sign in sign up
<<
Home , Quinolone antibiotic

UNIVERSITY OF COPENH A G E N DEPARTMENT OF BIOLOG Y

PhD thesis Rasmus Nielsen Klitgaard, M.Sc.

Antibiotic Discovery Potentiation of the quinolones and targeting the initiation of DNA replication

This thesis has been submitted to the PhD School of The Faculty of Science, University of Copenhagen, Denmark, 28. February 2018.

Academic advisor Anders Løbner-Olesen Department of Biology Functional Genomics University of Copenhagen, Denmark.

Assessment committee Dan Andersson Department of Medical Biochemistry and Microbiology University of Uppsala, Sweden.

Mogens Kilstrup Department of Biochemistry and Biomedicine Metabolic Signaling and Regulation Danish Technical University, Denmark.

Signe Lo Svenningsen Department of Biology Biomolecular Sciences University of Copenhagen, Denmark.

Submitted: 28.02.2018

1 Acknowledgements

First, I would like to thank my supervisor Anders Løbner-Olesen for his excellent support and guidance throughout my PhD. I have highly appreciated that Anders has been available more or less every day and gladly discussed any questions I might have had.

I would also like to thank Godefroid Charbon, not only for our collaboration on paper II presented in this thesis, but also for always taking time to discuss and give advice on my other projects. It has been greatly cherished.

Furthermore, I would like to thank the staff at Naicons srl. and all of the people who have been part of the ALO lab: Thomas T. Thomsen, Jakob Frimodt-Møller, Maria S. Haugan, Christoffer Campion, Anna E. Ebbensgard, Michaela Lederer, Henrik Jakobsen, Leise Riber and Belén M. Chamizo.

A thanks, should also be given to my bachelor student, Anne Kristine Schack, who contributed to the construction of the screening system presented in paper III.

Finally, I would like to thank my girlfriend, Marie, my family and my friends for their great support and interest in my work.

2 Abstract – English

Antibiotic resistance has been deemed as one of the biggest threats to the global public health by the World health Organization. In 2050, an estimated 10 million deaths per year will be attributed to resistance, thus proper action needs to be taken to stop this negative development. An important mean in the arms race against antibiotic resistance is the discovery and development of novel , but also preserving the efficacy of the antibiotics that are already in clinical use.

In paper I, we search for helper drug targets in an effort to preserve the use of this widely applied antibiotic. Using a combined genetic and transcriptomic approach, the AcrAB-TolC pump and the SOS response genes, RecA and RecC, are identified as potential targets for helper in strains with low-level ciprofloxacin resistance. In addition, our results also indicate that reversing high-level ciprofloxacin resistance is likely not plausible.

In paper II, we present two novel cell based screens for identifying inhibitors of the chromosomal DNA replication initiation in bacteria. The screens are based on growth rescue of cells that rigorously over-initiate the DNA replication, due to either increased regeneration of the active ATP bound form of the replication initiator protein DnaA, or by being deficient in the process known as regulatory inactivation of DnaA (RIDA). Screening a library of 400 microbial extracts, revealed the iron chelator deferoxamine as a compound that rescues the growth of over-initiating cells. Albeit not by decreasing the replication initiation frequency, but by reducing the production of reactive oxygen species. Substantiating the model that oxidative DNA damage and its repair promotes the lethal action of hyper-replication.

In paper III, we constructed and verified a novel high throughput, cell based, fluorescence screen for inhibitors of chromosome replication initiation in bacteria. The screen utilizes an E. coli mutant that is resistant to replication initiation inhibitors and holds a fluorescence reporter system for DNA replication inhibitors. This screen was also subjected to the above-mentioned library of microbial extracts, though it did not lead to any positive hits.

3 Abstract – Danish

Verdens Sundheds organisationen, WHO, har udnævnt antibiotika resistens til at være en af de største trusler mod det globale sundhedssystem. Det er blevet estimeret at i 2050 vil ca. 10 millioner dødsfald årligt være associeret med antibiotika resistens. Det er derfor yderst vigtigt at der allerede nu tages de nødvendige initiativer til at begrænse denne negative udvikling. En af de væsentligste faktorer i kampen mod antibiotika resistens er udviklingen af nye antibiotika, samt at præservere virkningen af de antibiotika som allerede bruges i klinikken.

I et forsøg på at præservere den kliniske anvendelighed af det ofte benyttede antibiotika ciprofloxacin. Søger vi i artikel I efter gener i Escherichia coli hvis deletion reverserer ciprofloxacin resistens og dermed kan bruges som mål for ciprofloxacin hjælpestoffer. Ved hjælp af genetisk deletions analyse identificerede vi efflux pumpen, AcrAB-tolC, samt SOS-respons proteinerne, RecA og RecC som mulige mål for ciprofloxacin hjælpestoffer i lav-resistente stammer af E. coli. Ydermere viste vores resultater også at det formentlig ikke er muligt at reverserer ciprofloxacin resistens i høj-resistente stammer af E. coli.

