thesis ...... 5 Abbreviations ...... 6 Enkel sammanfattning på svenska ...... 8 I. Prologue ...... 10 1. Introduction ...... 10 1.1. Escherichia coli (Migula 1895) Castellani and Chalmers 1919 (Bacillus coli Migula 1895) ...... 10 1.2. Main characteristics ...... 11 1.2.1. Basic cellular morphology...... 12 1.2.2. Nutritional and metabolic characteristics...... 14 1.2.3. Basic genetics...... 19 1.3. Commensal and pathogenic Escherichia coli ...... 20 2. Extraintestinal Pathogenic Escherichia coli (ExPEC) ...... 24 2.1. General characteristics...... 24 2.2. Virulence factors ...... 25 2.2.1. K antigens ...... 25 2.2.2. Cytotoxins...... 25 2.2.3. Invasins...... 27 2.2.4. Siderophores ...... 27 2.2.5. Virulence-associated adhesins...... 28 3. Pathoadaptation ...... 33 3.1. The Eco-Evo dynamics of ExPEC ...... 33 3.2. The genetic plasticity of ExPEC ...... 35 3.2.1. Pathogenicity islands ...... 35 3.2.2. Pathoadaptive mutations ...... 36 3.3. Pathoadaptation and metabolism ...... 39 3.4. Pathoadaptation and developmental programs ...... 39 3.4.1. Signaling networks that regulates sessility and motility in ExPEC 39 3.4.2. Formation of Intracellular Bacterial Communities (IBCs) by UPEC bacteria...... 40 II. Thesis Scope ...... 42 III. Results and Discussion ...... 44 1. Absence of global stress regulation in Escherichia coli promotes pathoadaptation and novel c-di-GMP-dependent metabolic capability ...... 44

2. Elevated recombinant clyA gene expression in the uropathogenic Escherichia coli strain 536, a clue to explain pathoadaptive mutations in a subset of extraintestinal E. coli strains ...... 46 3. A c-di-GMP-mediated morphotypic pathoadaptability of neonatal meningitis- causing Escherichia coli ...... 48 4. Summary ...... 50 5. Epilogue ...... 52 V. Acknowledgements ...... 53 VI. Bibliography ...... 1

Abstract

Establishment of commensal bacteria within a new niche of their host usually promotes the transition from commensalism to pathogenicity. Extraintestinal Pathogenic Escherichia coli (ExPEC) represent different pathovars with biphasic lifestyle – they can reside in the gut as commensals or they can escape and cause diseases elsewhere in the human body. Depending on the disease they are linked to, ExPEC can be divided into Uropathogenic E. coli (UPEC), Newborn Meningitis-causing E. coli (NMEC) and Sepsis-associated E. coli (SEPEC).

Pathoadaptive mutations linked to c-di-GMP signaling were investigated in the NMEC strain IHE3034 which lacks the main global stress regulator RpoS. We investigated the role of ycgG2 in the lifestyle of NMEC. Deletion of ycgG2, shown by us to encode an YcgG allozyme with c-di-GMP phosphodiesterase (PDE) activity, and the restored RpoS led to a decrease in the S-fimbriae, otherwise robustly produced in artificial urine, hinting that the urinary tract could serve as a habitat for NMEC. We showed that NMEC were capable of aerobic citrate utilization in the presence of a co-substrate - a property that normally does not exist in E. coli. Our data hint that this metabolic upgrade is enhanced by the lack, or reduced activity, of c-di-GMP PDEs. We also found that citrate utilization is a property of ExPEC, since we reconstituted it in E. coli UTI89 (a cystitis isolate) via inactivation of its RpoS, and since a set of pyelonephritis E. coli isolates use citrate aerobically in the presence of glucose. The main reason for this metabolic capability is the absence of RpoS which leads to the production of the citrate transporter CitT. Furthermore, we highlighted the deletion of the fec operon (required for the ferric citrate uptake) in a large group of different ExPEC strains and we showed that NMEC can use CitT for in vitro ferric citrate uptake dependent on YcgG2 as an alternative system.

Another pathoadaptive mutation which influences the fitness of ExPEC is the clyA (cytolysin A) gene inactivation, resulting from different deletions in different ExPEC genomes. When we restored the clyA+ locus, the UPEC strain 536 displayed increased susceptibility to antimicrobial peptides and urea. We also showed that the ClyA expression in 536 was increased by the presence of the DNA-binding regulator SfaX and another stand-alone PDE similar to YcgG2, called SfaY. The results were further confirmed by ClyA downregulation in NMEC deficient in SfaY and SfaX.

We also studied the role of sfaY - a gene coding for another stand-alone c-di-GMP PDE. The expression of sfaY is under the regulation of the main promoter of the horizontally acquired sfa gene cluster. The latter is responsible for the regulation and assembly of the virulence-associated S-fimbriae, via which ExPEC bacteria

bind to sialylated receptors. We found that NMEC are competent for filamentation because of a c-di-GMP-dependent program under the control of a phase-variation event which selectively turns ‘ON’ the sfa promoter in a subpopulation of bacteria. When SfaY is present, c-di-GMP levels are reduced, thus inducing the SOS stress response via the canonical LexA-RecA pathway. The signaling resulted in an internal differentiation of the bacterial population into a subpopulation exhibiting a filamentous morphotype (bacteria with induced SOS stress response) and a subpopulation of small motile and non-motile bacteria. Hence, this molecular program could serve as a clue to explain the formation of the intracellular bacterial communities observed during urinary tract infection by UPEC.

Papers included in the thesis

Zlatkov, N, Uhlin, BE. Absence of Global Stress Regulation in Escherichia coli Promotes Pathoadaptation and Novel c-di-GMP-dependent Metabolic Capability /Submitted in Scientific Reports/

Enow, COA, Oscarsson J, Zlatkov, N, Westermark, M, Duperthuy, M, Wai, SN, Uhlin, BE. Elevated Recombinant clyA gene expression in the Uropathogenic Escherichia coli strain 536, a Clue to Explain Pathoadaptive Mutations in a Subset of Extraintestinal E. coli Strains. BMC Microbiol. 2014, 14:216

Zlatkov, N, Uhlin, BE. C-di-GMP-mediated Morphotypic Pathoadaptability of Neonatal Meningitis Escherichia coli /Manuscript/

Not included in the thesis:

Dimova, L, Zlatkov, N, Verkade, HJ, Uhlin, BE, Tietge, UJF. High-Cholesterol Diet Does Not Alter Gut Microbiota Composition in Mice. Nutr Metab (Lond). 2017,14:15

Abbreviations

AIEC – Adherent-Invasive E. coli

APEC – Avian Pathogenic E. coli

AUM – artificial urine medium cAMP – 3’, 5’- cyclic adenosine monophosphate

BBB – blood brain barrier

BMECs – brain microvascular endothelial cells

Byr – billion years c-di-GMP – bis-(3’, 5’)-cyclic dimeric guanosine monophosphate

CNS – central nervous system

CRP – cAMP receptor protein

CUP – chaperon/usher pathway

DAEC – Diffusely Adherent E. coli

EAEC – Enteroaggregative E. coli

EIEC – Enteroinvasive E. coli

EMP – Embden–Meyerhof–Parnas (pathway)

EPEC – Enteropathogenic E. coli

EPS – extracellular polysaccharide

ETEC – Enterotoxigenic E. coli

ExPEC – Extraintestinal Pathogenic E. coli

GIT – gastrointestinal tract

HGT – horizontal gene transfer

IBCs – intracellular bacterial communities

IPEC – Intestinal Pathogenic E. coli kbp – kilobase pair kDa – kilodaltons

LB – lysogenic broth

LPS – lipopolysaccharides

Lrp – leucine-responsive regulatory protein

Mbp – megabase pair

Myr – million years

OM – outer membrane

OMP – outer membrane protein

NMEC – Neonatal Meningitis-causing E. coli

PAI – Pathogenicity Island

PDE – Phosphodiesterase

SEM – Scanning Electron Microscopy

STEC – Shiga-toxin producing E. coli tRNA – transfer RNA

UPEC – Uropathogenic E. coli

TCA – Tricarboxylic acid (cycle)

UTIs – Urinary Tract Infections

Enkel sammanfattning på svenska

Om kommensala bakterier etablerar sig i en ny omgivning i sin värdorganism kan det ofta leda till ett skifte från kommensalism till patogenicitet. Extraintestinala patogena Escherichia coli (ExPEC) representerar olika patogena varianter med alternativ livsstil - de förekommer i tarmarna som kommensaler hos människan och när de har tagit sig därifrån kan de orsaka sjukdomstillstånd på andra ställen i kroppen. Beroende på vilken sjukdom de är förknippade med kan ExPEC delas in i uropatogena E. coli (UPEC), neonatal menigit-orsakande E. coli (NMEC) och sepsis-associerade E. coli (SEPEC). I detta avhandlingsarbete har patoadaptiva mutationer med koppling till c-di-GMP signalering studerats hos NMEC bakterier som saknar RpoS - en central regulator av stressrespons hos E. coli.

Vi undersökte vilken roll gene ycgG2, som vi visat kodar för ett YcgG-allozym med c-di-GMP fosfodiesteras (PDE) aktivitet, spelar i NMECs livsstil. I varianter med återställt fungerande RpoS kombinerat med en deletion av ycgG2 fann vi minskat uttryck av S-fimbrier, som annars normalt produceras rikligt vid bakteriernas tillväxt i artificiellt medium med sammansättning som motsvarar urin. Resultaten tyder på att urinvägarna kan fungera som en tillväxtmiljö för NMEC. Vi fann att NMEC under aeroba förhållanden, i närvaro av co-substrat, kunde tillgodogöra sig citrat - en egenskap som E. coli normalt inte har. Våra data tyder på att denna metaboliska uppgradering förstärks när c-di-GMP PDE aktiviteten saknas. Vi fann också att citratutnyttjande är en egenskap hos andra ExPEC, eftersom vi kunde återskapa det i E. coli UTI89 (ett cystit isolat) genom att inaktivera rpoS genen och att ett antal pyelonefrit isolat av E. coli också aerobt kan använda citrat i närvaro av glukos. Den huvudsakliga anledningen till denna metaboliska förmåga är att RpoS saknas, vilken leder till produktion av citrat- transportören CitT. Slutligen rapporterade vi att det skett deletion av fec operonet (krävs för järncitrat upptag) i en stor grupp av olika ExPEC stammar och vi visade att NMEC, som ett alternativt system beroende av YcgG2, kan använda CitT för järncitrat-upptag in vitro.

En annan patoadaptiv mutation som kan påverka ExPECs fitness är inaktivering av clyA-genen (cytolysin A) som har skett i genomet hos olika ExPEC. När vi återskapade clyA+ locuset i UPEC-stammen 536 påvisades en ökad känslighet för antimikrobiella peptider och urea. Vi visade också att ClyA-uttrycket i 536 ökade i närvaro av den DNA-bindande regulatorn SfaX och av SfaY, ett annat PDE enzym som liknar YcgG2. Detta resultat verifierades av resultat som visade nedreglering av ClyA i NMEC som saknar SfaY och SfaX.

Vi har också studerat rollen av sfaY genen närmare. Uttrycket av sfaY regleras av promotorn till det horisontellt förvärvade sfa-genklustret, som ansvarar för

produktion av virulensassocierade S-fimbrier med vilka ExPEC-bakterierna binder till sialylerade receptorer. Vi har påvisat att NMEC växer i form av långa filament på grund av ett c-di-GMP-beroende program, under kontroll av fas- variation som selektivt slår ‘PÅ’ sfa promotorn in a en bakteriesubpopulation. När SfaY är närvarande minskas c-di-GMP-nivån vilket inducerar SOS- stressrespons via LexA-RecA-systemet. Denna signal resulterade i en differentiering av bakteriepopulationen så att det uppstod en subpopulation med filamentös morfologi (bakterierna med inducerad SOS-stressrespons) och en med små rörliga eller icke-rörliga bakterier. Alltså, skulle detta molekylära program kunna fungera som en ledtråd för att förklara bildandet av de intracellulära bakterieansamlingar som påvisats vid urinvägsinfektioner av UPEC.

I. Prologue

1. Introduction “On any possible, reasonable or fair criterion, bacteria are – and always have been – the dominant form of life on earth” 1. And that is how Gould’s piercing comment introduces us to the importance of bacteria to nature and, in particular, to humans 1.

Bacteria (or prokaryotes) are microscopic, predominantly unicellular forms of life. On the origin of life and their evolution, the prokaryotes were the first organisms to appear 3.4 Byr ago, maintaining life and evolution then after on their own for almost 2 Byr until they gave rise to the eukaryotes, with which they have been co-existing ever since 2. Due to their genetic plasticity, bacteria have far and away the most versatile and strong metabolic machinery which provides them with almost limitless capacity for adaptation to life in virtually all ecosystems on our planet. Once being inhabitants, bacteria have the power to change and develop a certain ecosystem in various ways.

Homo sapiens, for example, is a multitude of such ecosystems, colonized by an unfathomable variety of bacteria. Some of them, called commensals, can be advantageous to humans either by producing useful metabolites or fighting off harmful infectious agents. Others, called pathogens, can damage their host which results in the development of a particular disease. Being a pathogen or commensal is only of value to the colonized organism, and transition to a new niche of its host can often provide a commensal bacterium with pathogenic potential.

1.1. Escherichia coli (Migula 1895) Castellani and Chalmers 1919 (Bacillus coli Migula 1895) In 1885, Dr. Theodor Escherich brought to existence one bacterial species which he named “Bacterium coli commune” (the common colon slender rod) 3. The species was isolated from the gut microbiota of neonates and it represented by Gram (-) small rod-shaped bacteria able to clot milk 3. Extensive work on the characterization of Escherich’s isolate had been performed by numerous scientists and in 1919 Castellani and Chalmers reformed the definition of this species, based on some additional features 4. They combined the facts that in addition to milk clotting, Escherich’s bacteria grow well in ordinary laboratory media, do not form endospores, are often facultative anaerobes which can completely ferment glucose and lactose with the production of acid and gas 4. Based on all the new evidence they renamed Escherich’s bacterium Escherichia

coli (Esch.er.ìchi.a. - M.L. fem. n. Escherichia named after Theodor Escherich; còli - Gr. n. colon large intestine, colon; M.L. gen. n. coli of the colon.) 3,4.

Little did they all know that this bacterium would play a fundamental role in the development of science and humanity, ever since it was discovered, and such would its role be for the development of this thesis.

1.2. Main characteristics The identity of each species is shaped by its basic cellular organization, metabolism and genetics. All together, these determinants place Escherichia coli in:

Domain Eubacteria Phylum Proteobacteria Class Gammaproteobacteria Family Enterobacteriaceae Genus Escherichia

According to Huxley, the significance of adaptation can only be understood in relation to the total biology of the species 5, that is why this subchapter of the thesis aims to present the general biology of E. coli, in such a way, that any positively selected deviation from it could be appreciated as an adaptive trait for certain conditions at a particular time.

In short, the species is represented by small (0.5-1X1-2µm), rod-shaped, Gram- negative, facultative anaerobic, chemoorganotrophic bacteria with a respiratory and a fermentative type of metabolism which are mostly motile by means of peritrichous flagella 6. E. coli strains can either be commensal inhabitants of the gastrointestinal tract (GIT) of homothermic animals and reptiles, or they can be pathogens causing infections inside and outside of the GIT 6,7. E. coli cells are neutrophiles (pH 5.0 to 9.0) with temperature growth range from 15 to 45⁰C with 37⁰C being the optimal temperature for most of the E. coli strains, which defines them as mesophiles 6.

Figure 1. The major antigens of E. coli, exemplified with the bacteria of the neuroinvasive strain IHE3034. The K antigen is depicted by attachment of K1-specific phages.

1.2.1. Basic cellular morphology. The overall cellular organization of E. coli follows the typical Gram-negative pattern, i.e., the cells are isolated from the environment with two membranes – an inner (cytoplasmic) membrane and an outer membrane (OM), separated by priplasmic space. The cell wall is a monomolecular murein sacculus, composed of β-1→4 linearly linked N-acetylglucosamine and N-acetylmuramic acid monomers, cross-linked by L-Ala-D-Glu-L-meso-DAP-D-Ala peptides, located on the outside surface of the inner membrane 6. The outer membrane contains different outer membrane proteins (OMPs), most of them functioning as porins allowing passive diffusion of different metabolites, ions and other compounds. Linked to the extracellular surface of the OM is the lipopolysaccharide (LPS) molecular complex. It is composed of a hydrophobic anchor (lipid A) and a core domain, i.e., polysaccharide that connects lipid A to a glycan polymer - the third element in the structure. The glycan polymer is exposed directly to the extracellular milieu and it can vary in number and composition of its sugar residues. This variation of modified simple sugar monomers gives the bacterial LPS strong antigenic properties referred to as the O antigen (Fig. 1).

Most strains of E. coli are motile by means of peritrichous flagella. Each flagellum is composed of a basal body, a hook and a filament consisting of a single protein called flagellin (encoded by the fliC gene). It can be up to 20µm long and 20nm in diameter. Flagellin is also known as the H antigen (Fig. 1) since it can also vary antigenically due to changes in the amino acid composition of its central region or due to posttranscriptional modifications (such as acetylation) 8.

In addition to the flagellar machinery, E coli cells possess extracellular attachment organelles, called fimbriae (or pili), via which bacteria adhere to different surfaces. E. coli bacteria can express more than 30 fimbriae 6,9. Each

organelle is composed of linearly polymerized fimbrial subunits protruding outside of the bacterial cell, ending with a lectin-like adhesion molecule, i.e., adhesion, on its tip via which bacteria can selectively recognize and adhere to different surfaces, including host tissues. One of the most abundant types of fimbriae belongs to the group of the chaperone/usher pathway (CUP) family to which the virulence-associated (e.g., P- and S-fimbriae) and the common type-1 fimbriae belong (Fig. 2). Once the fimbrial subunits are made, they are secreted to the periplasmic space, bound to a chaperone that preserves them partially unfolded guiding them to the usher – a big proteinaceous platform where the assembly of the fimbrial subunits happens in a step-wise linear manner (Fig. 2) 10-12. Once guided to the usher, the chaperone subsequently leaves the fimbrial subunit, exposing the N-terminal sequence of the guided structural monomer 10- 12. Thus, the sequence becomes a donor for the C-terminal incomplete immunoglobulin fold of the last structural unit of the growing filament (Fig. 2) 10-12. The most common type of fimbriae are the type-1 fimbriae (encoded by the fim gene cluster) produced by 80% of all E. coli strains via which bacteria can selectively adhere to mannosylated glucoconjugates 13. Even though the type-1 fimbriae are not considered virulence factors, since they are also produced by the commensal bacteria, they can assist in the colonization by the uropathogenic, neuroinvasive, and adherent-invasive E. coli strains 14-17.

Figure 2. A summary of the chaperone-usher pathway fimbrial family. The main steps of the regulation of fimbrial

expression and fimbrial biogenesis are depicted and exemplified with the assembly and regulation of the S-fimbriae. The table (below the model) combines the most abundant and well-studied members of the family of the fimbrial chaperone-usher family with their basic characteristics, i.e., their genetic determinants, chromosome location and the binding specificity of their adhesins 18,19.

E. coli cells are able to produce two different types of extracellular polysaccharides (EPSs). The first type is represented by EPS loosely associated to the cell surface or directly released into the extracellular environment under the form of a slime layer 6,20. One of the most common extracellular EPSs layer is produced by accumulation of colanic acid, known as M antigen 20,21. The extracellular colanic acid is not antigenic and is produced under certain conditions by most members of Eneterobacteriaceae, including E. coli 22. The M antigen is a highly viscous layer involved in biofilm stabilization, prevention from dehydration and it can also serve as a defense mechanism against bacteriophages since it covers the phage receptors located in the OM 21,23. The second type of EPSs is represented by heteropolysaccharides covalently anchored to the LPS, forming a structure called capsule 6,20. Directly exposed to the extracellular environment, bacterial capsules are evolutionary designed to provide adhesion, prevention from desiccation, and resistance to harmful agents such as bacteriophages, specific and non-specific host immunity 21,24,25. The heteropolysaccharide nature provides the capsule of E. coli with antigenic properties, so the capsule is known as the K antigen (from German word “Kapsel”) (Fig. 1).