I artikel II præsenterer vi to nye screeningssystemer til at identificere inhibitorer af initieringen af kromosomal DNA replikation i bakterier. Disse to screeningssystemer er baseret på celler der over-initierer DNA replikationen, via henholdsvis forhøjet regenerering af den ATP bundne form af initieringsproteinet DnaA eller mangel på processen kendt som regulativ inaktivering af DnaA (RIDA). Denne over-initiering er lethal for cellerne. Under screening af et bibliotek bestående af 400 mikrobielle ekstrakter, identificerede vi jern chelatoren deferoxamine, som et stof der kan redde væksten af celler der over-initierer replikationen. Dog ikke ved at nedsætte initierings frekvensen, men ved at reducere produktionen af reaktive oxygen radikaler. Hvilket ydermere fast slår modellen, at oxidativ DNA skade og dets reparation medierer celledød i bakterier det over-initierer DNA replikationen.

I artikel III konstruerede og verificerede vi endnu et nyt screeningssystem til inhibitorer af DNA replikations initieringsprocessen. Denne screen består af en E. coli mutant der er resistent over for stoffer der blokerer replikations initierings processen og samtidig indeholder et fluorescens baseret reporter system der aktiveres af replikations initierings inhibitorer. Denne screen blev også testet mod det ovennævnte bibliotek af mikrobielle ekstrakter, men gav ingen positive hits.

4 List of papers

Paper I

Can Ciprofloxacin Resistance be Reversed by Helper Drugs? Rasmus N. Klitgaard, Bimal Jana, Luca Guardabassi, Karen Leth Nielsen and Anders Løbner-Olesen.

Paper II

A strategy for finding DNA replication inhibitors in E. coli identifies iron chelators as molecules that promote survival of hyper-replicating cells. Godefroid Charbon, Rasmus Nielsen Klitgaard, Charlotte Dahlmann Liboriussen, Peter Waaben Thulstrup, Sonia Ilaria Maffioli, Stefano Donadio and Anders Løbner-Olesen.

Paper III

A Novel Fluorescence Based Screen for Inhibitors of the Initiation of DNA Replication in Bacteria. Rasmus N. Klitgaard and Anders Løbner-Olesen.

Papers not included in the thesis

Ciprofloxacin intercalated in fluorohectorite clay: Identical pure drug activity and toxicity with higher adsorption and controlled release rate. E. C. dos Santos, Z. Rozynek, E. L. Hansen, R. Hartmann-Petersen, R. N. Klitgaard, A. Løbner-Olesen, d L. Michels, A. Mikkelsen, T. S. Plivelic, H. N. Bordallo and J. O. Fossum.

Mutations in the Bacterial Ribosomal Protein L3 and Their Association with Antibiotic Resistance. Rasmus N. Klitgaard, Eleni Ntokou, Katrine Nørgaard, Daniel Biltoft, Lykke H. Hansen, Nicolai M. Trædholm, Jacob Kongsted, Birte Vester.

5 Table of contents

ACKNOWLEDGEMENTS ...... 2

ABSTRACT – ENGLISH ...... 3

ABSTRACT – DANISH ...... 4

LIST OF PAPERS ...... 5

TABLE OF CONTENTS ...... 6

A BRIEF HISTORY OF ANTIBIOTICS ...... 9

The early days ...... 9

The golden age of antibiotics ...... 9

The present and future of antibiotics ...... 10

PART I: POTENTIATION OF THE QUINOLONES ...... 11

Discovery and development of the quinolones ...... 11

The quinolone targets ...... 12

Mechanism of action ...... 14 Fragmentation of the bacterial chromosome ...... 14

Reactive oxygen species and quinolone lethality ...... 15 Are ROS involved in quinolone lethality? ...... 15

The SOS response, an endogenous defense against quinolones ...... 16 Regulation and induction of the SOS response ...... 17 Repair of quinolone mediated double stranded DNA breaks by the SOS response ...... 17

Quinolone resistance ...... 18 Target site ...... 18 Non-target site mutations involved in quinolone resistance ...... 19 Plasmid mediated quinolone resistance ...... 19

Reversing antibiotic resistance by helper drugs ...... 22 Potential targets for potentiation of quinolones ...... 23

6 PART II: TARGETING THE INITIATION OF CHROMOSOMAL DNA REPLICATION IN BACTERIA ...... 23

Initiation of chromosomal DNA replication in E. coli ...... 24 DNA replication and the cell cycle...... 24 Initiation of replication ...... 25 The origin of replication ...... 26 The initiator protein DnaA ...... 27

Replication initiation by DnaAATP ...... 29 Formation of the DnaAATP initiation complex ...... 29 DUE unwinding ...... 30 DnaB helicase loading ...... 31