1.2.2. Nutritional and metabolic characteristics. In order to utilize a substrate, bacteria need to have the right import systems and the enzymes that participate in the biochemical pathways for its utilization. Furthermore, the expression of the import systems and the enzymes depend on the presence or the lack of regulatory factors 26. E. coli strains can feed on different carbohydrates, polyols, carboxylates, two-carbon components and fatty acids, and some amino acids (see Table 1 for all the compounds used by E. coli). As a facultative anaerobe with a respiratory and a fermentative type of metabolism, E. coli can thrive in different environments, using a big variety of carbon and energy sources which in nature never appear individually. Mitchell et al. showed that the environmental changes can be predicted by E. coli and once predicted, they can be inherited, i.e., in short, with the assumption that early in the process of the gut colonization, E. coli bacteria first utilize lactose, they showed that the lactose induces the lac operon, but also pre-induces the maltose operon, preparing the bacteria for utilization of maltose which comes later in the life of the host 27. In the presence of different substrates, the bacteria have to decide which substrate will be most beneficial for utilization – a choice made via dedicated signaling programs. One of them involves the utilization of glucose over other non-glucose substrates, a process known as catabolite repression - a

dedicated signaling program that is highly dependent on a second messenger, called 3’, 5’- cyclic adenosine monophosphate (cAMP) 28. E. coli imports glucose through the glucose transporter which is a multiprotein complex that couples the glucose transport with its immediate phosphorylation 29. This transport system is composed of the phosphohistidine carrier protein (HPr), the Enzyme I (EI) and the Enzyme II (EIIA and EIIB) components 29. In presence of glucose, the HPr and EI proteins transfer a phosphoryl group from a phosphoenolpyruvate molecule to EIIA and EIIB protein complexes which eventually transmit it to a glucose molecule 29. If there is no glucose in the medium, the phosphorylated central domain of EIIA binds and activates adenylate cyclase which produces cAMP 30. When present in the cells, cAMP binds and switches on a transcriptional activator called CRP (cAMP receptor protein) 30. Once activated, the CRP-cAMP complex triggers the expression of many catabolic genes which are normally under the regulation of weak promoters 30. In E. coli, CRP often acts together with Lrp (leucine-responsive regulatory protein) - a transcriptional regulator of around 10% of the E. coli genes and it is allosterically controlled by the presence of leucine 31,32. When leucine is present in the environment, Lrp shuts down the expression of the genes coding for the biosynthesis of leucine and other branched- chain amino acids such as valine and isoleucine 31. CRP and Lrp exert not only a regulatory effect on the bacterial metabolism, but they also serve as metabolic sensors that the pathogenic E. coli strains use in the regulation of the virulence gene expression, especially in the regulation of the virulence-associated fimbriae 19,33.

Table 1. Compounds used by E. coli as a sole carbon and energy source 6,26 Hexoses Pentoses/Trioses Oligosaccharides Amino sugars Uronic acids D-Glucose L-Arabinose Maltose/Maltodextrin N-Acetyl-D- Hexourinides Glucosamine D-Fructose D-Xylose Melibiose D-Glucosamine Hexouronates D-Galactose D-Ribose β-Galactosides Hexonates (lactose) D-Mannose L-Fucose Trehalose L-Sorbose1 L-Rhamnose2 Sucrose3 Phosphorylated Mono- and Polyols Amino acids and 2-Carbon (P) sugars dicarboxylic oligopeptides components acids and fatty acids Glucose-6-P Lactate Mannitol L-Asparagine5 Acetate Glucose-1-P Succinate Sorbitol L-Aspartate5 Fatty acids Fructose-6-P Fumarate Dulcitol3 L-Proline Ethanol5 Fructose-1-P L-Malate Arabitol L-Tryptophan Glucosamine-6-P Ribitol4 D-Alanine Xylilol L-Alanine/L- Leucine Glycerol Protein hydrolysates Oligopeptides

Notes: 1. 9 out of 40 strains grow on sorbose, E. coli K-12 cannot. 2. ¾ of tested E. coli strains grow. 3. Half of the strains tested grow on dulcitol. 4. 10% of tested strains grow, but not E. coli K-12.

5. Use the substrate anaerobically.

E. coli cells can synthesize all of their biomolecules from a single C source and inorganic donors for P and N and for cofactor synthesis. Defined as a collection of all biochemical reactions required to sustain life, metabolism is composed of (i) reactions required for the synthesis of precursor metabolites and energy; (ii) reactions used for the generation of building block compounds and co-enzymes; and (iii) reactions required for the synthesis and processing of biopolymers 26. One of the key substrates that E. coli bacteria use for generation of ATP and metabolite precursors is glucose. It is normally metabolized through the Embden–Meyerhof–Parnas (EMP) pathway with the involvement of the methylglyoxal bypass, if bacteria face P-limited environment. To enrich its cells with various metabolites, E. coli can also metabolize glucose via the hexose monophosphate pathway. Another pathway that exists in E. coli is the Entner– Doudoroff pathway through which bacteria metabolize gluconate, produced by the mucus of the mammalian intestine 26,34. In aerobic conditions, the full degradation of glucose is achieved through the tricarboxylic acid (TCA) cycle also generating the energy equivalents used for the ATP synthesis. In bacteria the TCA is inducible by the presence of oxygen 35. If bacteria are grown on C sources, such as acetate or fatty acids, which are not metabolized via pyruvate or phosphoenol pyruvate intermediates, bacteria cannot supplement the TCA cycle 24,26. Hence, to be able to utilize substrates with two C atoms, bacteria activate the glyoxalate cycle 24,26. Even though oxygen is the preferred terminal electron acceptor, E. coli can also use nitrate, fumarate and dimethyl sulfoxide as terminal acceptors 6,26,35. In the absence of oxygen, E. coli bacteria grow without using any exogenous electron acceptors, i.e., they are able to ferment glucose and other compounds into pyruvate which is further converted to lactic, acetic and formic acids 6,24,26. A recent study by Antonovsky et al. showed that E. coli can become hemiauthotroph by evolving to synthesize its biomass directly from CO2 when a non-native Calvin- Benson cycle was introduced and phosphoglycerate mutase genes (gmpA and gmpB) were deleted 36.

1.2.2.1. Inability of E. coli for citrate utilization. Although E. coli can use a large number of substrates (listed in Table 1), there are also some that cannot be metabolized for different reasons (see Table 2). It is worth noting that among the unusable substrates, there are some for which E. coli strains have the genetic determinants required for the substrates’ degradation (see Table 2, the substrates highlighted in blue) but either the transporter(s) or the enzymes are expressed in different conditions or not expressed at all. One example is the inability of E. coli to use citrate – an important part of the diagnostics of this species. In 1923, Koser pointed out that E. coli bacteria cannot grow on citrate 37, later Lara and Stokes showed that

lyophilized bacteria are actually able to oxidase citrate but not the living ones, suggestion that living bacteria do not express the citrate transporter 38. Finally, Lutgens and Gottshalk fully investigated why E. coli cannot grow on citrate. Based on their studies, the bacteria cannot utilize citrate as a sole carbon and energy source because (i) the citrate transporter is expressed anaerobically; (ii) the bacteria do not have oxaloacetate decarboxylase; (iii) in anaerobic conditions the TCA cycle is repressed and even taken in, the citrate cannot be metabolized 39. They also proposed a scenario which provides a possibility for E. coli to use citrate in the presence of a co-substrate (such as glucose, glycerol, lactose), needed for the production of reduced equivalents to activate malate dehydrogenase and fumarate reductase activities to complete the fermentation of citrate to succinate (Fig. 3) 39.

1.2.2.2. Iron metabolism. Iron is the only metal known to be generally growth limiting in different environments. Most bacteria need iron for various vital processes such as TCA cycle, respiration, nucleotide synthesis, redox reactions, oxygen transport, etc. 40. Iron is available as ferric (Fe3+) and ferric (Fe2+) cations depending on the redox potential of the environment and the uptake system developed by the bacteria reflect the iron source of this environment at a certain time 40.

In a reducing anaerobic/microanaerophilic environment the predominant form of iron is Fe(II). Being inhabitants of such an environment, E. coli bacteria developed mechanism for the Fe2+ uptake from the periplasm since ferrous cations diffuse freely through porins 40-42. The ferrous iron uptake system is encoded by the feoABC operon 40,42. FeoB is the major component of the system and serves as permease through which the Fe2+ ions are transported into the cell 41. FeoA is a small cytoplasmic protein with a possible role in protein-protein interactions and FeoC might play a role as a transcriptional regulator of the operon 41.

In the GIT, iron (III) is less abundant since ferrous cations prevail 40,41. If present, it forms soluble complexes with various chelating compounds 40,43. The commensal E. coli developed different iron acquisition systems for Fe(III) uptake, depending on the formed complex. The general mechanism for Fe(III) import requires active transport across each membrane of the cell envelope 40. The Fe(III) – complex is recognized by a specific receptor coupled to the TonBExbBExbD system which provides the energy for the active import of the iron through the OM 40. In the periplasm, the free iron (III) is bound to a periplasmic protein which further delivers the iron to a specific ABC transporter in the plasma membrane 40,43. Commensal E. coli express the following receptors: FhuE for Fe3+- coprogen, Fe3+- rhodotorulic acid, Fe3+- ferroxamine; FhuA for Fe3+- ferrichrome; FecA for Fe3+-citrate; FepA for Fe3+-enterobactin and Fe3+-

dihydroxybenzyl serine; Fui for Fe3+-dihydroxybenxoic acid and Fe3+- dihydroxybenzoyl serine 40,43. Besides the receptors for iron (III) uptake, E. coli produce and secrete small chelators with high affinity to Fe3+, called siderophores. There are 5 groups of siderophores based on their structures: catecholates, phenolates, hydroxamates, α-hydroxy-carboxylates and mixed-type siderophores 40. The commensal E. coli K-12 produce only one type siderophore, i.e., enterobactin, from the group of the catecholates 40,43. It has to be mentioned that in addition to enterobactin, the group of extraintestinal pathogenic E. coli produce more siderophores which are virulence factors described in Subchapter 2.2.4. Generally, once Fe(III) ions are transported in the cytoplasm, it is immediately reduced by assimilatory flavin-containing ferric reductases 40. Too much iron can be toxic for the bacteria, since Fe(II) has strong catalytic properties of generating reactive oxygen species that have very strong biological damage potential 40,44. Therefore, the excess of iron is stored in iron- binding proteins such as ferritin A, bacterioferritin or Dps 40,45.

Table 2. Compounds not used by E. coli as a sole carbon and energy source 6,26,39 Substrates Reasons: Conditional Utilization: L-Sorbose No transporter and enzymes. It can be metabolized through the L-fucose D-Arabinose pathway, but the pathway is not induced by D- arabinose. Raffinose No transporter and enzymes. Inositol No transporter and enzymes. Butanol No enzymes for conversion to butyryl-CoA. Cells cannot metabolize butyryl-CoA. L-Glutamate No transport system. L-Arginine Do not utilize it, but they have the potential to metabolize it to succinate. L-Histidine Lack histidinase. L-Cysteine Toxic for E. coli in high concentration. L-Tyrosine Do not have the enzymes. L-Serine Do not utilize it, but they have the potential to It is metabolized when it is present metabolize it. together with glycine, L-leucine, L-isoleucine, and L-valine. Glycine Do not utilize it. L-Threonine Do not utilize it, but they have the potential to metabolize it. Citrate Do not code for oxaloacetate dehydrogenase. The It is metabolized anaerobically CitT transporter is expressed anaerobically when when a co-substrate /such as the TCA cycle is repressed. glucose/ is provided to supply the fermentation process with reduced equivalents. Notes: The substrates that E. coli can potentially metabolize are highlighted in blue.

Figure 3. Proposed model for citrate utilization by E. coli under anaerobic conditions by Lutgens and Gottschalk. 39

1.2.3. Basic genetics. Stress presents itself in unlimited ways. The coding potential of E. coli is limited. Yet, this limited coding potential of E. coli is characterized by high genetic plasticity 46-48. This subchapter deals with the main aspects of E. coli genetics important for this thesis.

In 1997, Blattner et al. showed that the complete sequence of the genome of E. coli K-12 strain MG1655 is elegantly organized into a 4,6-Mbp single chromosome 49. It codes for 4,500 genes (of which, 303 essential), i.e., 4,300 ORFs, 7 rRNA operons, 86 tRNA and 74 pseudogenes 49-51. 755 out of 4,300 E. coli ORFs were acquired via horizontal gene transfer (HGT) 47,52. The remnants of Sf6 and P4 prophages were inserted at tRNAargW and tRNAleuX, respectively, suggesting that the tRNA loci are the preferred site for integration of mobile genetic elements (such as some lysogenic coliphages) 47,49.

Today, the sequences of almost 700 genomes of different pathogenic and non- pathogenic E. coli strains are available out of 14,820 sequences for genome assembly or annotation report (www.ncbi.nlm.nih.gov/genome/genomes/). Based on the available sequences, the chromosome of E. coli can vary from 4,6 to 5,5 Mbp, ruling out that strain MG1655 contains the smallest chromosome of all E. coli strains 18,53. Studies on E. coli genomes of different sizes by Ochman et al. suggested that the acquisition of the novel genetic material is followed by the loss of needless one 53. For example, E. coli ECOR 37 (a marmoset isolate) has a 5,4- Mbp chromosome that lacks 350 kbp of the genome of MG1655 but acquired a 1.2

Mbp unique coding sequence 53. The comparative genomics of pathogenic and commensal strains revealed that the core genome of E. coli is composed of around 2,000 genes, while the pan-genome contains 10,000 genes 18,48. Touchon et al. showed that the core genome codes for the species characteristics including the anabolism of amino acids, nucleotides, co-factors and proteins; metabolism of DNA, fatty acids, and phospholipids; transcription, translation and protein processing 48. The pan-genome contains mostly genes coding for regulators, cell wall modifications, different biological processes and mobile genetic elements 48.

In addition to the dynamics of the E. coli chromosome, different strains can harbor various extrachromosomal circular DNA elements that can independently replicate in the cytoplasm of bacteria, called plasmids. Most of them are conjugative F plasmids (from fertility), belonging to the IncF family 54. E. coli plasmids can bring various traits which are involved in virulence (plasmids from the group of ColV and Vir found in ExPEC), drug resistance (pETEC6 found in ETEC E24377A, pEK499 found in E. coli ST131) and the metabolism of uncommon compounds 6,54,55. Even though that E. coli is mostly presented as a citrate-negative bacterium (cf. Subchapter 1.2.3.), there are some thermosensitive H1 plasmids found in very few E. coli isolates from birds and cattle providing the bacteria with the ability for citrate utilization and antibiotic resistance 54,56,57. The overview in Table 3 summarizes the main groups of intestinal pathogenic E. coli and further highlights the importance of the plasmids for the virulence of E. coli, since a lot of the virulence traits are plasmid- based.

1.3. Commensal and pathogenic Escherichia coli E. coli and Salmonella enterica diverged between 120-160 Myr ago which coincided with the appearance of the first eutherians 47,52,58. At that period E. coli established a commensal relationship with its host – the GIT of the lactose producing, homeothermic animals, i.e., the mammals. As suggested by Tenaillon et al., establishing a partnership with a host provides the bacteria with a relatively constant influx of nutrients, a steady environment, a safety from some insults, and finally the host can be used for dissemination 46. It is supposed that the last common ancestor of the modern E. coli strains emerged between 20 to 40 Myr ago 53.

Today’s E. coli come in many different types. The species is composed of a large number of ecovars that can be commensal, pathogenic, environmental, and laboratory strains. The majority of E. coli strains can be found as commensals in the GIT of mammals (including humans), birds and reptiles, but they can also escape the GIT of their host and survive in different environments such as soil and water 6,7,18. Besides commensals, E. coli can be professional or opportunistic

pathogen, causing colibacillosis with a variety of outcomes in the animals. Based on the disease they are associated with and the virulence factors they have, pathogenic E. coli are usually divided into two big groups. The first group is the Intestinal Pathogenic E. coli (IPEC) whose members cause diarrhoeagenic colibacillosis 6,18. Based on the adhesion/colonization mechanism and the virulence factors, the IPEC representatives are divided into: Enteropathogenic E. coli (EPEC), Enterotoxigenic E. coli (ETEC), Enteroinvasive E. coli (EIEC), Enteroaggregative E. coli (EAEC), Diffusely Adherent E. coli (DAEC), Shiga-toxin producing E. coli (STEC) and Adherent-Invasive E. coli (AIEC) 6,59-66. With regret, the scope of the thesis does not cover the biology of IPEC, however their main characteristics are presented in Table 3 and in Fig. 4A. Extraintestinal Pathogenic E. coli (ExPEC) form the second group of pathogenic E. coli. Its members escape the GIT and cause diseases outside the GIT with various outcomes. The ExPEC colibacillosis may result in urinary tract infections (UTIs), such as cystitis, pyelonephritis, in sepsis, in meningitis, or in other less frequent infections. Subchapter 1.4. is dedicated to a detailed representation of ExPEC. Fig. 4A shows the complex relationships among different pathogenic E. coli groups based on the fact that some groups share the same virulence factors and serotypes (see Table 3 for IPEC and Subchapter 1.4. for ExPEC).

Table 3. Main pathovars that belong to Intestinal Pathogenic E. coli (IPEC)* 6,18,61,62

IPEC Subgroup Host Diseases Properties Comments i) Attachment and effacing (AE) lesions EPEC due to the intimin and intimin receptor (Tir) injected by the Type III secretion system encoded on the LEE locus on PAI. ii) Do not produce Shiga toxin. tEPEC Humans Infantile Presence of BFP diarrhea in (bundle forming pili) (typical EPEC) developing and EAF (EPEC countries adherence factor) encoded on pEAF plasmid. aEPEC Animal Enteritis; i) Lack pEAF and Strains also share reservoir, diarrhea in hence the BFP. virulence (atypical EPEC) humans calves, rabbits ii) Specific virulence properties with included and humans fimbriae such as for STEC such as the AF/R2/2 (adhesive production of factor/rabbit 1 and 2) haemolysins. and Ral (REPEC adherence locus) for rabbits and Paa (Porcine attachment/effacement associated) for pigs.

Host-specificity Ruminants; Traveler’s i) 20 types of CSs of ETEC is due to calves; diarrhea; (surface antigens). the presence of pigs; dogs; watery ii) Heat stable (ST) different humans diarrhea in enterotoxin similar to adhesins, i.e., new born cholera toxin and/or animal ETEC do animals heat labile enterotoxin not produce ETEC (LT). adhesins typical iii) α-Hly (alpha- for the human haemolysin). ETEC strains.

Some of the adhesins and the toxins, including α-Hly are plasmid-encoded.

Self-limited Presence of pINV EIEC cells are Humans; shigellosis- coding for invasins non-motile, EIEC Not known like dysentery (from IpaA to IpaH), mostly lactose animal the Type III secretion negative and reservoir system and IcsA. unable to decarboxylate lysine – typical features for Shigella spp. Persistent i) Presence of plasmid- EAEC cells form Humans diarrhea (up encoded aggregating a pattern similar to 17 days) in adhesion fimbriae to the EAEC humans (AAFI to AAFV). arrangement of ii) Presence of bricks in a wall enteroaggregative on cells in a Hep- heat-stable toxin-1 2 cells monolayer (EAST-1); Pet toxin culture. (plasmid encoded) and Shigella enterotoxin (ShET1). STEC Ruminants Diarrhea to Presence of Shiga Stx toxins are haemolytic toxins (Stx). encoded in the uremic genome of syndrome temperate lambdoid bacteriophage EHEC Ruminants Diarrhea; i) AE lesions due to haemorrhagic presence of LEE PAI (AE-STEC) colitis, (see EPEC). haemolytic ii) Stx1 and/or Stx2 uremic that recognize Gb3 syndrome (globotriaosyl ceramide receptor found on endothelial cells) receptors. Enterotoxaemic Pigs Oedema i) Presence of Stx2e Stx2e recognizes disease, which might not be in Gb4 and the E. coli followed by temperate lambdoid primary target is (F18+ STEC) nervous signs phage. not the and sudden ii) Presence of HlyA; enterocytes. death; iii) Presence of F18 lack of adhesins. Humans do not diarrhea have the receptor for F18 which suggests that the strains cannot be transfer to humans. Humans Diarrhea and i) Daa adhesins from Diffuse and extraintestinal Afa/Dr family. adherence on 1 DAEC animals infections in ii) ExPEC S- and P- cells in cells animals, fimbrial adhesins. culture.

unclear role in iii) Aap dispersin Genetically human common for EAEC. diverse group. AIEC1 Humans Implicated in i) ExPEC Type-1, S-, It is a relatively Crohn’s P- and Mat – fimbriae. young, disease ii) ExPEC capsular heterogeneous development antigens of K1, K5 and group of E. coli K12 type. composed of iii) ExPEC strains that have siderophores. the ability to iv) ExPEC toxins – adhere and Vat, HlyA, CNF1, Cdt replicate within v) ExPEC invasins – macrophages and IbeA epithelial cells without inducing cell death. Note: * The specific pathovars shown in green belong to particular main subgroup of IPEC in blue. 1. The group has virulence factors, typical for ExPEC.