Regulation of the replication initiation ...... 31 The dual role of DiaA in regulating replication initiation ...... 32 Regulatory inactivation of DnaAATP (RIDA) ...... 33 datA-dependent DnaAATP-hydrolysis (DDAH) ...... 33 Regulation of DDAH activity ...... 34 SeqA, a negative regulator of the replication initiation ...... 35 Rejuvenation of the cellular DnaAATP pool ...... 36

The lethal action of severe over-initiation of the DNA replication ...... 39

Targeting the Initiation of replication ...... 40

PAPER I: CAN CIPROFLOXACIN RESISTANCE BE REVERSED BY HELPER DRUGS?...... 42

PAPER II: A STRATEGY FOR FINDING DNA REPLICATION INHIBITORS IN E. COLI IDENTIFIES IRON CHELATORS AS MOLECULES THAT PROMOTE SURVIVAL OF HYPER-REPLICATING CELLS...... 57

PAPER III: A NOVEL FLUORESCENCE BASED SCREEN FOR INHIBITORS OF THE INITIATION OF DNA REPLICATION IN BACTERIA...... 94

DISCUSSION ...... 102

Potentiation of the quinolones ...... 102

Targeting the commencement of DNA replication in bacteria ...... 104

Why is severe over-initiation of the DNA replication lethal? ...... 106

7 CONCLUSIONS ...... 107

FUTURE PERSPECTIVES ...... 107

BIBLIOGRAPHY ...... 109

8 A brief history of antibiotics

The early days

Most people, even without a background within life science, have heard the intriguing story of how Alexander Fleming by coincidence contaminated his agar plates with mould and discovered back in 1929 (1). Of less common knowledge is the pioneering work of Alexander Ehrlich and Sahachiro Hata, which led to the discovery of salvarsan, in 1909, a novel drug for treating the sexual transmitted disease syphilis that is caused by the spirochete Treponema pallidium (2). Salvarsan and its derivative neosalvarsan, were the most prescribed drugs until they were replaced by penicillin in the 1940s (3). The large-scale screening method used by Ehrlich and Hata in the discovery of salvarsan, became the gold standard for identifying novel drugs and led to the discovery of the first sulfa drug in 1934, sulfonamidochrysoidine, a precursor of the active compound , which inhibits folic acid synthesis in bacteria (3, 4).

The golden age of antibiotics

The discovery of the sulfa drugs and the release of penicillin for clinical use kick-started a period of 30 years known as the golden age of antibiotics (1940-1970), in which almost all of the antibiotic drug classes used in the clinic today were discovered (see Figure 1) (5, 6). Most of the antibiotics discovered in this period were isolated from natural extracts from different microorganisms. Following the isolation

Figure 1: The top panel indicates the time at which different antibiotics and classes of antibiotics were discovered. The bottom panel, indicates when resistance was observed for the given antibiotics. Modified from Clatworthy et al., 2007.

9 of streptomycin, in 1944, from the soil growing filamentous bacteria Streptomyces griseus. Soil samples were collected from around the world and in 1952 the producing Streptomyces orientalis was isolated from a soil sample from Borneo, leading to the release of vancomycin for clinical use in 1958 (5). Despite of its name it was also in the golden age that it became evident that clinical antibiotic resistance would become a problem. In 1945, Alexander Fleming, during his Nobel lecture, warned that underdosing of penicillin could potentially lead to the development of resistance (7). In the decade following Flemings warning, it became apparent that antibiotic resistance was a problem. To overcome resistance scientists started to make derivatives of already know drugs, this led to the development of antibiotics that were impervious to the resistance mechanisms and in some cases improved the pharmacodynamics and pharmacokinetics of the drugs (5). However, it was also the start of a race between the evolution of antibiotic resistance and the development and discovery of antibiotics. A race that currently seems to be led by the bacteria.

The present and future of antibiotics

In the last 40 years, the only truly novel class of antibiotics that has been introduced into the clinic are the oxazolidinones, initially represented by the synthetic compound that was released in 2000 (8). Due to its synthetic nature it was anticipated that linezolid resistance would evolve slowly (9). This presumption unfortunately turned out to be wrong, as soon after its release, linezolid resistance was identified in clinical isolates of Staphylococcus aureus and several enterococcus species (10). As of December 2017 an estimated 48 antibiotics are in phase I to III clinical trials. Most of these antibiotics are derivatives of known antibiotics, almost half do not target listed as being a critical threat by the World Health Organisation (WHO) and even fever are expected to display activity against the multi drug resistant group of Gram negative ESKAPE pathogens (11). Considering that on average only one third of these antibiotics will make it through the clinical trials and become a marketable product, the current antibiotic pipeline is not robust enough to support the current and future clinical need (12). In addition, a report commissioned by the government of the United Kingdom in 2014, estimated that the annual number of deaths attributable to would be 10 million by 2050 and that it will generate a loss of 100 trillion dollars globally (13). Even though these numbers are only estimates, there is no doubt; antibiotic resistance is a major global health care problem and it will only become more evident with time, if proper action is not taken.

10

Part I: Potentiation of the quinolones

The quinolone class of antibiotics includes some of the most widely used and prescribed antibiotics (14- 16). Due to their popularity and misuse, quinolone resistance has become a major problem in the clinic (17, 18). In context of the current lack of development of novel classes of antibiotics, potentiation of already known antibiotics will be essential. At least until the antibiotic development pipeline has become more robust. The following sections will introduce the reader to the quinolone class of antibiotics, quinolone resistance and how quinolones might be potentiated to overcome resistance.

Discovery and development of the quinolones

In 1964, Sterling Drugs released the first compound of a novel class of antibiotics for use in the clinic, named . Though nalidixic acid is based on a 1,8-naphthyridone core and therefore technically not a quinolone (see Figure 2A), it is in general acknowledged as the first quinolone antibacterial. The events that led to the discovery of nalidixic acid are somewhat unclear. The story goes that a by-product of the synthesis of the antimalarial drug , made at Sterling Drugs inc., showed antibacterial properties and contained a quinolone core. Sterling has newer commented on why the quinolone core was substituted for a 1,8- napthyridone core in nalidixic acid. However, it was likely because there had already been filled a patent, by Imperial Chemical Industries in 1960, on a Figure 2: A) Comparison of the quinolone core with compound similar to nalidixic acid, but with a the 1,8-naphthyridone core of nalidixic acid. quinolone core. In the years after the release of Adapted from Bisacchi et al., 2015. B) The structure nalidixic acid, a number of follow-up drugs were of the second-generation quinolone, ciprofloxacin. released (19, 20). The first generation of quinolones The at position C6 is marked by a blue were mainly used in treating uncomplicated urinary circle and the substituent at position C7 by a red circle.

11 tract , as their systemic absorption was poor. In the early 1980s, the first second-generation compounds were released, including ciprofloxacin and . The major differences from the first- to the second-generation compounds were the addition of a fluorine at position C6 and a piperazine or methyl-piperazine substituent at C7 (See Figure 2B). The addition of the fluorine, led to the term fluoroquinolones. These two additions to the quinolone core, improved both the bacterial spectrum, but also the pharmacokinetic and pharmacodynamics significantly (21). Since then both third and fourth generation fluoroquinolones has made its way into the clinic. The third generation fluoroquinolones like, , and expanded the bacterial spectrum to include streptococci and had prolonged half-lives. The fourth generation fluoroquinolones, was the first generation with activity against anaerobes like, bacteroides fragilis, in addition to an enhanced activity against Gram- positives (22). Furthermore, the 8-methoxy group possessed by two of the fourth generation drugs, and , eliminated the phototoxicity observed for earlier generations (22). Throughout the rest of this thesis the term quinolone, will be used for both first generation quinolones and the fluoroquinolones, unless differences are specified.

The quinolone targets

The cellular pathway targeted by nalidixic acid was revealed already in 1964. By measuring the incorporation of C14-labeled thymine in DNA, it was shown that it inhibited the DNA synthesis (23). Five years later, in 1969, Hane et al. genetically mapped mutations in two distinct genes, nalA and nalB, that conferred different levels of nalidixic acid resistance (24). nalA was subsequently identified as gyrA, encoding the subunit of the DNA gyrase responsible for nicking and re-ligation of the DNA (25, 26). The DNA gyrase is not the sole target of the quinolones, in 1990 a novel topo-isomerase, topo-isomerase IV (topo IV), was discovered (27). Topo IV is the gene-product of parC and parE, which have a high degree of sequence homology Figure 3: A, B) Crystal structure of a topo IV cleavage complex bound by with gyrA and gyrB, especially in the regions where there have two molecules moxifloxacin. In been identified mutations conferring quinolone resistance (28). green: the ParE subunits. In blue: the The DNA gyrase and topo IV are both type II ParC subunits. In red: moxifloxacin , essential for numerous processes involving and in yellow: the DNA strand. Modified from Aldred et al. 2013