With the discovery of so many and so different E. coli strains, some order was needed. At the end of the 1980’s, Selander et al. and Goullet et al. showed by the use of 38 enzymes for multilocus enzyme electrophoresis that E. coli strains form stable subphylogenetic groups – A, B1, B2, C, D, E and F 26,67,68. This was followed by further studies with the genomes of more representative strains, which confirmed the clonality of the groups and more subphylogenetic groups were added to the tree (Fig. 4B) 48. Groups A and B1 gather mostly commensal strains, while Groups B1, E and S are represented mainly by IPEC strains (Fig. 4B). The ExPEC are generally distributed in Groups B2, D and F (Fig. 4B).

Figure 4. Phenotypic and phylogenetic relationships among different E. coli strains (adapted from 18,48,69).

Fig. 4A Vann diagram representing the complex relationships among different human pathogenic E. coli groups. ExPEC strains include NMEC (for example, strains RS218 and IHE3034) and UPEC (for example, strains 536, J96 and UTI89), but also strains isolated from different extraintestinal infections such as pneumonia, peritonitis, sepsis, cholecystitis and others. The ExPEC strains share virulence factors and there are common serotypes that can cause UTIs and meningitis, for example O18:K1:H7 (common for UTI89 and IHE3034). Some UPEC strains display diffuse adherence to tissue culture cells and share the same adhesins with DAEC, a heterogeneous pathotype linked to diarrhea. DAEC can also be recovered from UTIs. STEC pathovars are also reported to cause UTI and other extraintestinal infections. The STEC pathovar is defined by the presence of a bacteriophage- encoded Shiga toxin. Some STEC contain the LEE pathogenicity island, typical for EPEC. STEC strains carrying the LEE pathogenicity islands are known as EHEC (the most important serotype is O157:H7). The EPEC strains besides the attachment and effacing phenotype, they may produce a bundle-forming pilus and attach to cells in a localized manner and these EPEC are known as typical EPEC, while the EPEC strains that do not produce the bundle-forming pilus are known as atypical EPEC. Some EPEC can be diffusely adherent. EIEC bacteria invade intestinal epithelial cells and multiply in their cytoplasm, and spread from cell to cell. This pathogroup includes genus Shigella which is part of E. coli phylogeny. S. dysenteriae serogroup 1 strains produce Shiga toxins and they can be also included in the groups of EIEC and STEC. EAEC strains are associated with persistent diarrheae and are defined by their brick-in-wall pattern of adherence. In 2011 a large outbreak in Europe was caused by EAEC that acquired the gene encoding the Shiga toxin (strain O104:H4). EAEC are also associated with UTI. ETEC are associated with acute diarrheae and are grouped based on the production of heat- labile and/or heat-stable enterotoxins. Fig. 4B The main phylogenetic groups of E. coli strains. Sub- groups B2, D and F include strains associated with extraintestinal infections while sub-groups S, E and B1 with intestinal diseases. Sub-group A includes predominantly commensal isolates.

2. Extraintestinal Pathogenic Escherichia coli (ExPEC)

2.1. General characteristics. The ExPEC group is represented by one of the most diverse group of strains with three main characteristics – i) they can reside as commensals in the GIT of their hosts without causing any disease; ii) once they escape the GIT of their host, they can colonize other parts of the body, resulting in infection and iii) their virulence genes are encoded on large mobile genetic elements, called pathogenicity islands (PAIs) 16,70,71. Based on the host and the disease they are associated with, the ExPEC group is divided into Uropathogenic E. coli (UPEC), Neonatal Meningitis- causing E. coli (NMEC), and Avian Pathogenic E. coli (APEC). The thesis focuses mainly on some model UPEC and NMEC strains whose biology will be presented in detail.

UPEC bacteria are the most frequent isolates from UTIs in humans 18,63,72. Due to anatomical differences, women are more susceptible to UTIs than men and it is estimated that 40% of women will encounter UTIs in their lives 73. The route of infection starts with a transfer of bacteria (infectious dose of 200 bacteria) from the fecal site to the ureter and once established in the urine (bacteriuria), bacteria can ascend and infect the bladder (cystitis) and depending on the strain, the infection can spread from the bladder to the kidneys (pyelonephritis) 72,74,75.

Further complications can arise if the kidneys are damaged, leading to sepsis and kidney failure 72,74,75.

NMEC together with the Group B streptococci are the leading cause of neonatal meningitis and early onset sepsis 76,77. NMEC meningitis (6.8 per 1000 live births) is a very severe disease that can reach up to 40% mortality, requiring low infectious dose (50-200 bacteria in the blood) and can lead to severe neurological abnormalities after recovery (in 10-15%) such as hydrocephalus, seizures, mental retardation, cerebral palsy, hearing loss and memory impairment 78-81. In line with the severity of new born meningitis, the NMEC early onset sepsis has three times higher mortality compared to the sepsis caused by the Group B Streptococcus 82.

As depicted in Fig. 4A and Fig. 5, NMEC and UPEC strains share a lot of traits which suggest for an on-going evolution among them. Based on the similarities and the differences, one can speculate that ExPEC virulence results from a combinatorial effect of one or several virulence factors or special characteristics.

2.2. Virulence factors

2.2.1. K antigens E. coli bacteria can produce around 100 K antigens (see Subchapter 1.2.1.), and the most frequently isolated UPEC serotypes are K1, K2, K3, K5, K12, K13 and K15 83-85. Their heteropolysaccharides have low molecular weight; at least one of the monomers has acidic properties and the capsules are not produced at temperatures below 30⁰C 83-85. The main virulence factor of NMEC is the K1 capsular antigen (very similar to the major antigen of Neisseria meningitidis group B) encoded by the kps-neu-kps gene cluster 79,83,86-93. The capsule is composed of polysialic acid which blinds the host immune system for the invasive bacteria 25,94. The polysialic pattern of the K1 capsule resembles the glycosylation pattern of the embryonic form of N-CAM (Neural Cell Adhesion Molecule) which creates immune mimicry and the immune system cannot recognize the K1 bacteria as invaders 89,90,95. Even though the K1 serotypes are predominantly neonatal meningitis isolates (80% of the cases), they are also associated with sepsis and UTIs 79,86,96. Together with the different iron acquisition systems (salmochelin, as the main player, produced by iroBCD), the K1 capsule reinforces the survival of the bacteria in the blood 87-89,97.

2.2.2. Cytotoxins. α-Haemolysin (HlyA). Production of toxins, i.e., haemolysins that can lyse erythrocytes is a property of UPEC rather than of NMEC 16,18,98,99. UPEC produce HlyA which can lyse the erythrocytes of all mammals, even fish erythrocytes 100.

HlyA is a 110-kDa exotoxin that belongs to the family of RTX toxins. The determinants for production, activation and secretion of HlyA are located on PAIs organized into hlyCABD operon 101,102. The hly operon is usually linked to the operons of other virulence genes located on the same PAI, as exemplified in Fig. 103,104 9 (see PAI I536 and PAI II536) with the PAIs of UPEC strain 536 . HlyA is synthesized in its inactive form. It is later activated in the cytoplasm by HlyC (a fatty acid acetyl transferase) and then - secreted in the extracellular environment. There HlyA subunits oligomerize into cation-selective channels and in the presence of Ca2+ attack the eukaryotic membranes, resulting in an increased permeability of erythrocytes to Ca2+ and K+ 18,105,106. Studies with bladder epithelial cells infected with UPEC strains UTI89 and UTI89∆hlyA demonstrated that HlyA induces a mesotrypsin-mediated proteolysis of focal adhesion- associated proteins such as paxillin (a cytoskeletal scaffolding protein) which triggers the exfoliation of the bladder epithelial cells 107. HlyA can also induce the proteolysis of different macrophage proteins such as IκBα and RelA 107.

Cytotoxic necrotizing factor-1 (CNF1). CNF1 is produced by UPEC and by some NMEC strains 108,109. It belongs to the family of dermonectrotic toxins which activate Rho proteins. Its gene, cnf1, is also located on NMEC and UPEC PAIs (in UPEC, often linked to HlyA and P-fimbrial genetic determinants) 110. Generally, the CNF1 toxin has been implicated in the actin rearrangement of the host cells upon invasion by ExPEC bacteria, i.e., the invasion of the bladder epithelial cells by the cystitis UPEC isolates and transcytosis of NMEC through the brain microvascular endothelial cells (BMECs) 111,112.

Serine Protease Autotransporters of Enterobacteriaceae (SPATEs). SPATEs is a subfamily of extracellular serine proteases produced by enterobacteria 18,113. UPEC bacteria produce three different SPATE toxins whose genetic determinants are also located on PAIs 113,114. Secreted autotransporter toxin (Sat) is one of them with proteolytic activity towards casein, factor V and spectrin. Sat exerts a cytotoxic effect on VERO kidney cells, HK-2 human bladder cells, and HEp-2 cell lines which also triggers their vacuolation 115. Sat is also shown to exert in vitro proteolytic activity on cytoskeletal (fodrin and leukocyte function-associated molecule 1) and also on nuclear (microtubule-associated proteins, poly-(ADP-ribose) polymerase and signal-induced proliferation- associated 1-like protein 1) 116. Protease involved in intestinal colonization (Pic) is another SPATE protease, not a toxin, shown to mediate serum resistance by proteolytic deactivation of the components of the complement system 117-119. Temperature – sensitive haemagglutinin (Tsh), also known in UPEC as Vat, is a protein first described in APEC (later found in UPEC) to release haem from haemoglobin which can later be taken up by the bacteria 118,120. Tsh can also bind collagen IV and fibronectin which would suggest that it, in addition to its other functions, serves as a mean for adhesion 121. It was shown that both, Pic and Tsh,

can cleave leukocyte surface glycoproteins (such as CD43, PSGL-1, CD44, CD45, CD93, and fractalkine/CX3CL1), interfering with their attraction and migration 122. Genes coding for SPATE proteins were also found in NMEC, especially in strains associated with neonatal septicemia 123. The most frequently produced SPATEs by ExPEC are Tsh and Sat 123.

2.2.3. Invasins. This group of virulence factors will be presented in light of NMEC pathogenicity since the hallmark of this disease is the successful crossing of the blood brain barrier (BBB) 18,124. NMEC use two strategies to cross the BBB. The first one is the penetration into BMECs resulting in transcytosis into the central nervous system (CNS), while the second strategy is to enter and multiple in neutrophils which during inflammation, cross the BBB and deliver the bacteria in the CNS 18,125. In both cases, the NMEC bacteria trigger their own uptake by the host cells, a process called invasion. NMEC invasion is accomplished by the help of a group of OMPs called invasins.

IbeA (Invasion of Brain Epithelium, also called Ibe10) is the first discovered invasin in NMEC with its gene located on PAI 126-129. IbeA is a unique virulence factor for NMEC compared to the other human pathogenic E. coli (its gene can also be found in the genomes of some APEC strains) 130,131. IbeA contributes to the invasion of BMECs via binding to vimentin, located on the membrane of the host cells 128,132. Binding to vimentin triggers the actin rearrangement through the activation of Rac1 (a molecular switch from the Rho family GTPases for cytoskeleton organization) – a process in which CNF1 also participates 111 133. IbeA is further involved in the invasion of neutrophils 134-136. In short, IbeA induces NF- κB activation via CaMKII that phosphorylates TAK1 and Erk1/2 which leads to the activation of IKKα/β 134-136.

OmpA in addition to its role as a porin and receptor for different bacteriophages, OmpA plays a major role in NMEC pathogenesis as an invasin. The protein is conserved among proteobacteria and is important for the NMEC invasion of neutrophils and BMECs 125,137. OmpA is especially important for the survival and multiplication of NMEC in neutrophils since upon binding to gp96, it triggers signaling that leads to suppression of the oxidative burst via downregulation of rac1, rac2 and gp91 125. Besides OmpA, NMEC bacteria utilize other OMPs present also in the commensal strains as invasins, such as IbeB, YijP and AslA 131.

2.2.4. Siderophores As discussed in Subchapter 1.2.2.2., iron is an essential and limiting factor for bacterial growth. Outside of the GIT, the iron is present in its less soluble form, i.e., Fe(III), and is constantly bound to the host’s compounds such as haem, haemoglobin, transferrin and lactoferrin 40. For example, there is 20µM of iron

in the blood plasma, but the free iron is only 10-18M 138. Bearing in mind that bacteria need 105 Fe3+ per generation, iron limitation by the host is a logical, widely applied defense mechanism against invading pathogens 40.

To retrieve iron (III) from the specific niche and to outcompete the host’s iron- binding proteins, in addition to the existing iron acquisition systems (discussed in Subchapter 1.2.2.2.), ExPEC (NMEC and UPEC) synthesize and secrete more siderophores which are superior to enterobactin 16,40,71,106. The genetic determinants of the additional siderophores are also located on mobile genetic elements 16,40,71,106. Generally the siderophore production increases the survival of ExPEC in the blood and urine 106,124,139,140. The function of three siderophores, typical for ExPEC is discussed.

Aerobactin. It is a member of the mixed type siderophores, i.e., it is a citryl- hydroxamate siderophore 40,106,139. Its operons iutAiucABCD are located on PAI, on ColV or ColBM plasmids 139,141. Aerobactin is considered superior to the other siderophores for several reasons: it is has higher binding constant at neutral pH than enterobactin, it does not require recycling once taken up by its receptor, and it is more soluble and stable than enterobactin 106,139.

Salmochelin. Together with enterobactin, it belongs to the group of the catecholates and represents different glycosylated forms of enterobactin 139. Its production is genetically determined by the iroNiroBCDE gene cluster either located on PAI, or on ColV or ColBM virulence plasmids 139,141,142. It was shown that in a UTI infection model, the catecholate siderophores had a minor role, compared to the non-cholate and haem receptors which contributed significantly to the virulence 143.

Yersiniabactin. It belongs to the group of the phenolates 139. Its genes are located on High Pathogenicity Island of Yersinia that was acquired horizontally by the ExPEC bacteria 139. Yersiniabactin is required for the iron uptake during infection, and promotes biofilm formation in urine 143,144.

2.2.5. Virulence-associated adhesins. Establishing in a new environment always starts with the difficult necessity to physically remain there. To achieve that, ExPEC bacteria are equipped with different kinds of adhesins via which they can selectively attach to different biotic and/or abiotic surfaces, and establish a physical contact with the host. The adhesins could be present on fimbriae (basic features explained in Subchapter 1.2.1., and exemplified with the biogenesis of type-1 fimbriae), on fimbrillae (flexible, extended organelles with adhesins present all over the structure) or they can be directly associated with the OM (the so-called afimbrial adhesins) 18.

Different adhesins selectively adhere to different host receptors, which provide the bacteria with the ability to colonize their niche and to perform tissue tropism.

2.2.5.1. Virulence-associated fimbriae. Although, a prototypical ExPEC strain, such as UPEC strain CFT073, encodes 12 different fimbrial gene clusters (10 of which code for CUP fimbriae), only the role of the most widely studied CUP fimbriae, i.e., type-1, S- and P- fimbriae will be discussed 18,145. The biogenesis of the CUP fimbriae is exemplified with the production of type-1 fimbriae in Subchapter 1.2.1.. The biogenesis of S- and P- fimbriae follows the same trend (Fig. 2) with the main difference being the binding specificity of the S- fimbrial and P- fimbrial adhesins. Unlike type-1 fimbriae, the genetic determinants of the P and S fimbriae are encoded on PAIs and are specific for the genomes of ExPEC.

Type-1 fimbriae. As explained in Subchapter 1.2.1., type-1 fimbriae are produced by both commensal and pathogenic E. coli strains. Type-1 fimbriation assists in the virulence of UPEC and NMEC during different stages of their pathogenesis 17,33,146-149. For example, experiments done in infant rats infected with type-1 and S fimbriated NMEC bacteria demonstrated that type-1 fimbriated cells are less virulent with low mortality 150. At the same time, Teng et al. showed that type-1 fimbriae are important for the interaction of NMEC bacteria with human BMECs 149. Type-1 fimbriae enhance the urovirulence and trigger the uptake of the UPEC bacteria into the bladder endothelial cells by binding to mannosylated glucoconjugates and uroplakin 1a with their FimH adhesin 17,146,148,151,152. P-fimbriae. P-fimbriae are encoded by the pap (pyelonephritis associated pili) operons in the genomes of up to 70% of UPEC isolates 153,154. Eleven genes are required for the biogenesis of the P fimbriae (Fig. 2, Fig. 5 and Fig. 6A) which includes papA, coding for the main fimbrial subunit and papG encoding the adhesin via which UPEC bind to the digalactoside epitope of vascular endothelium and muscular tissues 155,156. P-fimbriae are typically produced by pyelonephritis UPEC isolates and the presence of the pap gene cluster is normally associated with highly virulent UPEC strain 153. The role of the P-fimbriae in the pathogenesis of UPEC is not yet clear since Mobley et al. showed that there is not any difference between pap deletion mutant and its wild type in the colonization of the mouse of ascending UTI 157. On the other hand, in a study done by Roberts et al E. coli strain DS17∆papG did not cause significant kidney damage in a pyelonephritis model in monkeys while its wild type caused kidney failure 158.

S-fimbriae. Studies on the sfa (sialic acid fimbrial adhesion) gene cluster coding for the S fimbriae is one of the main aspects of this thesis. S-fimbriae are found in 15% of UPEC and 30% of NMEC and they allow bacteria to adhere to sialylated glucoconjugates 99,150,153,159. They were discovered by their ability to perform

mannose-insensitive haemagglutination 160. The genes coding for the biogenesis of the S-fimbriae are located in the sfa (sialic acid fimbrial adhesion) gene cluster, coding for the major (sfaA), minor (sfaD, sfaG, sfaH) subunits of the fimbriae, the adhesin (sfaS), the chaperone (sfaE), usher (sfaF) and the local regulators (sfaB and sfaC) involved in the S-fimbrial biogenesis (Fig. 5 and 2) 161,162. S- fimbriae are normally produced by the cystitis UPEC isolates, allowing binding to vascular and glomerular epithelium as well as to connective tissues 156,163. The S- fimbriae, together with the K1 capsule, are considered a major virulence factor of NMEC 79,86,164. S-fimbriae can bind brain tissue and cultured endothelial cells 165,166. Additional experiments performed with bovine BMECs indicated that SfaS fimbrial adhesin promotes the attachment of the NMEC bacteria to BMECs due to the sialic acid containing glycoproteins and interestingly, SfaA also plays a role as an adhesive molecule that recognizes sulfated glycolipids 159,167.

2.2.5.2. Regulation of virulence-associated fimbrial gene expression. Given the presence of more than one fimbrial gene cluster and the fact that different fimbriae recognize different receptors on different host tissues, it is rational for a differential fimbrial gene expression to exist. The differential fimbrial expression allows the bacteria to sequentially colonize different niches and to perform tissue tropism. Fimbrial gene expression is a phase-variation event, i.e., it happens in a subpopulation of cells resulting in different types of fimbriated and non-fimbriated bacteria 162.

The fim gene cluster (coding for the type-1 fimbriae) expression is regulated via the recombinase activity of FimB and FimE that leads to positioning the main promoter, located on a 314bp DNA invertible element, fimS, in line with the fim structural genes 168,169. FimB can invert fimS from “ON” to “OFF” and vice versa, while FimE activity predominantly results in “ON” to “OFF” transition 168,169. It is interesting to mention the presence of some proteins with high homology to FimB, i.e., IpbA (55% identity to FimB) and IpuA (48% identity to FimB) in the genomes of UPEC and NMEC strains which can additionally control type-1 fimbrial phase variation 170,171.

The sfa gene cluster shares high homology with the foc, prf and pap gene clusters (Fig. 6A) coding for the type-1C, P-like fimbriae and P-fimbriae and their gene expression regulation follows the same trend (Fig. 6B) 162,172,173. The main sfa promoter is a platform for many regulatory events with hierarchical organization. Unlike the expression of the fim operons, the expression of the sfa gene cluster is under epigenetic regulation. The first event is the methylation of the proximal GATC site of the promoter region by the Dam methyltransferase so that the involved regulators can interact with the DNA elements in the neighborhood of the distal, non-methylated GATC site 174-176. The regulation of the S-fimbrial biogenesis is a product of local - SfaB and SfaC, and global regulators - Lrp,

cAMP-Crp and H-NS 19,162,177,178 (Fig 6B). SfaC, which shares 100% homology with the positive local regulator of the pap gene cluster PapI, is the main positive regulator of the sfa gene cluster which turns on the sfa gene expression by binding to the main promoter region, stabilizing the DNA-cAMP-Crp-Lrp complex and exposing the promoter to the RNA polymerase while H-NS is the main repressor which binds to AT-rich DNA regions and hinders the access of the RNA polymerase to the promoter 19,162,177,178. This complex regulatory network triggers the expression of the structural genes of the S-fimbriae, but it also regulates the expression of all the genes part of this cluster. To ensure the presence of particular type of fimbriae in certain niche, ExPEC can suppress the expression of type-1 fimbriae by expressing SfaB (or PapB) which inhibits FimB activity, in addition to the inhibition of fim expression, SfaB can upregulate the expression of the pap genes and PapB – the expression of the sfa genes, which suggest that their presence is required for the colonization of the same tissue 179,180.

In the end of the section, I would like to highlight the importance of the environmental regulation of the expression of fimbriae. Their expression is dependent on osmolarity, temperature and nutrient availability. For example, the expression of type-1 fimbriae, P-fimbriae and S-fimbriae is optimal at 37⁰C, while in rich medium or in medium with glucose or in high osmolarity, bacteria downregulate the expression of P- and S-fimbriae 181-184.

Figure 5. Comparison between NMEC (in blue) and UPEC (in red) bacteria. Even though that both groups share a lot of virulence traits, the virulence factors of NMEC are less diverse and more defined.

Figure 6. The genetic determinants of virulence-associated fimbriae encoded on ExPEC PAIs.

Fig. 6A Schematic comparison of the operons coding for P-, Type-1C and S- fimbriae showing similar organization and homology. One of the main differences is the type of adhesin they determine. Fig. 6B The regulation of virulence-associated fimbriae follows the same trend on transcriptional and post-transcriptional levels.

3. Pathoadaptation Pathoadaptation is any of the changes that adapt bacteria to a new host’s niche which eventually results in a disease. It is a process that involves positive selection of random mutagenic and non-mutagenic (e.g., epigenetic) events which increase the fitness of bacteria at a certain place in a certain time.

3.1. The Eco-Evo dynamics of ExPEC Ecological evolutionary (Eco-Evo) biology presents the environment as a way to induce and select for genotypic and phenotypic variation in bacteria. Once the new variants appear, natural selection can privilege certain phenotypes/genotypes by making them dominant in the bacterial population. This concept also holds true for pathogens, i.e., they have to access their habitat by crossing certain barriers, sustain and thrive in that habitat and in the end, they have to be transmitted to a new one. There are three groups of pathogens based on their ecology – professional, opportunistic and accidental.

The species of E. coli provides examples for all ecological groups of pathogens. The professional pathogens, such as EIEC/Shigella spp., always cause disease as part of their natural life cycle. The group of ExPEC is represented by opportunistic and accidental pathogens. UPEC strains are opportunistic pathogens – a group that is quite diverse and with complex ecological and evolutionary relationships. As opportunistic pathogens, UPEC bacteria live as commensals in GIT, and only when opportunity appears, i.e., upon breach of the defense systems or barriers of the host, they are able to transmit to another niche, i.e., the urinary tract, and cause disease (Fig. 7). An important feature of their life cycle is their ability to always return to their commensal life in GIT (as their original habitat), once they are shed by the urine. NMEC, on the other hand, can be considered as accidental pathogens for the fact that they establish infections in dead-end zones such as the CNS which does not allow them any further dissemination and thus, they cannot return to their primary habitat (Fig. 7).

Figure 7. Ecology of ExPEC, the routes of the commensal life style are shown in black, while the routs for extraintestinal infections are shown in blue for UTIs and in red for CNS invasion.

It is interesting to hypothesize that ExPEC virulence might have appeared coincidentally. Tenaillon et al. and Sokurenko explore in their reviews this possibility 46,185. In both reviews it is suggested that certain virulence traits should be considered as traits that increase the fitness of ExPEC bacteria not only in their pathogenic niche, but also in the gut 46,185. Such is the example with the iron acquisition systems and the O antigenic variation which protects the commensal E. coli bacteria against grazing predators in the gut 186-188. To further speculate on the hypothesis, if ExPEC virulence is of an accidental origin in the gut, one possible scenario would be the selection for K1 encapsulated E. coli bacteria. As discussed in Subchapter 2.2.1., in addition to the fact that the K1 antigen is a shared virulence factor by UPEC and NMEC, it is the main virulence factor in NMEC 79,96,189. Gut microbiota is composed not only of bacteria and archaea, but also of protozoa and bacteriophages 190,191. Studies on interactions of E. coli K1 with Acanthamoeba castellanii (a keratitis isolate) showed that in addition to LPS and OmpA, the K1 capsule is an important factor for the association, invasion and survival of the bacteria in the amoebae 187,192,193. The protection from grazing predators is one of the reasons that can result in “enrichment” of a certain E. coli population with K1 bacteria. Another reason for the selection of K1 bacteria in the gut is the presence of bacteriophages, such as bacteriophage T7, to which the K1 capsule serves as a barrier 194. Taken together, these examples suggest that the selective forces in the gut also shape the adaptive potential of ExPEC population and only after the bacteria escape the gut, can this adaptive potential prove pathogenic.

3.2. The genetic plasticity of ExPEC The genetic plasticity of ExPEC, i.e., the presence of core and flexible gene pools in their genomes, is the main prerequisite for their evolution which can result in new phenotypes selected by the environmental forces. The flexible gene pool is represented by mobile genetic elements which are integrated on the bacterial chromosome or they exist independently, plasmid-borne. The genes that belong to the flexible pool are acquired and it can also be transferred intra- and inter- specially via the means of HGT – conjugation, transduction and natural transformation.

3.2.1. Pathogenicity islands Genomic islands are horizontally acquired blocks of novel genes, encoding adaptive traits proved useful before, integrated on the chromosome. The genomic islands share several common features: (i) they carry genes coding for integrases and IS elements and sequences of phage or plasmid origin; (ii) their size is between 10 and 100 kb; (iii) their sites of insertion are predominantly the tRNA loci; (iv) they have different GC content, codon usage; and (v) they may be unstable 195,196. If the new traits acquired via HTG on genomic islands contribute to the pathogenic potential of the strain, the genomic islands are called PAIs 196. First PAIs were described in the genomes of pathogenic E. coli and later, they were discovered in the genomes of other bacterial species 104,195,197. The presence of PAIs is one of the main characteristics of ExPEC and is the reason for the bigger genomes of ExPEC bacteria compared to the genome of E. coli K-12 commensals. PAIs are the mosaic component of ExPEC chromosomes (Fig. 8) and they are part of the flexible gene pool. The PAIs were extensively studied in three prototypical UPEC strains 536, J96 and CFT073 196. There are 5 PAIs described 196-199 in the genome of strain 536 . The structure and organization of PAI I536, PAI 196-199 II536 and PAI IV536 are depicted in Fig. 9 . The evolutionary advantage provided by the PAIs is the acquisiton of different operons coding for different traits. For example, the genetic determinants of HlyA and P fimbriae are often encoded on the same PAI (Fig. 9, see PAI II536), such as in the case of the sfa gene 196,198 cluster and the iro operon encoded together on PAI III536 . PAIs whose sequences are flanked with direct repeats are unstable and the bacteria can lose them at high frequencies as studies on strain 536 showed 197. NMEC genomes are also mosaic due to the presence of PAIs. Studies on prototypical NMEC strains IHE3034 and RS218 showed that most of their PAIs are similar to the UPEC ones, for example RDI4 (from RS218-derived islands) carries the sfa gene cluster, which can also be found on PAI of the genome of strain IHE3034 200,201. NMEC genomes also carry unique PAIs, e.g., ibeA gene (typical only for NMEC and APEC) is located on RDI22 201.

Figure 8. Genomic comparison of three ExPEC chromosomes. The PAIs are present in fragments colored in red, blue, yellow and green due to the different algorithms used for their detection (CG content, tRNA locus for integration, presence of genes coding for genetic mobility and so on…). The differently colored dots, explained in the figure, represent pathoadaptive mutations linked to the virulence of the particular strain.

3.2.2. Pathoadaptive mutations Mutations in the core gene pool that increase the pathogenic potential of bacteria are called pathoadaptive mutations. They affect genes found in ExPEC and commensal bacteria. It is suggested that they are acquired in the process of pathogenesis, unlike PAIs, and that is why their presence represents true pathoadaptive trait 185. There are three main type of pathoadaptive mutations gene inactivation, gene amplification, and allelic variation 185.

Gene inactivation due to different point mutations or deletions can be due to bacterial adaptation to a very specific niche, such as the case with the intracellular pathogens Buchnera, Chlamydia trachomatis, Mycoplasma, Mycobacterium leprae 185,202. This kind of gene reduction results from the lack of gene usage. Another type of gene inactivation eliminates genes whose products antagonize the virulence of the pathogen 202. These kind of genes are known as anti-virulence genes, and their products – anti-virulence factors 202. The concept behind is that expression of a certain gene is beneficial in one niche but may be deleterious in another. A classic example for an anti-virulence genes is the cadBAcadC genetic system coding for the production of cadaverine in E. coli 203-205. Lack of lysine decarboxylase activity, required for cadaverine production, is one of the main differences between Shigella/EIEC and the other E. coli strains 6. Cadaverine is a polyamine that plays an important role in cell adaptation to low pH 206. When the functional cadBAcadC operons were restored in wild-type Shigella, it was shown that cadaverine disrupts ShET1/ShET2 activity, inhibits neutrophil transepithelial migration and blocks the escape of bacteria from the phagosome 205,207,208. It is also worth mentioning that the cadBAcadC operons are also deactivated in a set of STEC and EHEC strains due to the fact that restored cadaverine production reduced their adherence ability 209,210.

The second group of pathoadaptive mutations result in gene amplificiation, i.e., the increased gene dosage leads to improved pathogenicity. Such as the case with the duplication of hlyCABD operon located on two PAIs of UPEC strains 536 and J96 (Fig. 9) 211. In another UPEC strain, CFT073, carries two copies of the pap gene cluster also on two PAIs 211. In principle, overproduction of specific fimbriae can improve the strength of adhesion and lead to better colonization of the pathogenic niche.

Allelic variations belong to the third group of pathoadaptive mutations. They represent point mutations that do not truncate or delete genes, but lead to the production of mutated proteins that can have different specificities compared to its wild-type version. One example is the allelic variation of the fimH gene, coding for the adhesin of the type-1 fimbriae. Sokurenko et al. showed that there are 2 FimH variants – one variant, produced by the commensal E. coli, has low binding specificity to mono-mannose and the other FimH variant, produced by UPEC, exerts high binding affinity 212. The selection for FimH variants that bind strongly mono-mannose optimizes the tissue tropism of UPEC due to broad cell receptor specificity which provides an advantage during the colonization of urinary tract 213.

Figure 9. Genetic organization of three PAIs found in the genome of UPEC strain 536. The site of insertion (the tRNA locus) is shown in green; in yellow are depicted the genes coding for integrases or IS; the genes coding for HlyA production are shown in red, while the genes coding for fimbriae – in blue. The genetic determinants for the yersiniabactin iron acquisition system are shown in purple.

3.3. Pathoadaptation and metabolism The virulence factors of ExPEC provide these bacteria with the ability to reside in the blood and the cerebrospinal fluid which, in terms of nutrients, contain enough carbon and energy sources (such as glucose) to allow for bacterial growth.

Urine, on the other hand, is one of the harsh environments which UPEC bacteria occupy. It is a complex, nutrient-poor environment, characterized with buffered pH (6-7), high osmolarity (due to the presence of 0.5M urea and 0.29M NaCl), and dissolved oxygen of 4 ppm 214-216. From nutritional point of view, urine represents a dilute medium which contains amino acids, small peptides, glucose (its concentration varies between 0.05 and 3.4 mM), citrate (concentration between 1 – 4 mM) and other compounds 217-220. A shift from an anaerobic, reducing environment enriched with nutrients into new oxidative niches with limited nutrients often requires changes in the key metabolic pathways of these ExPEC bacteria. Alteri et al. showed that the primary carbon, energy and nitrogen source for UPEC in the urine are the short peptides and the amino acids 221. Besides amino acids and peptides, UPEC bacteria can also utilize hexouronates and hexanates 222. The import of peptides, gluconeogenesis and the TCA cycle are necessary for UPEC in the course of UTI, while glycolysis, the pentose phosphate pathway and Entner-Doudoroff pathway are not 221. Snyder, et al. demonstrated that UPEC bacteria in a mouse urinary tract infection model downregulate many genes involved in the anaerobic bacterial growth 222. Though urea is abundant in urine, E. coli are urease negative and cannot use urea as a nitrogen source. The nitrogen limitation is overcome by producing the GluP and GluQ glutamine importers 222.

3.4. Pathoadaptation and developmental programs

3.4.1. Signaling networks that regulates sessility and motility in ExPEC The lifestyles and behavior of bacterial cells may include regulatory signaling mechanisms controlling switches between motility and sessility. The key element in their signaling cascades is the second messenger bis-(3’, 5’)-cyclic dimeric guanosine monophosphate (c-di-GMP), an important molecule which serves as a main motility-sessility switch, manufactured only by some eubacteria 223-225. When c-di-GMP is needed in E. coli, the second messenger is produced by a group of enzymes that have the GGDEF domain - diguanylate cyclases, and degraded, when no longer needed, by c-di-GMP phosphodiesterases that carry the EAL domain 226,227. There are 29 genes encoding GGDEF/EAL proteins in E. coli - 12 genes coding for GGDEF proteins, 10 genes for EAL proteins, and 7 genes that encode proteins carrying both domains 228,229. The third component of the c-di- GMP signaling module in E. coli is the receptor protein that is activated upon binding to the messenger. The receptor proteins belong to the PilZ-like family and

so far, there are 2 effector proteins in E. coli – YcgR and BcsA 225,230. Previous studies done with commensal E. coli showed that c-di-GMP is involved in motility, curli-mediated adhesion, and ribosome-triggered biofilm formation 231- 233. Even though c-di-GMP signaling has been thoroughly investigated, the role of certain EAL PDEs, such as YcgG, in the lifestyle of the commensal E. coli strains remains unknown. The role of c-di-GMP was also studied in some ExPEC strains, mainly in uropathogenic E. coli CFT073 and to a lesser extend in E. coli IHE3034 200,234-237. Overall, besides sessility and motility, in E. coli CFT073, c-di-GMP represses type-1 fimbriae expression in a pst mutant variant, and the presence of YdiV (a degenerated PDE) represses the formation of P-fimbriae 234-236.

Downstream of the sfa/foc/pap structural genes (also downstream of foc and pap genes), there are two novel regulatory ORFs, part of the sfa gene cluster – sfa/pap/focX and sfa/pap/focY (Fig. 6B) 164,200. Their products are also involved in the regulation of motility. The SfaX-like proteins belong to the MarR family of transcriptional regulators 238. The family is represented by alpha helical members that bind palindromic DNA repeats via a winged, helix-turn-helix motif 238,239. PapX/SfaX-like proteins are mainly involved in the regulation of the type-1 fimbrial phase variation and bacterial motility 240. The role of the ‘X’ proteins in motility was demonstrated for the first time in NMEC strain IHE3034 240. Sjöström et al. found that overproduction of the sfaX gene abolishes the bacterial motility, because as it was discovered later, SfaX binds to the main promoter of class I flagellar operons, repressing the expression of flhCD which codes for the FlhCD transcriptional activator of fliC 238,240. Studies on PapX and FocX confirmed that the ‘X’ proteins repress the expression of the flagellar genes also in UPEC 241,242. Sfa/Pap/FocY proteins are novel, active, stand-alone c-di-GMP EAL PDEs, produced by NMEC and UPEC 200. Previous studies done in our lab showed that SfaY is an active PDE and its gene, i.e., sfaY, is under the regulation of the main promoter of the sfa gene cluster, responsible for the assembly and formation of the virulence-associated S-fimbriae 200.

3.4.2. Formation of Intracellular Bacterial Communities (IBCs) by UPEC bacteria. Surviving in many disparate host niches is the major reason for the ExPEC strains not only to be equipped with a diverse arsenal of virulence factors, but also to develop morphological plasticity which supports their adaptation and survival. One of the good examples for this morphological plasticity is the formation of the intracellular bacterial communities (IBCs) by the UPEC strain UTI89 upon invasion of the urothelial bladder cells during infection 243. During their intracellular lifestyle, after a couple of generations they start differentiating into long, filamentous bacteria, and small motile and non-motile cells 244. It is assumed that the filamentous bacteria break the uroepithelial cells open, propel the small bacteria and prevent them from being phagocytosed by the immune

cells 244. Thus, the filamentous UPEC species play an important role in the UPEC pathogenicity and the IBCs play a major role of the chronic phase of the infection 245. Genetic studies done on UPEC filamentation revealed two ways for the bacteria to elongate their cells – one study showed lack of filamentation of UTI89 cells that carried deletion in the sulA gene and the other study showed lack of filamentation in damX UTI89 deletants 243,246.

II. Thesis Scope

The results of this thesis are organized into three parts, depending on the aspects of the ExPEC pathoadaptive events which were investigated.

In the first part of the thesis research, we questioned the reasoning behind the inactivation of the main global stress regulator RpoS in NMEC. To provide better understanding of such an event, we asked the following questions:

1. Since RpoS is the main regulator of c-di-GMP signaling 247, we asked how much c-di-GMP NMEC bacteria produce compared to the UPEC and the commensal K- 12 strain.

2. We addressed the link between the rpoS inactivation and the production of the main virulence factors of NMEC strain IHE3034, i.e. the S fimbriae.

3. Since it has been reported that the loss of RpoS increases the spectrum of compounds that E. coli can metabolize 248, we also asked if this particular loss can contribute to the utilization of substrates that E. coli bacteria do not metabolize even though they code for the whole repertoire of enzymes and transporters. And if such metabolic pathoadaptability exists, how conserved it is in the group of ExPEC.

The second part of the thesis research involved the investigation of clyA which codes for a cryptic toxin with heamolytic and cytotixic activities. It is a cryptic toxin because its gene expression is repressed by H-NS in the genomes of commensal E. coli. The situation with clyA is different in the ExPEC genomes – the gene expression was abolished via various deletion events which took place either in the gene or in its promoter region. We tried to find the reason behind these events and if they carried some pathoadaptive potential to the strains. To achieve that, we used UPEC (J96 and 536) and NMEC (IHE3034) strains and we asked the questions:

1. What happens if the deleted clyA allele is replaced with its intact version? Will the haemolytic activity be restored?

2. Is there some regulatory interference by regulators, specific for ExPEC, which can lead to overproduction of the restored ClyA?

3. How the restored clyA affects the fitness of the ExPEC bacteria grown in media, mimicking their natural habitats? And how is their fitness affected when the ClyA+ ExPEC are exposed to antimicrobial peptides?

In the last part of the thesis research, we were in a quest to understand the role of SfaY, which we previously showed to be able to code for a stand-alone c-di-GMP PDE, in the lifestyle of NMEC 200. We asked the questions:

1. How does the phase variation of the main sfa promoter affect the expression of sfaY?

2. Is there specific phenotype caused by the deletion of the gene or by its overexpression? And under what conditions?

III. Results and Discussion

1. Absence of global stress regulation in Escherichia coli promotes pathoadaptation and novel c-di-GMP-dependent metabolic capability Trade-offs in bacterial evolution often represent mutually exclusive adaptations which suggest that loss of one function can lead to the gain of another. In the first project we investigated if the loss of global stress regulation in a subset of NMEC strains is such a trade-off that results in pathoadaptation. NMEC strain IHE3034 was isolated in Finland in 1976, while RS218 is a neuroinvasive American pathovar 249,250. Both strains lack active RpoS due to the premature interruption of its translation 251.

RpoS belongs to the family of the σ70 factors that act in the RNA polymerase complex at initiation of transcription 252,253. RpoS is the main global stress regulator that provides cross-resistance against starvation (due to limitation of P, N, Mg and C); against changes (increase and decrease) in temperature, pH, osmolarity; and against oxidative stress 253-255. RpoS triggers the c-di-GMP signaling in E. coli and the deficiency in RpoS results in reduced biofilm formation 247,256.

The first step we made was to compare the c-di-GMP levels of RS18 and IHE3034 to the levels produced by the commensal E. coli MG1655 and UPEC strains 536 and UTI89. What we found was that NMEC strains deficient in RpoS maintained extremely low c-di-GMP levels even compared to the UPEC strains with which they share a lot of common traits. After restoration of active RpoS, NMEC c-di- GMP levels increased to the ones produced by the commensal and UPEC strains.

Then, we involved the role of ycgG (pdeG) 257. The ycgG gene codes for a membrane-bound PDE in the genomes of different E. coli strains, including the commensal 229. Sarenko et al. and Reinders et al. showed that the expression levels of ycgG are very low under different laboratory conditions 258,259. In most of the genomes of E. coli strains belonging to the B2 group, the ycgG gave rise to a new allelic variant – ycgG2. The new allelic variant resulted from a deletion event which removed the part of the ORF coding for the membrane-binding domain, followed by a point mutation that turned the first codon of the coding sequence into a rare start codon, i.e. TTG. Via in-vitro transcription/translation we showed that ycgG2 is, indeed, an ORF and then we showed the cellular expression of the protein in LB and artificial urine medium (AUM). In the end, we showed that ycgG2 has also a phenotypic impact on biofilm formation by

overexpressing it in Vibrio cholerae luxOc strain which resulted in a transition from rugose to smooth colony morphology.