12 nucleic acids, including DNA replication and chromosome segregation. They modulate the topology of the DNA by controlling the level of under- and over-winding and are able to sort tangles and knots in the DNA. The modulation of the DNA topology is achieved by inducing transient double-stranded brakes in the DNA, thereby releasing the torsional stress. To maintain the integrity of the chromosome during the opening of the double-strand, the gyrase and topo IV binds covalently to the generated 5´-DNA ends, creating a so-called “cleavage complex” (see Figure 3AB). Following the cleavage reaction the DNA is re-ligated again by the bound enzyme(21). The gyrase and topo IV are both heterotetramers with an A2B2 quaternary structure. The gyrase consists of two GyrA and two GyrB subunits, while Topo IV contains two ParC and two ParE subunits. The GyrA/ParC subunits holds the active site tyrosine residues and are responsible for the DNA cleavage and ligation reactions, while GyrB and ParE both contain an ATPase domain delivering the energy for the cleavage and Figure 4: A) Structure of the water-metal ligation reaction by hydrolysis of ATP (29). Studies of crystal ion bridge between the C3/C4 keto acid of structures of type II topoisomerases bound by different the quinolone, the serine and either quinolones have revealed that the binding is mediated by a aspartic acid or glutamic acid. B) Alignment of GyrA and ParC (GrlA) from water-metal ion bridge (29-31), between the C3/C4 keto acid of baumanii (Ab), (Ba), the quinolone and Ser83, Asp87 in GyrA, or Ser80 and Glu84 in Escherichia coli (Ec), Staphylococcus aureus ParC (E. coli numbering, see Figure 4A). These findings are (Sa) and (Sp). In supported by the fact that the most common amino acid red the serine and the acidic amino acid substitutions conferring resistance to quinolones, have been that forms the water-metal ion bridge. Note that the human homologs hTIIα and identified at these specific amino acids (32). Notefully, the hTIIβ do not contain the serine or the acidic human type II topoisomerases , hTIIα and hTIIβ, do not contain amino acid. Modified from Aldred et al., these residues (see Figure 4B). It has therefore been proposed 2014. that lack of the necessary amino acids to mediate the water- metal bridge, is one of the main reasons why quinolones do not target hTIIα and hTIIβ (21). Earlier, based on studies in E. coli, S. aureus and , it was pressumed that the DNA gyrase was the primary target in Gram-negative bacteria, while topo IV was the primary target in Gram-positive bacteria (33-35). However, this presumption turned out to be incorrect,

13 as later studies have shown that the target specificity is both species and drug dependent (21). For instance, in S. aureus nalidixic acid was shown to target the gyrase and norfloxacin preferentially topo IV, while ciprofloxacin targeted both the gyrase and topo IV (36).

Mechanism of action

Fragmentation of the bacterial chromosome

In 1979 Kreuzer et al. proposed that quinolones were acting as poisons, corrupting the function of the gyrase, rather than directly targeting the catalytic effect of the gyrase(37). This hypothesis turned out to be true. Quinolones act by blocking the ability of topo IV and the gyrase to re-ligate the cleaved DNA in the cleavage complex, in turn leading to fragmentation of the bacterial chromosome (21). The exact events that leads to the fragmentation is still under debate. Earlier, it was believed that the quinolone bound cleavage-complex was converted to a permanent break if hit by a replication fork or other complexes moving along the DNA. This idea originated from the findings that eukaryotic I, trapped on the DNA, created double stranded breaks when colliding with a replication fork (38, 39). However, no one have been able to show that this is also the case in bacteria. Though it was shown that collision between the quinolone bound cleavage-complex and the replication fork, stalled the replication fork and rendered the quinolone-cleavage-complex in an irreversible state (40-43). These findings spawned the idea, that stalling of the replication fork was followed by endonuclease mediated clevage of the DNA at the replication fork (44). This model was later challenged, as halting DNA replication, by using a temperature sensitive DnaB helicase mutant, did not affect quinolone lethality (45). Studies of the lethal action of nalidixic acid and gatifloxacin treatment in combination with , a protein synthesis inhibitor, revealed that the lethal action of nalidixic acid was blocked without ongoing protein synthesis, while the lethality of gatifloxacin was retained (39, 46). These findings indicated the existence of two pathways leading to the lethal action of quinolones; a protein synthesis dependent pathway and a pathway independent of protein synthesis. It has been proposed that binding of quinolones to the cleavage complex destablizes the complex, thereby releasing the double stranded DNA break from the cleavage complex (39). This model is independent of protein synthesis and is supported by the fact, that gatifloxacin can fragment chromosomes in vitro in the presence of purfied gyrase (39). Additionally, it has also been shown that an E. coli mutant, where the gyrase has been destabilized by introduction of a GyrA A67S , is killed by nalidixic acid in the presence of chloramphenicol, in contrary to the wild type (39). The chromosome fragmentation pathway dependent on protein synthesis is less clearly understood. However, protease digestion of the

14 gyrase or nuclease-mediated cleavage on either side of the cleavage complex, have been suggested to mediate the release of the DNA from the cleavage complex (47).