Next, we investigated the effect of RpoS inactivation and the presence of ycgG2 on the production of virulence factors by NMEC strain IHE3034. We decided to run the experiments in AUM, since we were interested if the NMEC strains might exhibit urofitness, since it is suggested that for the vertical transmission from mother to infant, NMEC can reside in the maternal urinary tract. The initial results indicated that NMEC bacteria grow well in AUM. Moreover, the expression of the S-fimbriae was higher than the S-fimbrial expression by the bacteria cultured in LB which suggested that NMEC bacteria can also colonize UTI. When IHE3034 cells were cultured in AUM and LB, deletion of ycgG2 and the restoration of RpoS led to reduced S-fimbrial expression.

We also considered that lack of RpoS in NMEC might contribute to potential alteration in NMEC metabolism due to a regulatory trade-off between nutrient utilization and stress resistance. The primary interest was to find if NMEC bacteria can use substrates for which E. coli encode the enzymes and the transporters but due to their separate production, the utilization and the metabolic pathways are incomplete. We tested if NMEC can utilize citrate aerobically according to Gottschalk’s conditions, i.e., the presence of a co- substrate that provides the reducing power to the malate dehydrogenase and fumarate reductase 39. To achieve that, we modified Simmon’s agar 260 by adding 0.2% D-glucose embedded on the surface. Results indicated that IHE3034 bacteria were able to grow on Simmon’s medium, i.e., to use citrate. As described in this thesis, the ability of E. coli to use citrate is prevented because the main citrate importer, CitT, is only expressed anaerobically and during anaerobic growth the TCA cycle is repressed 39. To test if CitT is expressed aerobically by IHE3034, we deleted its gene and showed that the IHE3034∆citT bacteria were not able to grow on modified Simmon’s agar. We also showed that the restored RpoS activity could abolish the growth of IHE3034 in presence of citrate, which suggested for the existence of a trade-off that favored the nutrient competence over the cross-resistance.

The finding that IHE3034 is able to grow on citrate stimulated us to investigate how conserved this metabolic reprogramming is in the group of ExPEC. To do that, we used two RpoS+ UPEC pathovars – strain UTI89 (a cystitis isolate, serotype O18:K1:H7) and strain 536 (a pyelonephritis isolate, serotype O6:K15:H31). UTI89 strain showed interesting behavior. In the beginning, the wild type did not use citrate and exhibited retarded growth at 25⁰C on Simmon’s medium with glucose, but after we inactivated rpoS, UTI89 bacteria started using citrate and their growth at 25⁰C on the modified Simmon’s agar was restored.

These findings further extended the advantages of RpoS inactivation. The results from the experiments with strain 536 were also interesting. Even though 536 has active RpoS, its bacteria can use citrate. Moreover, we tested more pyelonephritis Finnish isolates which also showed the ability use citrate aerobically which suggested that this type of citrate utilization is property of ExPEC and most probably it improves the fitness of bacteria when they thrive in the urine during their infection process.

Finally, we were curious if NMEC bacteria have adapted their CitT transporter not only for citrate uptake, but also for iron (III) citrate uptake. On the quest to investigate this property, it came to our attention that a vast majority of ExPEC, including the NMEC and UPEC strains used in our experiments, have lost the major operons coding for the iron (III) citrate uptake, i.e., fecIfecABCDE. This further stimulated us to test if CitT can also serve as a portal for iron (III) citrate. We performed an in vitro Fe(III) citrate utilization assay by comparing the iron accumulation in IHE3034 and its citT deletion mutant. Interestingly, only the wild type bacteria could accumulate iron while the mutant cells failed. We also tested if the ∆ycgG2 mutant bacteria can take iron (III) citrate. Surprisingly, they failed. Having in mind that iron is chelated with citrate in the gut, one can speculate that in their commensal phase of growth, IHE3034 bacteria express YcgG2 which stimulates the iron (III) citrate uptake. Once they escape the gut, YcgG2 is decreased and the bacteria trigger the citrate fermentation program due to nutrient restriction. Thus, YcgG2 could serve as a switch, i.e. when present, it triggers iron (III) uptake, while when it is downregulated – citrate fermentation. CitT is a citrate/succinate antiporter and one plausible mechanism for the switch could be that under different conditions it can export anions, different from succinate, which may allow the uptake not only of citrate but also of iron (III) citrate.

2. Elevated recombinant clyA gene expression in the uropathogenic Escherichia coli strain 536, a clue to explain pathoadaptive mutations in a subset of extraintestinal E. coli strains Sometimes expression of a certain gene can reduce the fitness of bacteria in their new pathogenic niche. That is why gene loss, as a pathoadaptive mutation, does not only happen to anti-virulence genes, but also to genes whose products severely exacerbate bacterial growth. In the second project, we explored the role of another pathoadaptive mutation - the clyA (coding for cytolysin A) gene inactivation done by different mutations in the ExPEC genomes.

The clyA gene is the only genetic factor found in non-pathogenic E. coli strains that codes for a pore-forming toxin cytolysin A 261,262. The transcription of clyA

gene is silenced by the H-NS DNA-binding protein and the gene is considered inactive 263. Interestingly, the predominantly H-NS-repressed clyA gene was turned into a pseudogene, via various insertion, deletion and recombination events, in the genomes of a large set of ExPEC strains. This led us on the track to test if, even though assumed silent, loss of clyA gene could have some pathoadaptive value to the lifestyle of ExPEC.

In the beginning, we restored the clyA+ locus in UPEC strains 536 and J96. We confirmed that ClyA was produced intracellularly, but to confirm that ClyA was active, we grew the ClyA+ strains on Ca2+ - depleted blood agar plates (compared to HlyA, ClyA activity is Ca2+- independent) and stimulated the ClyA release by prophage induction with mitomycin C. After the phage release was induced, the beta-haemolytic phenotype of ClyA+ J96 and 536 was confirmed. We also showed via clyA::luxAB fusion that the clyA transcription happens during the exponential phase of the growth of the UPEC derivatives. The results so far suggest that the clyA+ expression in J96 and 536 is not under H-NS silencing, which leads to its higher production in the UPEC derivatives compared to the amount produced by the commensal isolates, and finally, the produced ClyA is retained intracellularly, i.e., it is not secreted.

Since the restored clyA+ locus was not silenced by H-NS, we tried to find if some ExPEC-specific regulators might play a role in its expression. We stopped our attention on SfaX/PapX DNA-binding proteins from the MarR transcriptional regulators whose genes (sfaX/papX) are encoded on the sfa/pap gene clusters in ExPEC 200,238,241,264. We speculated that the presence of SfaX/PapX might have anti-silencing effect that could counteract H-NS. UPEC strains code for more than one fimbrial gene cluster and hence, they express more than one type of SfaX/PapX regulators. To avoid cross-talk interference, we moved the experiments in IHE3034 background (an NMEC strain carrying only the sfa gene cluster and coding only for SfaX). We cloned the clyA+ gene on a plasmid under the regulation of its own promoter and we examined the ClyA expression in the wild-type IHE3034 strain and its sfaX deletion mutants. In the experiments we included another mutant strain, i.e., IHE3034∆sfaY. SfaY is a stand-alone c-di- GMP PDE which functionally resembles YcgG2 (the PDE studied in the first project) and whose gene sfaY is also part of the sfa gene cluster of IHE3034 200. The ClyA+ IHE3034 wild type produced higher levels of ClyA and displayed stronger haemolytic activity compared to the ones produced by ClyA+ deletion mutants. To further confirm that SfaX is a positive regulator of clyA transcription, we introduced sfaX in trans in strain 536 clyA::luxAB. The luciferase activity showed that SfaX enhanced greatly the transcription of clyA which further confirmed the role of the SfaX/PapX proteins as positive regulators of clyA+ in the ExPEC strains.

Thus far our results suggest that ExPEC do not silence the expression of the restored clyA, and on contrary, its expression is further stimulated by the activity of the SfaX/PapX regulators. The initial experiments also hinted that the produced ClyA is retained in the cells of J96 and 536, and its secretion requires phage induction. It was shown before that the accumulation of heamolysins in E. coli decreases the fitness of their cells 265, so the final part of the project was to check if the clyA gene was inactivated in the ExPEC genomes as a result of pathoadaptive mutation due to decrease in ExPEC fitness. We found out that the UPEC strain 536 clyA+ showed increased susceptibility to antimicrobial peptides (Polymixin B) and its fitness also decreased when the bacteria were cultured in AUM with higher concentration of urea.

3. A c-di-GMP-mediated morphotypic pathoadaptability of neonatal meningitis-causing Escherichia coli Another fascinating aspect of ExPEC is their ability to integrate the core, “house- keeping” signaling networks with the signaling networks of their pathogenicity. This integration may result in new phenotypic outcomes, bridging signaling cascades via novel molecular mechanisms. That would be the take-home message of the final project in which we explored the competence of NMEC bacteria for filamentation.

We started this project with investigating the role of SfaY as a stand-alone c-di- GMP PDE in the lifestyle of IHE3034. To follow up on the previous experiments done in our laboratory, we created a scarless deletion sfaY mutant in strain IHE3034 and explored its phenotype. The mutant cells did not display any changes behavioral changes in terms of motility, biofilm formation and growth under different conditions, such as changes in temperature, osmolarity and iron availability, compared to its wild type. Then we decided to overproduce SfaY in IHE3034 and to look for different phenotypes. This time, after we incubated the IHE3034 bacteria on solid media and induced the expression of SfaY and we performed scanning electron microscopy (SEM). The results showed curious phenomenon – the bacteria in which SfaY was overproduced formed heterogeneous population composed of long, filamented bacteria (30µm in length) and bacteria of normal size (1-2µm in length). The frequency of the filamented bacteria was similar to the frequency that is normally due to phase variation.

Then we aimed to clarify NMEC filamentation was a result from a phase-variation event since sfaY expression is under the regulation of the main promoter of the sfa gene cluster. To do that, we tested the frequency of filamentation when the sfa promoter was locked in its ‘ON’ phase by expressing in-trans the local positive regulator SfaC, together with a plasmid that carries sfaY. We carried out the main

experiments in IHE3034∆sfaY mutant background to avoid interference from the native SfaY regulator. The results confirmed our hypothesis – only the ‘ON’ bacteria overexpressing SfaY displayed 100% filamentation while the negative control bacteria remained with their normal size. In this experimental step, we also turned ‘ON’ the main promoter of IHE3034∆hns mutant since H-NS was shown to be main repressor of the virulence-associated fimbrial gene expression 19. IHE3034∆hns bacteria became elongated as soon as their sfa promoter was in its ‘ON’ state. In addition, the wild-type IHE3034 exhibited the same filamentation when its sfa promoter was in its ‘ON’ state and SfaY was overproduced.

Bacterial filamentation is a phenomenon of different values. From one aspect, it has been used as a read-out system in studies on bacterial cell division and that is how the essential genes that determine the bacterial cell division components were discovered under the form of fts (filamentation temperature-sensitive) conditional mutants 266,267. Another aspect involved the fact that for many years the filamented bacteria have been assumed to be the severely stressed and dying part of the bacterial population which holds true for many environmental conditions 244. In its third aspect, bacterial filamentation represents a survival strategy, part of the non-mutagenic events of the bacterial SOS stress response 268-271. In short, the SOS response is induced when bacteria are exposed to DNA- damaging agents and it is triggered by the interaction between RecA-ssDNA complex and LexA (a suppressor) 268-271. This interaction triggers the self-cleavage of LexA and results in derepression of genes whose products participate in the process of DNA repair 268-271. If the genotoxin persists, then bacteria enter in the non-mutagenic phase of the SOS stress response, they express a protein, called SulA, which interacts with the FtsZ monomers of the bacterial divisome and the formation of the division ring is prevented, so bacteria stop dividing until the stressor disappears 268-272. Bearing in mind that the SOS response is an inducible system, our next step was to see if there was a link between the filamentation that is under phase-variation and the induction of the DNA SOS stress response. We checked the levels of RecA and LexA in the filamented IHE3034 bacteria compared to the non-filamented ones and we observed reduced LexA production, coupled with increased levels of RecA in the filamented bacteria, indicating that indeed the SOS stress response was on. To further confirm that the filamented bacteria already entered the non-mutagenic events of the SOS stress response via SfaY, we worked only with the ‘ON’ strains of IHE3034 and its ∆sfaY deletion mutant. When stressed with mitomycin C (a potent DNA cross-linker), only IHE3034 ‘ON’ bacteria underwent filamentation compared to the ∆sfaY ‘ON’ mutant.

After we linked SfaY production to the DNA SOS stress response, we tried to characterize the filamented bacteria. We showed that the IHE3034 filamented

bacteria are non-fimbriated and carry multiple chromosomes. The fact that the filamentation was reversible, and the viable counts of bacteria remained the same, suggested that the filamented bacteria are alive.

The last step from this project was to test if filamentation was an NMEC pathovar- specific event. Surprisingly, when we included the American NMEC isolate RS218, its cells already exhibited the filamented morphotype. So, while we were trying to explore the filamentous potential of IHE3034, we embarked on the adventure to turn the filamentous RS218 cells non-filamentous. In doing so, we found out that SulA was the effector that triggered RS218 filamentation since after turning ‘ON’ the RS218 main sfa promoter, only the RS218∆sulA bacteria remained non-filamented. As far as IHE3034 filamentation is concerned, we hypothesize that it could be due to a proteolytic cleave of the FtsZ monomers since we detected an overproduction of ClpP - a component of the ClpXP proteolytic machinery and a slight change in the FtsZ levels. The filamentation of IHE3034 could also be due to overproduction of DamX, which we detected only in IHE3034 ‘ON’ strain overexpressing SfaY. DamX is a non-essential ‘SPOR’- domain protein involved in bacterial cell division, which, when overproduced, triggers filamentation 273,274.

4. Summary Regulation and mutation are the two main mechanisms via which bacteria adapt to new environments. Most of the studies done in this thesis are studies on pathoadaptive mutations which resulted in the improvement of either the fitness or the pathogenicity of ExPEC. For example, loss of RpoS resulted in the increase of the urofitness of NMEC strain IHE3034 and also triggered its ability for citrate utilization – an ability that normally does not exist in E. coli.

On the other hand, acquiring new regulators opens the way to new perspectives for investigating the anti-silencing of the so-called cryptic genes, as it is the case with the upregulation of ClyA in the UPEC strains studied here. Once expressed, the product of the cryptic genes might be disadvantageous for the bacteria and that is why these genes may be inactivated in the process of evolution such as the case with the inactive clyA allelic variants in the genomes of a large set of ExPEC strains.

Finally, the different ‘house-keeping’ signaling networks might serve different purposes for good reasons, such as the case with the c-di-GMP-dependent developmental program, under the control of phase variation that integrates the DNA SOS stress response in its final stage of execution in strains IHE3034 and RS218. The competence of NMEC for filamentation, shown here, could also serve as a clue to explain the formation of the filamentous bacteria of UPEC

intracellular bacterial communities for which also the two effectors are suggested – SulA and/or DamX 243,246.

5. Epilogue

The presence of ~ 100 K, 56 H and 164 O in E. coli can give rise to 10 000 possible serotypes 106. It is interesting to know what evolutionary forces selected for the existence of only a couple of hundred serotypes and it is even more curious to discover the conditions that can lead to the selection of new ones. Why is it, for example, that E. coli K1 are the predominant neonatal meningitis isolates and not E. coli K92 25?

Another interesting aspect that requires further exploration is the zoonotic potential of certain ExPEC strains causing diseases in animals (other than humans). One can also speculate that humans can also be reservoirs for ExPEC strains that can potentially infect animals. In line with this, certain strains are restricted to particular hosts due to certain environmental barriers but what if they acquire a virulence gene that would allow them to cross that barrier? One example is strain F18+STEC which causes sudden death in pigs due to the specificity of the receptor of its Shiga toxin (it is Gb4, not Gb3) 62. This strain cannot be transmitted to humans because they lack the receptor for its fimbriae, but what if this strain acquires the LEE pathogenicity island? One of the future perspectives for bacterial adhesion might be the discovery of new bacterial adhesins and their cognate host’s receptors.

To finalize this chapter, I would like to expand the view of the evolution of ExPEC strains. It is right to assume that ExPEC virulence is of human origin, since they reside as commensals in the gut and only when they escape the GIT can cause disease elsewhere inside the human body. Another interesting speculation could be if the ExPEC pathogenicity of certain strains evolve in the GIT of other animals. For instance, the UTIs in dogs have the same epidemiology as the human’s ones. Even though that specific NMEC strains have not been characterized in mammals (other than humans), certain APEC and NMEC strains are more similar to each other rather than to certain UPEC strains.

Microevolutionary processes always select for the fittest, on the other hand, microevolution eventually leads to macroevolution that results in the appearance of a new species. Even though that Escherichia coli is the ambassador of science, the bacterium of many flavors and the hero of this thesis, one should not forget that there are other species within genus Escherichia which can be as equally interesting and dangerous as E. coli, especially if they remain understudied.

In the end, I would like to remind you on what Pasteur said: “[Ladies and] Gentlemen, it is the microbes who will have the last word.”

V. Acknowledgements

My dear professor, together we embarked on this wondrous adventure by promising not to promise anything to each other. And so it began. Throughout this journey, you - Professor Bernt Eric Uhlin, have become nothing but my father in science, when I needed the inspiring world of science; my best friend, when I needed someone to lean on in times of trouble; my strongest and, yet, most gentle adversary when I had to learn how to grow and survive in science… And finally - you, Professor Uhlin, have been the greatest teacher I could have ever wished for! And when I pursue science in my next life, we shall meet again! Thank you!

It is said that in the darkest hour Sun will rise. I am so thankful to have you, Sun, as my co-supervisor! Your support and the way you perceive science devoid my days of any darkest hour! Thank you!

In Prof. Uhlin’s Lab, I have met not only unique bacteria, but also unique personalities who have become my true friends. I met you, Connie. Your vast personality is a true blessing. Thank you for letting me be your “small brother” and your friend! Patricia, you were my direct mentor in the beginning of this quest. Your kind soul and mind prepared me for all the challenges which were ahead of me! Thank you! Another friend that I met here is you, Monica! Thank you for letting me in your life and that you were always there for me! Almost in the end of my PhD studies, I met Prof. Rhen. I enjoy your spirit and scientific approach so much that I see us as brothers in arms!

I would like to express also my deepest gratitude to you, Annika! You have that special something in your personality that immediately made me become your friend. And I can honestly say that I have never seen such an unconditional friendship! I am truly honored! Thank you for that!

Another source of light that I found in this department is you, Svetle! Thank you for that and for all the nice adventures we had been through…

The world is so big and bacteria are so small… It is incredible how certain bacteria manage to find each other and form strong biofilm… That is a very rare event. It is the same with people, especially if they work with Listeria… Jorgen, your mind aligns in such a harmony with your heart that it is always a pleasure to present and discuss science with you! Karolis, thank you for your time and help when I needed it the most! You are a pure example of action speaks louder than words! Teresa – a beautiful girl with a beautiful mind and with a very big heart! Thank you for all the nice conversations and advice! I am amazed the way you perceive

life and I hope I got a little bit of your perception. Sabine, your presence here meant a lot to me. I miss you so much!

Teaching is important, but teaching can also be fun! Anders and Vicky, thank you for letting me teach in Molecular Genetics. I enjoyed every moment spent with you in that course! Teaching Molecular Genetics is fun, especially if you do it together with someone that is as crazy and weird as I am. Maybe that is why we became friends, what do you say, Hasan? Never change! From Molecular Genetics, I moved to Bacterial Physiology and Pathogenesis, where I met Lisa whose positive energy can recharge the whole department and sometimes we really need it! Thank you, Lisa! In the end of my PhD studies, I had to teach again, and again, I taught Bacterial Physiology and Pathogenesis. This time I had Matthew as a course leader. Matthew, whatever strong discussions we might have had, I know that you have a good heart and I really like you for that! Thank you for listening and thank you for caring! Salah, the last TA with whom I taught. It was an honor!

Marianella, the goddess of purge and care! Thank you that you have my back at the moments I misbehave and at those particular moments when I cannot be there for my students…

Maria, you are the first person I met here! Your solid support behind the scene of science is unbelievable! You are the guru of economy!

Johnny, I do not know what this department will do without you, honestly… I was in your shoes on several occasions and it was incredible what you deal with! Thank you for your miracles which solved my problems fast! Really fast!

Why is it that the best moments happen in the end of every story? Never mind, I am so happy that I met Teresa and Anna. Teresa, you bring color, happiness and enlightenment in a way that only a real muse can do it! And everyone needs at least one muse in their life… Anna, you came as a friend I had already had, I just didn’t realize it back then. It was a matter of time to meet. And we met! Thank you for your friendship!