Reactive oxygen species and quinolone lethality

In E. coli, deleterious reactive oxygen species (ROS) are continuously formed during respiration, when

.- .- auto-oxidation of its redox enzymes generates superoxide (O2 ) (48). To prevent accumulation of O2 it is converted by superoxide dismutases to oxygen and hydrogen peroxide (H2O2). Intracellularly H2O2 can react with iron (II), leading to generation of highly reactive hydroxyl radicals (OH•) through Fenton chemistry (49): • + → + +

The generated hydroxyl radicals can essentially react with and damage most biomolecules, including DNA, proteins and lipids. As there are no known cellular pathways degrading hydroxyl radicals, its generation is limited by peroxidases and catalases that degrade H2O2 (50). From 2002 to 2006, a number of papers reported that treatment of bacteria with bactericidal antibiotics lead to heightened levels of ROS and that ROS was involved in the lethal action of bactericidal antibiotics (51-54). In 2007, the first model for a ROS mediate cellular death pathway induced by bactericidal antibiotics was published. It was proposed that bactericidal antibiotics stimulate oxidation of NADPH to NAD+ by the electron transport chain, leading to a boost in superoxide production. Superoxide mediated damage of iron-sulphur-cluster proteins then releases iron (II), which reacts with H2O2 and generates hydroxyl radicals through the Fenton reaction. At the time, cell death was explained by general hydroxyl radical mediated damage to proteins, lipids and DNA (55, 56). Later, oxidation of the cells nucleotide pool, specifically generation of 8-oxo-dGTP and its incorporation into DNA was proposed as the dominant mechanism by which ROS mediates cell death by bactericidal antibiotics (57, 58). Proposing that ROS significantly contributed to the lethal action of bactericidal antibiotics was controversial and it is still a matter of debate.

Are ROS involved in quinolone lethality?

The first clues indicating that quinolone treatment of E. coli led to an increase in ROS production, came from the observation that nalidixic acid significantly increased the expression of the superoxide dismutase, encoded by sodA (59). The increase in expression of sodA was later shown to be mediated by activation of the soxRS regulon (60, 61), a major oxidative stress response system fund in most Gram-negative bacteria (62, 63). Investigations of the involvement of the soxRS regulon in quinolone resistance revealed that over-expression of soxS in E.coli and constitutive activation of the

15 soxRS regulon in enterica increased the level of quinolone resistance. However, it should be noted that the observed resistance was likely, due to the fact that activation of SoxRS results in posttranscriptional negative regulation of the OmpF porin, involved in quinolone transport into cells (64) and overproduction of the AcrAB-TolC efflux pump (65). Several different quinolones have been shown to increase ROS production, as detected by both chemiluminescence and fluorescence methods, in E. coli, S. aureus and Enterococcus faecalis (51-53, 55, 56). Furthermore, blocking ROS generation by either antioxidants or iron chelators lowers the susceptibility of E. coli to some quinolones (54, 66). In addition, deletion analysis, in E. coli, of the genes involved in H2O2 metabolism, katG, ahpCF and katE , showed that a katG, ahpCF double mutant and a katG, ahpCF, katE triple mutant were more susceptible to ciprofloxacin than the wild-type(54). To challenge the proposed model for ROS mediated killing by bactericidal antibiotics described above. The efficacy of a number of quinolones was investigated under anaerobic conditions, where ROS cannot be generated. This included the first generation quinolone, nalidixic acid and the fluoroquinolones norfloxacin, ciprofloxacin and (67-69). The results showed that anaerobic growth did not lead to an increase in MIC. However, anaerobic conditions blocked the killing by nalidixic acid, but not by norfloxacin, ciprofloxacin or ofloxacin, though higher concentrations of norfloxacin and ciprofloxacin were required to kill the cells when grown anaerobically (67-69). Furthermore, quenching of ROS production by treatment with the iron chelator, dipyridyl and the reducing agent thiourea, blocked the lethality of oxolonic acid, but only partially reduced the lethal action of moxifloxacin, while the C8-methoxy fluorquinolone, PD161144, was unaffected. Interestingly, there is an inverse correlation between the lethal action of quinolones under anaerobic conditions and the observed degree of protein synthesis dependency for lethality (66, 67). Indicating that the protein dependent pathway relies on generation of ROS, while the protein synthesis independent pathway does not (66). However, further research is needed to elucidate the exact mechanism that connects ROS with the lethal action of the protein synthesis dependent pathway.

The SOS response, an endogenous defense against quinolones

Maintaining genome integrity is vital for bacteria, therefore most bacteria express an inducible DNA damage repair system termed the SOS response(70). As quinolones fragments the chromosome, they are strong induceres of the SOS response (55, 71), which acts as a first line of defence against this group of antibiotics. In addition, the activation of the SOS response leads to high mutation rates, which in turn can result in occurrence of mutations conferring resistance to quinolones (32, 72). Therefore the SOS response is a key process in both quinolone susceptibility and in the evolution of quinolone resistance.

16 Regulation and induction of the SOS response

The SOS response is regulated by two key proteins, the LexA repressor and the activator; RecA. During regular cell growth, the LexA repressor binds to a specific sequence in the promoter regions of the SOS response genes called the SOS box. The binding of LexA to the SOS box blocks the expression of the SOS response genes. In addition, the binding of LexA to the SOS box also regulates the sequence by which the SOS response genes are expressed during DNA damage. Genes expressed early in the SOS response have a low affinity SOS box, while the SOS box in the promoter region of late SOS genes has a high affinity for LexA. When the DNA is damaged, filaments of activated RecA are assembled on persisting regions of single stranded DNA. The assembly of the RecA filaments facilitates the autocleavage of the LexA repressor, thereby leading to expression of the SOS response genes. In E. coli more than 40 genes are regulated by LexA cleavage in response to DNA damage, including genes responsible for DNA repair and cell cycle control (73-75).

Repair of quinolone mediated double stranded DNA breaks by the SOS response

One of the major tasks carried out by the SOS response genes is DNA damage repair. The DNA repair systems that are part of the SOS response can repair a number of different types of DNA damage. Repair of double stranded breaks (DSB) in bacteria, like those caused by quinolones, is achieved by homologous recombination (HR). In E. coli there are two known pathways of HR , the RecBCD- and RecF-pathway, where RecBCD is the predominate one (see Figure 5) (76). RecBCD is a multi-functional enzyme complex, having both nuclease and helicase activity, and is responsible for processing the open DNA ends formed at DSBs in the DNA. RecBCD initiates the DSB repair by binding to the open DNA- end at the DSB and starts unwinding the DNA. Hereafter, a combination of helicase and nuclease activity leads to formation of a single stranded 3´-overhang. When the RecBCD complex have reached a so-called chi site on the strand with the open 3´-end, it loads RecA onto the 3`-tail, Figure 5: Schematic of DNA double stranded creating a RecA filament, and dissociates from the DNA (77). break repair by homologous recombination via the RecBCD pathway. Modified from RecA then catalyzes strand invasion of a homologous dsDNA, Wyman et al. 2004

17 creating a displacement loop (D-loop). Hereafter the intact homologous DNA strands are used as templates for the DNA polymerase. DNA crosses termed Holiday junctions now physically link the hetero duplex DNA strands. To resolve the Holiday junctions, the RuvAB protein complex extends the heteroduplex DNA region by migrating the Holiday junctions in an outward direction. Following the hetero-duplex extension, the RuvC protein associated with RuvAB, resolves the Holiday junctions by nicking the crossed strands. A DNA ligase then ligates the nicks in the DNA, reconstituting the two double strands (78).

Quinolone resistance

Quinolones have become one of the most prescribed antibacterial drugs in the world today (14-16), It is therefore not surprising that quinolone resistance has been identified in almost all bacterial species of clinical interest (17, 18). A number of different resistance mechanisms conferring quinolone resistance have been identified this far, including; target site mutations, enzymatic inactivation, target protection and efflux systems.

Target site mutations

Quinolone resistance is most frequently caused by target site mutations in the gyrase and topo IV, eventhough the mutations confering quinolone resistance have been mapped to wide range of positions in both subunits of the gyrase and topo IV. The most frequent mutations are found at the serine and acidic residues of gyrA and parC that are critical for binding of quinolones via the water-metal-ion bridge (21). Studies have revealed that more than 90% of all clinical isolates and laboratory derived strains with lowered susceptibility to quinolones generally have a mutation at the specific serine residue, and that 85% of these also have parC mutations (79). In vitro selection of quinolone resistant mutants have shown that the mutations in the gyrase and topo IV are selected for in a stepwise manner (34, 80). In most cases multiple mutations are needed to confer clinical quinolone resistance. In an E. coli background without any other quinolone confering mechanisms, two amino acid substitutions in gyrA and one in parC is needed for the ciprofloxacin MIC to exceed the CLSI clinical breakpoint of 1 µg/mL (81, 82). Often, target site mutations confering antibiotic resistance have a negative impact on the fitness of the bacteria (83, 84). It has therefore, to some degree, been surprising that quinolone resistance caused by target site mutations, has become such a serious problem in the clinic. An explanation to this paradox is likely that third-step quinolone resistance mutations have been shown to both restore fitness and increase resistance significantly (80, 81). The increase in fitness could thereby catalyze the selection of mutants highly resistant to quinolones, without exposure to high quinolone

18 concentrations (81). Furthermore, there is evidence that accumulation of quinolone resistance mutations lead to increased mutation rates (85). Taken together with the fact that quinolones, as mentioned earlier, are induceres of the SOS response and thereby mediates expression of the error- prone DNA polymerase IV (86). The frequent occurrence of quinolone resistance might not be so suprising after all.