Uli, my soulmate! You are my support in good, bad and very bad times! Thank you!

Saskia, your friendship is very rare! I am so honored and so pleased to have you besides me! Thank you!

In science, there is this special sub-population of remarkable beings whom you almost fall in love with. And then you realize that as long as science is done by the

people, for the people, these remarkable creatures become your main priority! Now it is time to thank my students… Jonas, Falk, Petteri, Crystal, Tiago, Michael, Saskia, Nils, Merhnoush, and Bart. I have no doubt in your success and I am absolutely sure that whatever you become, each of you will be a good one! Thank you for your time, efforts, help and unconditional devotion to science!

Мило мое семейство, благодаря ви, че винаги сте ми давали това, което имате и това, което нямате! Обичам ви!

This work was performed within the Laboratory for Molecular Infection Medicine Sweden (MIMS) and the Umeå Centre for Microbial Research (UCMR) Linnaeus Programme with support by grants from the Swedish Research Council (2015-03007, 2015-06824, 2016-06598, 349-2007-8673, 829-2006-7431).

VI. Bibliography:

1 Gould, S. J. in Full House: The Spread of Excellence from Plato to Darwin 175-192 ( 1996). 2 Schopf, J. W. Disparate rates, differing fates: tempo and mode of evolution changed from the Precambrian to the Phanerozoic. Proc Natl Acad Sci U S A 91, 6735-6742 (1994). 3 Escherich, T. Die Darmbakterien des Säuglings und ihre Beziehungen zur Physiologie der Verdauung. (Enke, 1886). 4 Castellani, A. & Chalmers, A. J. Manual of tropical medicine. 3d edn, (Baillière, Tindall and Cox, 1919). 5 Huxley, J. Evolution, the modern synthesis. (Harper & Bros., 1942). 6 Strockbine, F. S. a. N. Escherichia. 2nd edn, (Springer,, 2005). 7 Gordon, D. M. & Cowling, A. The distribution and genetic structure of Escherichia coli in Australian vertebrates: host and geographic effects. Microbiology 149, 3575-3586, doi:10.1099/mic.0.26486-0 (2003). 8 Wang, L., Rothemund, D., Curd, H. & Reeves, P. R. Species-wide variation in the Escherichia coli flagellin (H-antigen) gene. J Bacteriol 185, 2936-2943 (2003). 9 Low D., B., B., van der Woude, M. in Escherichia and Salmonella: Cellular and Molecular Biology (ASM Press, 1996). 10 Puorger, C., Vetsch, M., Wider, G. & Glockshuber, R. Structure, folding and stability of FimA, the main structural subunit of type 1 pili from uropathogenic Escherichia coli strains. J Mol Biol 412, 520-535, doi:10.1016/j.jmb.2011.07.044 (2011). 11 Lillington, J., Geibel, S. & Waksman, G. Biogenesis and adhesion of type 1 and P pili. Biochim Biophys Acta 1840, 2783-2793, doi:10.1016/j.bbagen.2014.04.021 (2014). 12 Geibel, S. & Waksman, G. The molecular dissection of the chaperone- usher pathway. Biochim Biophys Acta 1843, 1559-1567, doi:10.1016/j.bbamcr.2013.09.023 (2014). 13 Buchanan, K., Falkow, S., Hull, R. A. & Hull, S. I. Frequency among Enterobacteriaceae of the DNA sequences encoding type 1 pili. J Bacteriol 162, 799-803 (1985). 14 O'Hanley, P., Lark, D., Falkow, S. & Schoolnik, G. Molecular basis of Escherichia coli colonization of the upper urinary tract in BALB/c mice. Gal-Gal pili immunization prevents Escherichia coli pyelonephritis in the BALB/c mouse model of human pyelonephritis. J Clin Invest 75, 347- 360, doi:10.1172/JCI111707 (1985). 15 Dreux, N. et al. Point mutations in FimH adhesin of Crohn's disease- associated adherent-invasive Escherichia coli enhance intestinal inflammatory response. PLoS Pathog 9, e1003141, doi:10.1371/journal.ppat.1003141 (2013). 16 Vila, J. et al. Escherichia coli: an old friend with new tidings. FEMS Microbiol Rev 40, 437-463, doi:10.1093/femsre/fuw005 (2016). 17 Connell, I. et al. Type 1 fimbrial expression enhances Escherichia coli virulence for the urinary tract. Proc Natl Acad Sci U S A 93, 9827-9832 (1996).

18 Donnenberg, M. S. Escherichia coli : pathotypes and principles of pathogenesis. 2nd edition. edn, (Academic Press, 2013). 19 Uhlin, B. E. Regulation of E. coli Fimbrial Expression. (CRC Press, Inc., 1994). 20 Whitfield, C. & Keenleyside, W. J. Regulation of expression of group IA capsular polysaccharides in Escherichia coli and related extracellular polysaccharides in other bacteria. J Ind Microbiol 15, 361-371 (1995). 21 Whitfield, C. Structure and Assembly of Escherichia coli Capsules. EcoSal Plus 3, doi:10.1128/ecosalplus.4.7.3 (2009). 22 Drummelsmith, J. & Whitfield, C. Gene products required for surface expression of the capsular form of the group 1 K antigen in Escherichia coli (O9a:K30). Mol Microbiol 31, 1321-1332 (1999). 23 Cieslewicz, M. & Vimr, E. Reduced polysialic acid capsule expression in Escherichia coli K1 mutants with chromosomal defects in kpsF. Mol Microbiol 26, 237-249 (1997). 24 Roberts, I. S. The biochemistry and genetics of capsular polysaccharide production in bacteria. Annu Rev Microbiol 50, 285-315, doi:10.1146/annurev.micro.50.1.285 (1996). 25 Jann, B. & Jann, K. Structure and biosynthesis of the capsular antigens of Escherichia coli. Curr Top Microbiol Immunol 150, 19-42 (1990). 26 Neidhardt, F. C. & Curtiss, R. Escherichia coli and Salmonella : cellular and molecular biology. 2nd edn, (ASM Press, 1996). 27 Mitchell, A. et al. Adaptive prediction of environmental changes by microorganisms. Nature 460, 220-224, doi:10.1038/nature08112 (2009). 28 Makman, R. S. & Sutherland, E. W. Adenosine 3',5'-Phosphate in Escherichia coli. J Biol Chem 240, 1309-1314 (1965). 29 Escalante, A., Salinas Cervantes, A., Gosset, G. & Bolivar, F. Current knowledge of the Escherichia coli phosphoenolpyruvate-carbohydrate phosphotransferase system: peculiarities of regulation and impact on growth and product formation. Appl Microbiol Biotechnol 94, 1483- 1494, doi:10.1007/s00253-012-4101-5 (2012). 30 Gorke, B. & Stulke, J. Carbon catabolite repression in bacteria: many ways to make the most out of nutrients. Nat Rev Microbiol 6, 613-624, doi:10.1038/nrmicro1932 (2008). 31 Calvo, J. M. & Matthews, R. G. The leucine-responsive regulatory protein, a global regulator of metabolism in Escherichia coli. Microbiol Rev 58, 466-490 (1994). 32 Tani, T. H., Khodursky, A., Blumenthal, R. M., Brown, P. O. & Matthews, R. G. Adaptation to famine: a family of stationary-phase genes revealed by microarray analysis. Proc Natl Acad Sci U S A 99, 13471-13476, doi:10.1073/pnas.212510999 (2002). 33 Müller, C. M. et al. Type 1 fimbriae, a colonization factor of uropathogenic Escherichia coli, are controlled by the metabolic sensor CRP-cAMP. PLoS Pathog 5, e1000303, doi:10.1371/journal.ppat.1000303 (2009). 34 Kim, B. H. & Gadd, G. M. Bacterial physiology and metabolism. (Cambridge University Press, 2008).

35 Guest, J. R. Oxygen-regulated gene expression in Escherichia coli. The 1992 Marjory Stephenson Prize Lecture. J Gen Microbiol 138, 2253- 2263, doi:10.1099/00221287-138-11-2253 (1992). 36 Antonovsky, N. et al. Sugar Synthesis from CO2 in Escherichia coli. Cell 166, 115-125, doi:10.1016/j.cell.2016.05.064 (2016). 37 Koser, S. A. Correlation of Citrate Utilization by Members of the Colon- Aerogenes Group with Other Differential Characteristics and with Habitat. J Bacteriol 9, 59-77 (1924). 38 Lara, F. J. & Stokes, J. L. Oxidation of citrate by Escherichia coli. J Bacteriol 63, 415-420 (1952). 39 Lutgens, M. & Gottschalk, G. Why a co-substrate is required for anaerobic growth of Escherichia coli on citrate. J Gen Microbiol 119, 63- 70, doi:10.1099/00221287-119-1-63 (1980). 40 Crichton, R. R. Iron metabolism : from molecular mechanisms to clinical consequences. 3rd edn, (John Wiley & Sons, 2009). 41 Cartron, M. L., Maddocks, S., Gillingham, P., Craven, C. J. & Andrews, S. C. Feo--transport of ferrous iron into bacteria. Biometals 19, 143-157, doi:10.1007/s10534-006-0003-2 (2006). 42 Kammler, M., Schon, C. & Hantke, K. Characterization of the ferrous iron uptake system of Escherichia coli. J Bacteriol 175, 6212-6219 (1993). 43 Braun, V. Iron uptake by Escherichia coli. Front Biosci 8, s1409-1421 (2003). 44 Winterbourn, C. C. Toxicity of iron and hydrogen peroxide: the Fenton reaction. Toxicol Lett 82-83, 969-974 (1995). 45 Smith, J. L. The physiological role of ferritin-like compounds in bacteria. Crit Rev Microbiol 30, 173-185, doi:10.1080/10408410490435151 (2004). 46 Tenaillon, O., Skurnik, D., Picard, B. & Denamur, E. The population genetics of commensal Escherichia coli. Nat Rev Microbiol 8, 207-217, doi:10.1038/nrmicro2298 (2010). 47 Lawrence, J. G. & Ochman, H. Molecular archaeology of the Escherichia coli genome. Proc Natl Acad Sci U S A 95, 9413-9417 (1998). 48 Touchon, M. et al. Organised genome dynamics in the Escherichia coli species results in highly diverse adaptive paths. PLoS Genet 5, e1000344, doi:10.1371/journal.pgen.1000344 (2009). 49 Blattner, F. R. et al. The complete genome sequence of Escherichia coli K-12. Science 277, 1453-1462 (1997). 50 Ochman, H. & Davalos, L. M. The nature and dynamics of bacterial genomes. Science 311, 1730-1733, doi:10.1126/science.1119966 (2006). 51 Baba, T. et al. Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2, 2006 0008, doi:10.1038/msb4100050 (2006). 52 Ochman, H. & Wilson, A. C. Evolution in bacteria: evidence for a universal substitution rate in cellular genomes. J Mol Evol 26, 74-86 (1987). 53 Ochman, H. & Jones, I. B. Evolutionary dynamics of full genome content in Escherichia coli. EMBO J 19, 6637-6643, doi:10.1093/emboj/19.24.6637 (2000).

54 Johnson, T. J. & Nolan, L. K. Pathogenomics of the virulence plasmids of Escherichia coli. Microbiol Mol Biol Rev 73, 750-774, doi:10.1128/MMBR.00015-09 (2009). 55 Mathers, A. J., Peirano, G. & Pitout, J. D. Escherichia coli ST131: The quintessential example of an international multiresistant high-risk clone. Adv Appl Microbiol 90, 109-154, doi:10.1016/bs.aambs.2014.09.002 (2015). 56 Sato, G., Asagi, M., Oka, C., Ishiguro, N. & Terakado, N. Transmissible citrate-utilizing ability in Escherichia coli isolated from pigeons, pigs and cattle. Microbiol Immunol 22, 357-360 (1978). 57 Ishiguro, N. & Sato, G. The distribution of plasmids determining citrate utilization in citrate-positive variants of Escherichia coli from humans, domestic animals, feral birds and environments. J Hyg (Lond) 83, 331- 344 (1979). 58 Luo, Z. X., Yuan, C. X., Meng, Q. J. & Ji, Q. A Jurassic eutherian mammal and divergence of marsupials and placentals. Nature 476, 442-445, doi:10.1038/nature10291 (2011). 59 Szunerits, S. et al. Differentiation of Crohn's Disease-Associated Isolates from Other Pathogenic Escherichia coli by Fimbrial Adhesion under Shear Force. Biology (Basel) 5, doi:10.3390/biology5020014 (2016). 60 Croxen, M. A. & Finlay, B. B. Molecular mechanisms of Escherichia coli pathogenicity. Nat Rev Microbiol 8, 26-38, doi:10.1038/nrmicro2265 (2010). 61 Tozzoli, R., Sheutz, F. in Pathogenic Escherichia coli (ed S. Morabito) (Caister Academic Press, 2014). 62 Mainil, J. & Fairbrother, J. in Pathogenic Escherichia coli (ed S. Morabito) (Caister Academic Press, 2014). 63 Kaper, J. B., Nataro, J. P. & Mobley, H. L. Pathogenic Escherichia coli. Nat Rev Microbiol 2, 123-140, doi:10.1038/nrmicro818 (2004). 64 Nataro, J. P. & Kaper, J. B. Diarrheagenic Escherichia coli. Clin Microbiol Rev 11, 142-201 (1998). 65 Boudeau, J., Glasser, A. L., Masseret, E., Joly, B. & Darfeuille-Michaud, A. Invasive ability of an Escherichia coli strain isolated from the ileal mucosa of a patient with Crohn's disease. Infect Immun 67, 4499-4509 (1999). 66 Glasser, A. L. et al. Adherent invasive Escherichia coli strains from patients with Crohn's disease survive and replicate within macrophages without inducing host cell death. Infect Immun 69, 5529-5537 (2001). 67 Selander, R. K. et al. Methods of multilocus enzyme electrophoresis for bacterial population genetics and systematics. Appl Environ Microbiol 51, 873-884 (1986). 68 Goullet, P. & Picard, B. Comparative electrophoretic polymorphism of esterases and other enzymes in Escherichia coli. J Gen Microbiol 135, 135-143, doi:10.1099/00221287-135-1-135 (1989). 69 Sims, G. E. & Kim, S. H. Whole-genome phylogeny of Escherichia coli/Shigella group by feature frequency profiles (FFPs). Proc Natl Acad Sci U S A 108, 8329-8334, doi:10.1073/pnas.1105168108 (2011).

70 Dobrindt, U., Chowdary, M. G., Krumbholz, G. & Hacker, J. Genome dynamics and its impact on evolution of Escherichia coli. Med Microbiol Immunol 199, 145-154, doi:10.1007/s00430-010-0161-2 (2010). 71 Subashchandrabose, S. & Mobley, H. L. Virulence and Fitness Determinants of Uropathogenic Escherichia coli. Microbiol Spectr 3, doi:10.1128/microbiolspec.UTI-0015-2012 (2015). 72 Kucheria, R., Dasgupta, P., Sacks, S. H., Khan, M. S. & Sheerin, N. S. Urinary tract infections: new insights into a common problem. Postgrad Med J 81, 83-86, doi:10.1136/pgmj.2004.023036 (2005). 73 Warren, J. W. in Urinary tract infections: molecular pathogenesis and clinical management (ed H. L. Mobley) (ASM Press, Washington, D.C., 1996). 74 Foxman, B. et al. Uropathogenic Escherichia coli are more likely than commensal E. coli to be shared between heterosexual sex partners. Am J Epidemiol 156, 1133-1140 (2002). 75 Marrs, C. F., Zhang, L. & Foxman, B. Escherichia coli mediated urinary tract infections: are there distinct uropathogenic E. coli (UPEC) pathotypes? FEMS Microbiol Lett 252, 183-190, doi:10.1016/j.femsle.2005.08.028 (2005). 76 Agrawal, S. & Nadel, S. Acute bacterial meningitis in infants and children: epidemiology and management. Paediatr Drugs 13, 385-400, doi:10.2165/11593340-000000000-00000 (2011). 77 Gaschignard, J. et al. Neonatal Bacterial Meningitis: 444 Cases in 7 Years. Pediatr Infect Dis J 30, 212-217 (2011). 78 Saez-Llorens, X. & McCracken, G. H., Jr. Bacterial meningitis in children. Lancet 361, 2139-2148, doi:10.1016/S0140-6736(03)13693-8 (2003). 79 McCracken, G. H., Jr. et al. Relation between Escherichia coli K1 capsular polysaccharide antigen and clinical outcome in neonatal meningitis. Lancet 2, 246-250 (1974). 80 Stoll, B. J. et al. Changes in pathogens causing early-onset sepsis in very- low-birth-weight infants. N Engl J Med 347, 240-247, doi:10.1056/NEJMoa012657 (2002). 81 Barichello, T. et al. Neonatal Escherichia coli K1 meningitis causes learning and memory impairments in adulthood. J Neuroimmunol 272, 35-41, doi:10.1016/j.jneuroim.2014.05.003 (2014). 82 Mayor-Lynn, K. et al. Comparison of early-onset neonatal sepsis caused by Escherichia coli and group B Streptococcus. Am J Obstet Gynecol 192, 1437-1439, doi:10.1016/j.ajog.2004.12.031 (2005). 83 Silver, R. P., Aaronson, W. & Vann, W. F. The K1 capsular polysaccharide of Escherichia coli. Rev Infect Dis 10 Suppl 2, S282-286 (1988). 84 Jann, K. & Jann, B. Polysaccharide antigens of Escherichia coli. Rev Infect Dis 9 Suppl 5, S517-526 (1987). 85 Chen, B., Sysoeva, T. A., Chowdhury, S., Guo, L. & Nixon, B. T. ADPase activity of recombinantly expressed thermotolerant ATPases may be caused by copurification of adenylate kinase of Escherichia coli. FEBS J 276, 807-815, doi:10.1111/j.1742-4658.2008.06825.x (2009). 86 Robbins, J. B. et al. Escherichia coli K1 capsular polysaccharide associated with neonatal meningitis. N Engl J Med 290, 1216-1220, doi:10.1056/NEJM197405302902202 (1974).

87 Pluschke, G., Mayden, J., Achtman, M. & Levine, R. P. Role of the capsule and the O antigen in resistance of O18:K1 Escherichia coli to complement-mediated killing. Infect Immun 42, 907-913 (1983). 88 Pluschke, G. & Achtman, M. Degree of antibody-independent activation of the classical complement pathway by K1 Escherichia coli differs with O antigen type and correlates with virulence of meningitis in newborns. Infect Immun 43, 684-692 (1984). 89 Bortolussi, R., Ferrieri, P., Björkstén, B. & Quie, P. G. Capsular K1 polysaccharide of Escherichia coli: relationship to virulence in newborn rats and resistance to phagocytosis. Infect Immun 25, 293-298 (1979). 90 Bortolussi, R. & Ferrier, P. Protection against Escherichia coli K1 infection in newborn rats by antibody to K1 capsular polysaccharide antigen. Infect Immun 28, 111-117 (1980). 91 McCarthy, A. J., Stabler, R. A. & Taylor, P. W. Genome-Wide Identification by Transposon Insertion Sequencing of Escherichia coli K1 Genes Essential for In Vitro Growth, Gastrointestinal Colonizing Capacity, and Survival in Serum. J Bacteriol 200, doi:10.1128/JB.00698-17 (2018). 92 Silver, R. P. et al. Molecular cloning of the K1 capsular polysaccharide genes of E. coli. Nature 289, 696-698 (1981). 93 Silver, R. P., Vann, W. F. & Aaronson, W. Genetic and molecular analyses of Escherichia coli K1 antigen genes. J Bacteriol 157, 568-575 (1984). 94 Finne, J., Leinonen, M. & Mäkelä, P. H. Antigenic similarities between brain components and bacteria causing meningitis. Implications for vaccine development and pathogenesis. Lancet 2, 355-357 (1983). 95 Troy, F. A., 2nd. Polysialylation: from bacteria to brains. Glycobiology 2, 5-23 (1992). 96 Sarff, L. D. et al. Epidemiology of Escherichia coli K1 in healthy and diseased newborns. Lancet 1, 1099-1104 (1975). 97 Gao, Q. et al. Roles of iron acquisition systems in virulence of extraintestinal pathogenic Escherichia coli: salmochelin and aerobactin contribute more to virulence than heme in a chicken infection model. BMC Microbiol 12, 143, doi:10.1186/1471-2180-12-143 (2012). 98 Evans, D. J. et al. Hemolysin and K antigens in relation to serotype and hemagglutination type of Escherichia coli isolated from extraintestinal infections. J Clin Microbiol 13, 171-178 (1981). 99 Korhonen, T. K. et al. Serotypes, hemolysin production, and receptor recognition of Escherichia coli strains associated with neonatal sepsis and meningitis. Infect Immun 48, 486-491 (1985). 100 Hughes, C., Müller, D., Hacker, J. & Goebel, W. Genetics and pathogenic role of Escherichia coli haemolysin. Toxicon 20, 247-252 (1982). 101 Bhakdi, S. et al. The hemolysin of Escherichia coli. Eur J Epidemiol 4, 135-143 (1988). 102 Cross, M. A., Koronakis, V., Stanley, P. L. & Hughes, C. HlyB-dependent secretion of hemolysin by uropathogenic Escherichia coli requires conserved sequences flanking the chromosomal hly determinant. J Bacteriol 172, 1217-1224 (1990).

103 Lloyd, A. L., Rasko, D. A. & Mobley, H. L. Defining genomic islands and uropathogen-specific genes in uropathogenic Escherichia coli. J Bacteriol 189, 3532-3546, doi:10.1128/JB.01744-06 (2007). 104 Knapp, S., Hacker, J., Jarchau, T. & Goebel, W. Large, unstable inserts in the chromosome affect virulence properties of uropathogenic Escherichia coli O6 strain 536. J Bacteriol 168, 22-30 (1986). 105 Menestrina, G., Mackman, N., Holland, I. B. & Bhakdi, S. Escherichia coli haemolysin forms voltage-dependent ion channels in lipid membranes. Biochim Biophys Acta 905, 109-117 (1987). 106 Johnson, J. R. Virulence factors in Escherichia coli urinary tract infection. Clin Microbiol Rev 4, 80-128 (1991). 107 Dhakal, B. K. & Mulvey, M. A. The UPEC pore-forming toxin alpha- hemolysin triggers proteolysis of host proteins to disrupt cell adhesion, inflammatory, and survival pathways. Cell Host Microbe 11, 58-69, doi:10.1016/j.chom.2011.12.003 (2012). 108 Landraud, L., Gauthier, M., Fosse, T. & Boquet, P. Frequency of Escherichia coli strains producing the cytotoxic necrotizing factor (CNF1) in nosocomial urinary tract infections. Lett Appl Microbiol 30, 213-216 (2000). 109 Khan, N. A. et al. Cytotoxic necrotizing factor-1 contributes to Escherichia coli K1 invasion of the central nervous system. J Biol Chem 277, 15607-15612, doi:10.1074/jbc.M112224200 (2002). 110 Blum, G., Falbo, V., Caprioli, A. & Hacker, J. Gene clusters encoding the cytotoxic necrotizing factor type 1, Prs-fimbriae and alpha-hemolysin form the pathogenicity island II of the uropathogenic Escherichia coli strain J96. FEMS Microbiol Lett 126, 189-195 (1995). 111 Wang, M. H. & Kim, K. S. Cytotoxic necrotizing factor 1 contributes to Escherichia coli meningitis. Toxins (Basel) 5, 2270-2280, doi:10.3390/toxins5112270 (2013). 112 Landraud, L. et al. E. coli CNF1 toxin: a two-in-one system for host-cell invasion. Int J Med Microbiol 293, 513-518, doi:10.1078/1438-4221- 00295 (2004). 113 Ruiz-Perez, F. & Nataro, J. P. Bacterial serine proteases secreted by the autotransporter pathway: classification, specificity, and role in virulence. Cell Mol Life Sci 71, 745-770, doi:10.1007/s00018-013-1355-8 (2014). 114 Guyer, D. M., Henderson, I. R., Nataro, J. P. & Mobley, H. L. Identification of sat, an autotransporter toxin produced by uropathogenic Escherichia coli. Mol Microbiol 38, 53-66 (2000). 115 Guyer, D. M., Radulovic, S., Jones, F. E. & Mobley, H. L. Sat, the secreted autotransporter toxin of uropathogenic Escherichia coli, is a vacuolating cytotoxin for bladder and kidney epithelial cells. Infect Immun 70, 4539- 4546 (2002). 116 Maroncle, N. M., Sivick, K. E., Brady, R., Stokes, F. E. & Mobley, H. L. Protease activity, secretion, cell entry, cytotoxicity, and cellular targets of secreted autotransporter toxin of uropathogenic Escherichia coli. Infect Immun 74, 6124-6134, doi:10.1128/IAI.01086-06 (2006). 117 Parham, N. J. et al. PicU, a second serine protease autotransporter of uropathogenic Escherichia coli. FEMS Microbiol Lett 230, 73-83 (2004).

118 Heimer, S. R., Rasko, D. A., Lockatell, C. V., Johnson, D. E. & Mobley, H. L. Autotransporter genes pic and tsh are associated with Escherichia coli strains that cause acute pyelonephritis and are expressed during urinary tract infection. Infect Immun 72, 593-597 (2004). 119 Navarro-Garcia, F. et al. Pic, an autotransporter protein secreted by different pathogens in the Enterobacteriaceae family, is a potent mucus secretagogue. Infect Immun 78, 4101-4109, doi:10.1128/IAI.00523-10 (2010). 120 Provence, D. L. & Curtiss, R., 3rd. Isolation and characterization of a gene involved in hemagglutination by an avian pathogenic Escherichia coli strain. Infect Immun 62, 1369-1380 (1994). 121 Kostakioti, M. & Stathopoulos, C. Functional analysis of the Tsh autotransporter from an avian pathogenic Escherichia coli strain. Infect Immun 72, 5548-5554, doi:10.1128/IAI.72.10.5548-5554.2004 (2004). 122 Ayala-Lujan, J. L. et al. Broad spectrum activity of a lectin-like bacterial serine protease family on human leukocytes. PLoS One 9, e107920, doi:10.1371/journal.pone.0107920 (2014). 123 Tapader, R. et al. The high prevalence of serine protease autotransporters of Enterobacteriaceae (SPATEs) in Escherichia coli causing neonatal septicemia. Eur J Clin Microbiol Infect Dis 33, 2015-2024, doi:10.1007/s10096-014-2161-4 (2014). 124 Sik Kim, K. Meningitis-Associated Escherichia coli. EcoSal Plus 2, doi:10.1128/ecosalplus.8.6.1.2 (2006). 125 Mittal, R. & Prasadarao, N. V. gp96 expression in neutrophils is critical for the onset of Escherichia coli K1 (RS218) meningitis. Nat Commun 2, 552, doi:10.1038/ncomms1554 (2011). 126 Huang, S. H. et al. A novel genetic island of meningitic Escherichia coli K1 containing the ibeA invasion gene (GimA): functional annotation and carbon-source-regulated invasion of human brain microvascular endothelial cells. Funct Integr Genomics 1, 312-322, doi:10.1007/s101420100039 (2001). 127 Huang, S. H. et al. Escherichia coli invasion of brain microvascular endothelial cells in vitro and in vivo: molecular cloning and characterization of invasion gene ibe10. Infect Immun 63, 4470-4475 (1995). 128 Prasadarao, N. V., Wass, C. A., Huang, S. H. & Kim, K. S. Identification and characterization of a novel Ibe10 binding protein that contributes to Escherichia coli invasion of brain microvascular endothelial cells. Infect Immun 67, 1131-1138 (1999). 129 Bloch, C. A., Huang, S. H., Rode, C. K. & Kim, K. S. Mapping of noninvasion TnphoA mutations on the Escherichia coli O18:K1:H7 chromosome. FEMS Microbiol Lett 144, 171-176 (1996). 130 Vidotto, M. C., Queiroz, M. B., de Lima, N. C. & Gaziri, L. C. Prevalence of ibeA gene in avian pathogenic Escherichia coli (APEC). Vet Microbiol 119, 88-89, doi:10.1016/j.vetmic.2006.08.016 (2007). 131 Huang, S. H., Wan, Z. S., Chen, Y. H., Jong, A. Y. & Kim, K. S. Further characterization of Escherichia coli brain microvascular endothelial cell invasion gene ibeA by deletion, complementation, and protein expression. J Infect Dis 183, 1071-1078, doi:10.1086/319290 (2001).

132 Zou, Y., He, L. & Huang, S. H. Identification of a surface protein on human brain microvascular endothelial cells as vimentin interacting with Escherichia coli invasion protein IbeA. Biochem Biophys Res Commun 351, 625-630, doi:10.1016/j.bbrc.2006.10.091 (2006). 133 Maruvada, R. & Kim, K. S. IbeA and OmpA of Escherichia coli K1 exploit Rac1 activation for invasion of human brain microvascular endothelial cells. Infect Immun 80, 2035-2041, doi:10.1128/IAI.06320-11 (2012). 134 Huang, S. H. et al. Vimentin, a Novel NF-κB Regulator, Is Required for Meningitic Escherichia coli K1-Induced Pathogen Invasion and PMN Transmigration across the Blood-Brain Barrier. PLoS One 11, e0162641, doi:10.1371/journal.pone.0162641 (2016). 135 Chi, F., Bo, T., Wu, C. H., Jong, A. & Huang, S. H. Vimentin and PSF act in concert to regulate IbeA+ E. coli K1 induced activation and nuclear translocation of NF-κB in human brain endothelial cells. PLoS One 7, e35862, doi:10.1371/journal.pone.0035862 (2012). 136 Che, X. et al. Involvement of IbeA in meningitic Escherichia coli K1- induced polymorphonuclear leukocyte transmigration across brain endothelial cells. Brain Pathol 21, 389-404, doi:10.1111/j.1750- 3639.2010.00463.x (2011). 137 Prasadarao, N. V. et al. Outer membrane protein A of Escherichia coli contributes to invasion of brain microvascular endothelial cells. Infect Immun 64, 146-153 (1996). 138 Crosa, J. H. The relationship of plasmid-mediated iron transport and bacterial virulence. Annu Rev Microbiol 38, 69-89, doi:10.1146/annurev.mi.38.100184.000441 (1984). 139 Garénaux, A., Caza, M. & Dozois, C. M. The Ins and Outs of siderophore mediated iron uptake by extra-intestinal pathogenic Escherichia coli. Vet Microbiol 153, 89-98, doi:10.1016/j.vetmic.2011.05.023 (2011). 140 Montgomerie, J. Z., Bindereif, A., Neilands, J. B., Kalmanson, G. M. & Guze, L. B. Association of hydroxamate siderophore (aerobactin) with Escherichia coli isolated from patients with bacteremia. Infect Immun 46, 835-838 (1984). 141 Johnson, T. J., Siek, K. E., Johnson, S. J. & Nolan, L. K. DNA sequence of a ColV plasmid and prevalence of selected plasmid-encoded virulence genes among avian Escherichia coli strains. J Bacteriol 188, 745-758, doi:10.1128/JB.188.2.745-758.2006 (2006). 142 Johnson, T. J., Johnson, S. J. & Nolan, L. K. Complete DNA sequence of a ColBM plasmid from avian pathogenic Escherichia coli suggests that it evolved from closely related ColV virulence plasmids. J Bacteriol 188, 5975-5983, doi:10.1128/JB.00204-06 (2006). 143 Garcia, E. C., Brumbaugh, A. R. & Mobley, H. L. Redundancy and specificity of Escherichia coli iron acquisition systems during urinary tract infection. Infect Immun 79, 1225-1235, doi:10.1128/IAI.01222-10 (2011). 144 Hancock, V., Ferrieres, L. & Klemm, P. The ferric yersiniabactin uptake receptor FyuA is required for efficient biofilm formation by urinary tract infectious Escherichia coli in human urine. Microbiology 154, 167-175, doi:10.1099/mic.0.2007/011981-0 (2008).

145 Welch, R. A. et al. Extensive mosaic structure revealed by the complete genome sequence of uropathogenic Escherichia coli. Proc Natl Acad Sci U S A 99, 17020-17024, doi:10.1073/pnas.252529799 (2002). 146 Martinez, J. J., Mulvey, M. A., Schilling, J. D., Pinkner, J. S. & Hultgren, S. J. Type 1 pilus-mediated bacterial invasion of bladder epithelial cells. EMBO J 19, 2803-2812, doi:10.1093/emboj/19.12.2803 (2000). 147 Martinez, J. J. & Hultgren, S. J. Requirement of Rho-family GTPases in the invasion of type 1-piliated uropathogenic Escherichia coli. Cell Microbiol 4, 19-28 (2002). 148 Gunther, N. W. et al. Assessment of virulence of uropathogenic Escherichia coli type 1 fimbrial mutants in which the invertible element is phase-locked on or off. Infect Immun 70, 3344-3354 (2002). 149 Teng, C. H. et al. Escherichia coli K1 RS218 interacts with human brain microvascular endothelial cells via type 1 fimbria bacteria in the fimbriated state. Infect Immun 73, 2923-2931, doi:10.1128/IAI.73.5.2923-2931.2005 (2005). 150 Saukkonen, K. M., Nowicki, B. & Leinonen, M. Role of type 1 and S fimbriae in the pathogenesis of Escherichia coli O18:K1 bacteremia and meningitis in the infant rat. Infect Immun 56, 892-897 (1988). 151 Zhou, G. et al. Uroplakin Ia is the urothelial receptor for uropathogenic Escherichia coli: evidence from in vitro FimH binding. J Cell Sci 114, 4095-4103 (2001). 152 Brinton, C. C., Jr. Non-flagellar appendages of bacteria. Nature 183, 782-786 (1959). 153 Spurbeck, R. R. et al. Fimbrial profiles predict virulence of uropathogenic Escherichia coli strains: contribution of ygi and yad fimbriae. Infect Immun 79, 4753-4763, doi:10.1128/IAI.05621-11 (2011). 154 Marklund, B. I. et al. Horizontal gene transfer of the Escherichia coli pap and prs pili operons as a mechanism for the development of tissue- specific adhesive properties. Mol Microbiol 6, 2225-2242 (1992). 155 Hull, R. A., Gill, R. E., Hsu, P., Minshew, B. H. & Falkow, S. Construction and expression of recombinant plasmids encoding type 1 or D-mannose- resistant pili from a urinary tract infection Escherichia coli isolate. Infect Immun 33, 933-938 (1981). 156 Virkola, R. et al. Binding characteristics of Escherichia coli adhesins in human urinary bladder. Infect Immun 56, 2615-2622 (1988). 157 Mobley, H. L. et al. Isogenic P-fimbrial deletion mutants of pyelonephritogenic Escherichia coli: the role of alpha Gal(1-4) beta Gal binding in virulence of a wild-type strain. Mol Microbiol 10, 143-155 (1993). 158 Roberts, J. A. et al. The Gal(alpha 1-4)Gal-specific tip adhesin of Escherichia coli P-fimbriae is needed for pyelonephritis to occur in the normal urinary tract. Proc Natl Acad Sci U S A 91, 11889-11893 (1994). 159 Stins, M. F. et al. Binding characteristics of S fimbriated Escherichia coli to isolated brain microvascular endothelial cells. Am J Pathol 145, 1228- 1236 (1994). 160 Korhonen, T. K. et al. Escherichia coli fimbriae recognizing sialyl galactosides. J Bacteriol 159, 762-766 (1984).

161 Ott, M. et al. Analysis of the genetic determinants coding for the S- fimbrial adhesin (sfa) in different Escherichia coli strains causing meningitis or urinary tract infections. Infect Immun 54, 646-653 (1986). 162 Morschhäuser, J., Vetter, V., Emödy, L. & Hacker, J. Adhesin regulatory genes within large, unstable DNA regions of pathogenic Escherichia coli: cross-talk between different adhesin gene clusters. Mol Microbiol 11, 555-566 (1994). 163 Korhonen, T. K. et al. Binding of Escherichia coli S fimbriae to human kidney epithelium. Infect Immun 54, 322-327 (1986). 164 Korhonen, T. K. et al. Serotypes, hemolysin production, and receptor recognition of Escherichia coli strains associated with neonatal sepsis and meningitis. Infect Immun 48, 486-491 (1985). 165 Parkkinen, J., Korhonen, T. K., Pere, A., Hacker, J. & Soinila, S. Binding sites in the rat brain for Escherichia coli S fimbriae associated with neonatal meningitis. J Clin Invest 81, 860-865, doi:10.1172/JCI113395 (1988). 166 Parkkinen, J., Ristimaki, A. & Westerlund, B. Binding of Escherichia coli S fimbriae to cultured human endothelial cells. Infect Immun 57, 2256- 2259 (1989). 167 Prasadarao, N. V., Wass, C. A., Hacker, J., Jann, K. & Kim, K. S. Adhesion of S-fimbriated Escherichia coli to brain glycolipids mediated by sfaA gene-encoded protein of S-fimbriae. J Biol Chem 268, 10356-10363 (1993). 168 Abraham, J. M., Freitag, C.S., Clements, J.R., Eisenstein, B.I. An invertible element of DNA controls phase variation of type 1 fimbriae of Escherichia coli. Proc Natl Acad Sci U S A 82, 5724-5727 (1985). 169 Klemm, P. Two regulatory fim genes, fimB and fimE, control the phase variation of type 1 fimbriae in Escherichia coli. EMBO J 5, 1389-1393 (1986). 170 Bryan, A. et al. Regulation of type 1 fimbriae by unlinked FimB- and FimE-like recombinases in uropathogenic Escherichia coli strain CFT073. Infect Immun 74, 1072-1083, doi:10.1128/IAI.74.2.1072- 1083.2006 (2006). 171 Xie, Y., Yao, Y., Kolisnychenko, V., Teng, C. H. & Kim, K. S. HbiF regulates type 1 fimbriation independently of FimB and FimE. Infect Immun 74, 4039-4047, doi:10.1128/IAI.02058-05 (2006). 172 Lindberg, S. et al. Regulatory Interactions among adhesin gene systems of uropathogenic Escherichia coli. Infect Immun 76, 771-780, doi:10.1128/IAI.01010-07 (2008). 173 Ott, M., Hoschützky, H., Jann, K., Van Die, I. & Hacker, J. Gene clusters for S fimbrial adhesin (sfa) and F1C fimbriae (foc) of Escherichia coli: comparative aspects of structure and function. J Bacteriol 170, 3983- 3990 (1988). 174 van der Woude, M., Braaten, B. & Low, D. Epigenetic phase variation of the pap operon in Escherichia coli. Trends Microbiol 4, 5-9 (1996). 175 Peterson, S. N. & Reich, N. O. Competitive Lrp and Dam assembly at the pap regulatory region: implications for mechanisms of epigenetic regulation. J Mol Biol 383, 92-105, doi:10.1016/j.jmb.2008.07.086 (2008).

176 Hale, W. B., van der Woude, M. W., Braaten, B. A. & Low, D. A. Regulation of uropathogenic Escherichia coli adhesin expression by DNA methylation. Mol Genet Metab 65, 191-196, doi:10.1006/mgme.1998.2744 (1998). 177 Ott, M. & Hacker, J. Analysis of the variability of S-fimbriae expression in an Escherichia coli pathogen. FEMS Microbiol Lett 63, 233-238 (1991). 178 Morschhäuser, J., Uhlin, B. E. & Hacker, J. Transcriptional analysis and regulation of the sfa determinant coding for S fimbriae of pathogenic Escherichia coli strains. Mol Gen Genet 238, 97-105 (1993). 179 Xia, Y., Gally, D., Forsman-Semb, K. & Uhlin, B. E. Regulatory cross-talk between adhesin operons in Escherichia coli: inhibition of type 1 fimbriae expression by the PapB protein. EMBO J 19, 1450-1457, doi:10.1093/emboj/19.7.1450 (2000). 180 Lindberg, S. et al. Regulatory Interactions among adhesin gene systems of uropathogenic Escherichia coli. Infect Immun 76, 771-780, doi:10.1128/IAI.01010-07 (2008). 181 White-Ziegler, C. A., Villapakkam, A., Ronaszeki, K. & Young, S. H-NS controls pap and daa fimbrial transcription in Escherichia coli in response to multiple environmental cues. J Bacteriol 182, 6391-6400 (2000). 182 Gally, D. L., Rucker, T. J. & Blomfield, I. C. The leucine-responsive regulatory protein binds to the fim switch to control phase variation of type 1 fimbrial expression in Escherichia coli K-12. J Bacteriol 176, 5665- 5672 (1994). 183 Schmoll, T., Ott, M., Oudega, B. & Hacker, J. Use of a wild-type gene fusion to determine the influence of environmental conditions on expression of the S fimbrial adhesin in an Escherichia coli pathogen. J Bacteriol 172, 5103-5111 (1990). 184 Göransson, M., Forsman, K. & Uhlin, B. E. Regulatory genes in the thermoregulation of Escherichia coli pili gene transcription. Genes Dev 3, 123-130 (1989). 185 Sokurenko, E. Pathoadaptive Mutations in Uropathogenic Escherichia coli. Microbiol Spectr 4, doi:10.1128/microbiolspec.UTI-0020-2015 (2016). 186 Adiba, S., Nizak, C., van Baalen, M., Denamur, E. & Depaulis, F. From grazing resistance to pathogenesis: the coincidental evolution of virulence factors. PLoS One 5, e11882, doi:10.1371/journal.pone.0011882 (2010). 187 Alsam, S. et al. Escherichia coli interactions with Acanthamoeba: a symbiosis with environmental and clinical implications. J Med Microbiol 55, 689-694, doi:10.1099/jmm.0.46497-0 (2006). 188 Wildschutte, H., Wolfe, D. M., Tamewitz, A. & Lawrence, J. G. Protozoan predation, diversifying selection, and the evolution of antigenic diversity in Salmonella. Proc Natl Acad Sci U S A 101, 10644-10649, doi:10.1073/pnas.0404028101 (2004). 189 Ewers, C. et al. Avian pathogenic, uropathogenic, and newborn meningitis-causing Escherichia coli: how closely related are they? Int J Med Microbiol 297, 163-176, doi:10.1016/j.ijmm.2007.01.003 (2007).

190 Chabe, M., Lokmer, A. & Segurel, L. Gut Protozoa: Friends or Foes of the Human Gut Microbiota? Trends Parasitol 33, 925-934, doi:10.1016/j.pt.2017.08.005 (2017). 191 Samie, A. et al. Prevalence and species distribution of E. histolytica and E. dispar in the Venda region, Limpopo, South Africa. Am J Trop Med Hyg 75, 565-571 (2006). 192 Matin, A. & Jung, S. Y. Interaction of Escherichia coli K1 and K5 with Acanthamoeba castellanii trophozoites and cysts. Korean J Parasitol 49, 349-356, doi:10.3347/kjp.2011.49.4.349 (2011). 193 Jung, S. Y., Matin, A., Kim, K. S. & Khan, N. A. The capsule plays an important role in Escherichia coli K1 interactions with Acanthamoeba. Int J Parasitol 37, 417-423, doi:10.1016/j.ijpara.2006.10.012 (2007). 194 Scholl, D., Adhya, S. & Merril, C. Escherichia coli K1's capsule is a barrier to bacteriophage T7. Appl Environ Microbiol 71, 4872-4874, doi:10.1128/AEM.71.8.4872-4874.2005 (2005). 195 Hacker, J. & Kaper, J. B. Pathogenicity islands and the evolution of microbes. Annu Rev Microbiol 54, 641-679, doi:10.1146/annurev.micro.54.1.641 (2000). 196 Hacker, J. & Carniel, E. Ecological fitness, genomic islands and bacterial pathogenicity. A Darwinian view of the evolution of microbes. EMBO Rep 2, 376-381, doi:10.1093/embo-reports/kve097 (2001). 197 Blum, G. et al. Excision of large DNA regions termed pathogenicity islands from tRNA-specific loci in the chromosome of an Escherichia coli wild-type pathogen. Infect Immun 62, 606-614 (1994). 198 Dobrindt, U. et al. S-Fimbria-encoding determinant sfa(I) is located on pathogenicity island III(536) of uropathogenic Escherichia coli strain 536. Infect Immun 69, 4248-4256, doi:10.1128/IAI.69.7.4248- 4256.2001 (2001). 199 Middendorf, B. et al. The pathogenicity islands (PAIs) of the uropathogenic Escherichia coli strain 536: island probing of PAI II536. J Infect Dis 183 Suppl 1, S17-20, doi:10.1086/318843 (2001). 200 Sjöström, A. E. et al. Analysis of the sfaX(II) locus in the Escherichia coli meningitis isolate IHE3034 reveals two novel regulatory genes within the promoter-distal region of the main S fimbrial operon. Microb Pathog 46, 150-158, doi:10.1016/j.micpath.2008.12.001 (2009). 201 Xie, Y. et al. Identification and characterization of Escherichia coli RS218-derived islands in the pathogenesis of E. coli meningitis. J Infect Dis 194, 358-364, doi:10.1086/505429 (2006). 202 Bliven, K. A. & Maurelli, A. T. Antivirulence genes: insights into pathogen evolution through gene loss. Infect Immun 80, 4061-4070, doi:10.1128/IAI.00740-12 (2012). 203 Casalino, M., Latella, M. C., Prosseda, G. & Colonna, B. CadC is the preferential target of a convergent evolution driving enteroinvasive Escherichia coli toward a lysine decarboxylase-defective phenotype. Infect Immun 71, 5472-5479 (2003). 204 Day, W. A., Jr., Fernandez, R. E. & Maurelli, A. T. Pathoadaptive mutations that enhance virulence: genetic organization of the cadA regions of Shigella spp. Infect Immun 69, 7471-7480, doi:10.1128/IAI.69.12.7471-7480.2001 (2001).

205 Maurelli, A. T., Fernandez, R. E., Bloch, C. A., Rode, C. K. & Fasano, A. "Black holes" and bacterial pathogenicity: a large genomic deletion that enhances the virulence of Shigella spp. and enteroinvasive Escherichia coli. Proc Natl Acad Sci U S A 95, 3943-3948 (1998). 206 Samartzidou, H., Mehrazin, M., Xu, Z., Benedik, M. J. & Delcour, A. H. Cadaverine inhibition of porin plays a role in cell survival at acidic pH. J Bacteriol 185, 13-19 (2003). 207 Fernandez, I. M. et al. Cadaverine prevents the escape of Shigella flexneri from the phagolysosome: a connection between bacterial dissemination and neutrophil transepithelial signaling. J Infect Dis 184, 743-753, doi:10.1086/323035 (2001). 208 McCormick, B. A., Fernandez, M. I., Siber, A. M. & Maurelli, A. T. Inhibition of Shigella flexneri-induced transepithelial migration of polymorphonuclear leucocytes by cadaverine. Cell Microbiol 1, 143-155 (1999). 209 Torres, A. G., Vazquez-Juarez, R. C., Tutt, C. B. & Garcia-Gallegos, J. G. Pathoadaptive mutation that mediates adherence of shiga toxin- producing Escherichia coli O111. Infect Immun 73, 4766-4776, doi:10.1128/IAI.73.8.4766-4776.2005 (2005). 210 Vazquez-Juarez, R. C. et al. CadA negatively regulates Escherichia coli O157:H7 adherence and intestinal colonization. Infect Immun 76, 5072- 5081, doi:10.1128/IAI.00677-08 (2008). 211 Oelschlaeger, T. A., Dobrindt, U. & Hacker, J. Pathogenicity islands of uropathogenic E. coli and the evolution of virulence. Int J Antimicrob Agents 19, 517-521 (2002). 212 Sokurenko, E. V., Courtney, H. S., Maslow, J., Siitonen, A. & Hasty, D. L. Quantitative differences in adhesiveness of type 1 fimbriated Escherichia coli due to structural differences in fimH genes. J Bacteriol 177, 3680- 3686 (1995). 213 Sokurenko, E. V. et al. Pathogenic adaptation of Escherichia coli by natural variation of the FimH adhesin. Proc Natl Acad Sci U S A 95, 8922-8926 (1998). 214 Kurtz, I. Molecular mechanisms and regulation of urinary acidification. Compr Physiol 4, 1737-1774, doi:10.1002/cphy.c140021 (2014). 215 Plough, I. C. & Baker, E. M. Maximum physiological concentration of sodium in human urine. J Appl Physiol 14, 1036-1038, doi:10.1152/jappl.1959.14.6.1036 (1959). 216 Giannakopoulos, X. et al. Human bladder urine oxygen content: implications for urinary tract diseases. Int Urol Nephrol 29, 393-401 (1997). 217 Gupta, R. C., Goyal, A., Ghosh, R., Punjabi, M. & Singh, P. P. Normal range for glucose in urine: age-related changes. Clin Chem 28, 2335 (1982). 218 Jain, S. Reference interval for urinary glucose in elderly subjects. Clin Chem 32, 711-712 (1986). 219 Edin-Liljegren, A., Rodin, L., Grenabo, L. & Hedelin, H. The importance of glucose for the Escherichia coli mediated citrate depletion in synthetic and human urine. Scand J Urol Nephrol 35, 106-111 (2001).

220 Brooks, T. & Keevil, C. W. A simple artificial urine for the growth of urinary pathogens. Lett Appl Microbiol 24, 203-206 (1997). 221 Alteri, C. J., Smith, S. N. & Mobley, H. L. Fitness of Escherichia coli during urinary tract infection requires gluconeogenesis and the TCA cycle. PLoS Pathog 5, e1000448, doi:10.1371/journal.ppat.1000448 (2009). 222 Snyder, J. A. et al. Transcriptome of uropathogenic Escherichia coli during urinary tract infection. Infect Immun 72, 6373-6381, doi:10.1128/IAI.72.11.6373-6381.2004 (2004). 223 Römling, U., Gomelsky, M. & Galperin, M. Y. C-di-GMP: the dawning of a novel bacterial signalling system. Mol Microbiol 57, 629-639, doi:10.1111/j.1365-2958.2005.04697.x (2005). 224 Simm, R., Morr, M., Kader, A., Nimtz, M. & Römling, U. GGDEF and EAL domains inversely regulate cyclic di-GMP levels and transition from sessility to motility. Mol Microbiol 53, 1123-1134, doi:10.1111/j.1365- 2958.2004.04206.x (2004). 225 Hengge, R. Principles of c-di-GMP signalling in bacteria. Nat Rev Microbiol 7, 263-273, doi:10.1038/nrmicro2109 (2009). 226 Schmidt, A. J., Ryjenkov, D. A. & Gomelsky, M. The ubiquitous protein domain EAL is a cyclic diguanylate-specific phosphodiesterase: enzymatically active and inactive EAL domains. J Bacteriol 187, 4774- 4781, doi:10.1128/JB.187.14.4774-4781.2005 (2005). 227 Ausmees, N. et al. Genetic data indicate that proteins containing the GGDEF domain possess diguanylate cyclase activity. FEMS Microbiol Lett 204, 163-167 (2001). 228 Amikam, D. & Galperin, M. Y. PilZ domain is part of the bacterial c-di- GMP binding protein. Bioinformatics 22, 3-6, doi:10.1093/bioinformatics/bti739 (2006). 229 Povolotsky, T. L. & Hengge, R. Genome-Based Comparison of Cyclic Di- GMP Signaling in Pathogenic and Commensal Escherichia coli Strains. J Bacteriol 198, 111-126, doi:10.1128/JB.00520-15 (2015). 230 Ryjenkov, D. A., Simm, R., Römling, U. & Gomelsky, M. The PilZ domain is a receptor for the second messenger c-di-GMP: the PilZ domain protein YcgR controls motility in enterobacteria. J Biol Chem 281, 30310-30314, doi:10.1074/jbc.C600179200 (2006). 231 Pesavento, C. et al. Inverse regulatory coordination of motility and curli- mediated adhesion in Escherichia coli. Genes Dev 22, 2434-2446, doi:10.1101/gad.475808 (2008). 232 Boehm, A. et al. Second messenger-mediated adjustment of bacterial swimming velocity. Cell 141, 107-116, doi:10.1016/j.cell.2010.01.018 (2010). 233 Boehm, A. et al. Second messenger signalling governs Escherichia coli biofilm induction upon ribosomal stress. Mol Microbiol 72, 1500-1516, doi:10.1111/j.1365-2958.2009.06739.x (2009). 234 Crepin, S., Porcheron, G., Houle, S., Harel, J. & Dozois, C. M. Altered Regulation of the Diguanylate Cyclase YaiC Reduces Production of Type 1 Fimbriae in a Pst Mutant of Uropathogenic Escherichia coli CFT073. J Bacteriol 199, doi:10.1128/JB.00168-17 (2017).

235 Spurbeck, R. R., Tarrien, R. J. & Mobley, H. L. Enzymatically active and inactive phosphodiesterases and diguanylate cyclases are involved in regulation of Motility or sessility in Escherichia coli CFT073. MBio 3, doi:10.1128/mBio.00307-12 (2012). 236 Spurbeck, R. R., Alteri, C. J., Himpsl, S. D. & Mobley, H. L. The multifunctional protein YdiV represses P fimbria-mediated adherence in uropathogenic Escherichia coli. J Bacteriol 195, 3156-3164, doi:10.1128/JB.02254-12 (2013). 237 Enow, C. O. et al. Elevated recombinant clyA gene expression in the uropathogenic Escherichia coli strain 536, a clue to explain pathoadaptive mutations in a subset of extraintestinal E. coli strains. BMC Microbiol 14, 216, doi:10.1186/s12866-014-0216-4 (2014). 238 Paracuellos, P., Ohman, A., Sauer-Eriksson, A. E. & Uhlin, B. E. Expression and purification of SfaX(II), a protein involved in regulating adhesion and motility genes in extraintestinal pathogenic Escherichia coli. Protein Expr Purif 86, 127-134, doi:10.1016/j.pep.2012.09.006 (2012). 239 Saridakis, V., Shahinas, D., Xu, X. & Christendat, D. Structural insight on the mechanism of regulation of the MarR family of proteins: high- resolution crystal structure of a transcriptional repressor from Methanobacterium thermoautotrophicum. J Mol Biol 377, 655-667, doi:10.1016/j.jmb.2008.01.001 (2008). 240 Sjöström, A. E. et al. The SfaXII protein from newborn meningitis E. coli is involved in regulation of motility and type 1 fimbriae expression. Microb Pathog 46, 243-252, doi:10.1016/j.micpath.2009.01.007 (2009). 241 Simms, A. N. & Mobley, H. L. PapX, a P fimbrial operon-encoded inhibitor of motility in uropathogenic Escherichia coli. Infect Immun 76, 4833-4841, doi:10.1128/IAI.00630-08 (2008). 242 Reiss, D. J. & Mobley, H. L. Determination of target sequence bound by PapX, repressor of bacterial motility, in flhD promoter using systematic evolution of ligands by exponential enrichment (SELEX) and high throughput sequencing. J Biol Chem 286, 44726-44738, doi:10.1074/jbc.M111.290684 (2011). 243 Justice, S. S., Hunstad, D. A., Seed, P. C. & Hultgren, S. J. Filamentation by Escherichia coli subverts innate defenses during urinary tract infection. Proc Natl Acad Sci U S A 103, 19884-19889, doi:10.1073/pnas.0606329104 (2006). 244 Justice, S. S., Hunstad, D. A., Cegelski, L. & Hultgren, S. J. Morphological plasticity as a bacterial survival strategy. Nat Rev Microbiol 6, 162-168, doi:10.1038/nrmicro1820 (2008). 245 Garofalo, C. K. et al. Escherichia coli from urine of female patients with urinary tract infections is competent for intracellular bacterial community formation. Infect Immun 75, 52-60, doi:10.1128/IAI.01123- 06 (2007). 246 Khandige, S. et al. DamX Controls Reversible Cell Morphology Switching in Uropathogenic Escherichia coli. MBio 7, doi:10.1128/mBio.00642-16 (2016).

247 Weber, H., Pesavento, C., Possling, A., Tischendorf, G. & Hengge, R. Cyclic-di-GMP-mediated signalling within the sigma network of Escherichia coli. Mol Microbiol 62, 1014-1034, doi:10.1111/j.1365- 2958.2006.05440.x (2006). 248 Notley-McRobb, L., King, T. & Ferenci, T. rpoS mutations and loss of general stress resistance in Escherichia coli populations as a consequence of conflict between competing stress responses. J Bacteriol 184, 806-811 (2002). 249 Moriel, D. G. et al. Identification of protective and broadly conserved vaccine antigens from the genome of extraintestinal pathogenic Escherichia coli. Proc Natl Acad Sci U S A 107, 9072-9077, doi:10.1073/pnas.0915077107 (2010). 250 Silver, R. P., Aaronson, W., Sutton, A. & Schneerson, R. Comparative analysis of plasmids and some metabolic characteristics of Escherichia coli K1 from diseased and healthy individuals. Infect Immun 29, 200- 206 (1980). 251 Wang, Y. & Kim, K. S. Effect of rpoS mutations on stress-resistance and invasion of brain microvascular endothelial cells in Escherichia coli K1. FEMS Microbiol Lett 182, 241-247 (2000). 252 Schellhorn, H. E. & Hassan, H. M. Transcriptional regulation of katE in Escherichia coli K-12. J Bacteriol 170, 4286-4292 (1988). 253 Lange, R. & Hengge-Aronis, R. Identification of a central regulator of stationary-phase gene expression in Escherichia coli. Mol Microbiol 5, 49-59, doi:10.1111/j.1365-2958.1991.tb01825.x (1991). 254 Battesti, A., Majdalani, N. & Gottesman, S. The RpoS-mediated general stress response in Escherichia coli. Annu Rev Microbiol 65, 189-213, doi:10.1146/annurev-micro-090110-102946 (2011). 255 Hengge-Aronis, R. Signal transduction and regulatory mechanisms involved in control of the sigma(S) (RpoS) subunit of RNA polymerase. Microbiol Mol Biol Rev 66, 373-395, table of contents (2002). 256 Adams, J. L. & McLean, R. J. Impact of rpoS deletion on Escherichia coli biofilms. Appl Environ Microbiol 65, 4285-4287 (1999). 257 Hengge, R. et al. Systematic Nomenclature for GGDEF and EAL Domain- Containing Cyclic Di-GMP Turnover Proteins of Escherichia coli. J Bacteriol 198, 7-11, doi:10.1128/JB.00424-15 (2016). 258 Reinders, A. et al. Expression and Genetic Activation of Cyclic Di-GMP- Specific Phosphodiesterases in Escherichia coli. J Bacteriol 198, 448- 462, doi:10.1128/JB.00604-15 (2015). 259 Sarenko, O. et al. More than Enzymes That Make or Break Cyclic Di- GMP-Local Signaling in the Interactome of GGDEF/EAL Domain Proteins of Escherichia coli. MBio 8, doi:10.1128/mBio.01639-17 (2017). 260 Simmons, J. A Culture Medium for Differentiating Organisms of Typhoid-Colon Aerogenes Groups and for Isolation of Certain Fungi: With Colored Plate. The Journal of Infectious Diseases Volume 39, 209–214 (1926). 261 del Castillo, F. J., Leal, S. C., Moreno, F. & del Castillo, I. The Escherichia coli K-12 sheA gene encodes a 34-kDa secreted haemolysin. Mol Microbiol 25, 107-115 (1997).

262 Oscarsson, J. et al. Molecular analysis of the cytolytic protein ClyA (SheA) from Escherichia coli. Mol Microbiol 32, 1226-1238 (1999). 263 Westermark, M., Oscarsson, J., Mizunoe, Y., Urbonaviciene, J. & Uhlin, B. E. Silencing and activation of ClyA cytotoxin expression in Escherichia coli. J Bacteriol 182, 6347-6357 (2000). 264 Simms, A. N. & Mobley, H. L. Multiple genes repress motility in uropathogenic Escherichia coli constitutively expressing type 1 fimbriae. J Bacteriol 190, 3747-3756, doi:10.1128/JB.01870-07 (2008). 265 Goebel, W. & Hedgpeth, J. Cloning and functional characterization of the plasmid-encoded hemolysin determinant of Escherichia coli. J Bacteriol 151, 1290-1298 (1982). 266 Rothfield, L., Justice, S. & Garcia-Lara, J. Bacterial cell division. Annu Rev Genet 33, 423-448, doi:10.1146/annurev.genet.33.1.423 (1999). 267 Harry, E., Monahan, L. & Thompson, L. Bacterial cell division: the mechanism and its precison. Int Rev Cytol 253, 27-94, doi:10.1016/S0074-7696(06)53002-5 (2006). 268 Erill, I., Campoy, S. & Barbe, J. Aeons of distress: an evolutionary perspective on the bacterial SOS response. FEMS Microbiol Rev 31, 637- 656, doi:10.1111/j.1574-6976.2007.00082.x (2007). 269 Kreuzer, K. N. DNA damage responses in prokaryotes: regulating gene expression, modulating growth patterns, and manipulating replication forks. Cold Spring Harb Perspect Biol 5, a012674, doi:10.1101/cshperspect.a012674 (2013). 270 Zgur-Bertok, D. DNA damage repair and bacterial pathogens. PLoS Pathog 9, e1003711, doi:10.1371/journal.ppat.1003711 (2013). 271 Hanawalt, P. C. Historical perspective on the DNA damage response. DNA Repair (Amst) 36, 2-7, doi:10.1016/j.dnarep.2015.10.001 (2015). 272 Schoemaker, J. M., Gayda, R. C. & Markovitz, A. Regulation of cell division in Escherichia coli: SOS induction and cellular location of the sulA protein, a key to lon-associated filamentation and death. J Bacteriol 158, 551-561 (1984). 273 Lopez-Garrido, J. & Casadesus, J. The DamX protein of Escherichia coli and Salmonella enterica. Gut Microbes 1, 285-288, doi:10.4161/gmic.1.4.12079 (2010). 274 Arends, S. J. et al. Discovery and characterization of three new Escherichia coli septal ring proteins that contain a SPOR domain: DamX, DedD, and RlpA. J Bacteriol 192, 242-255, doi:10.1128/JB.01244-09 (2010).