Non-target site mutations involved in quinolone resistance

A number of different non-target site mutations often occur in quinolone resistant bacteria. Including, deleterious mutations in acrR and marR, a direct and an indirect repressor of the expression of the endogenous AcrAB-TolC efflux system in E. coli (87), thus leading to elevated levels of quinolone efflux. MarR acts by repressing expression of marA, a global transcriptional activator, that activates expression of acrAB (88). In addition, MarA also activates transcription of micF, encoding an antisense RNA that post-transcriptionally inhibits the outer membrane porin, OmpF (89). OmpF is important for quinolone entry into the cell in E. coli and its inhibition leads to lowered quinolone susceptibility (90, 91). Other Gram-negative bacteria have similar efflux systems. For instance, aeruginosa expresses the MexAB-OprM efflux system that is repressed by MexR (92). Quinolone resistant clinical isolates of P. aeruginosa often have mutations in mexR leading to overexpression of the MexAB-OprM efflux pump (93, 94). The Gram-positive bacteria where quinolone efflux is best characterized is S. aureus. Here overexpression of three different efflux pumps; NorA, NorB and NorC, have been shown to lower the quinolone susceptibility. The regulation of these three efflux pumps is somewhat more complex than for AcrAB-TolC and MexAB-OprM, as some transcriptional regulators, like GntR, is both an activator of norA and norB, but a repressor of norC (95). A number of other efflux systems in both Gram-negative and Gram-positive bacteria have been linked to quinolone resistance (95), but these will not be discussed here.

Plasmid mediated quinolone resistance

Eventhough target site mutations are the most frequent cause of quinolone resistance, a number of different plasmid mediated resistance mechanisms have also been identified in clincal isolates. These mechanisms do generally not confer clinical quinolone resistance, but have been shown to facilitate selection of high level quinolone resistance (96-98). The first claims of a plasmid mediated quinolone resistance (PMQR) mechanism were reported back in 1987 (99), but was later withdrawn. Though, it was first over 10 years later, by Martinez-Martines et al., that the existence of PMQR was confirmed by transfer of a plasmid that lowered the susceptibility for nalidixic acid and ciprofloxacin in an otherwise susceptible E. coli strain (96).

19 The qnr genes; a DNA mimic. The gene identified by Martinez-Martinez et al. was named qnr (later qnrA) (96), encoding an 218 amino acid long protein, belonging to the pentapeptide repeat protein (PRP) family. The PRP family contains more than a 1000 proteins, many of which are of unknown function (100). The PRPs are defined by being composed of or having domains of tandem peptide repeats with the consensus sequence; [S,T,A,V], [D,N], [L,F], [S,T,R] and [G] (101). It was the function of two other members of the PRP familly, MfpA and McbG, that led the way to the discovery of the function of QnrA. MfpA and McbG are both encoded on the chromosome and protect the DNA gyrase from ciprofloxacin and the natural DNA gyrase poison microcin B17, respectively (98). Knowing this, the in vitro supercoiling activity of DNA gyrase in the presence of ciprofloxacin and purified QnrA was assesed. The results reveald that QnrA protected the DNA gyrase from inhibition by ciprofloxacin, retaining its ability to supercoil DNA (102). The discovery of qnrA was followed by the discovery of six other families of plasmid born qnr genes; qnrS (103), qnrB (104), qnrC (105), qnrD (106), qnrE (107) and qnrVC (108). These six qnr families generally have around 65%, or less, sequence homology with qnrA and each other(100). Crystal structures of QnrB1 and a Qnr protein from the Gram-negative bacteria Aeromonas hydrophila showed that they are dimers linked at the C-termini, folding into a right-handed β-helix (see Figure 6). This strutucture resembles the size, shape and charge of β-DNA, which has led to the current opinion; that Qnr proteins are DNA mimics that bind to and destabilize quinolone bund cleavage-complexes, leading to release of the bund quinolone and reactivation of the topoisomerase (100, 109, 110). It still remains to be resolved, how Qnr proteins can compete with DNA for binding to the DNA gyrase without significantly inhibiting the gyrase actitivty in the bacteria.

Figure 6: Structure of the QnrB1 dimer. The two QnrB1 monomers are linked at the C-termini and fold into a right-handed quadrilateral β-helix, mimicking the size, structure and charge of β-DNA. Deletion of loop A´or B´ leads to lowered protection of DNA gyrase from ciprofloxacin. Modified from Vetting et al., 2011.

20 Inactivation by AAC(6´)-lb-cr mediated acetylation The AAC(6´)-lb protein family consists of 6´-N- acetyltransferases that can inactivate a number of aminoglycoside antibiotics by acetylation (98). It was therefore surprising when disruption of an aac(6´)-lb gene, on a multiple resistance plasmid from a clinical isolate of E. coli, led to increased ciprofloxacin susceptibility (111). An acetylation assay showed that this novel member of the AAC(6´)-lb family was able to N-acetylate ciprofloxacin at the amino nitrogen on its piperazinyl substituent (see Figure 7). The enzyme was therefore called AAC(6´)-Ib-cr, where “cr” stands for ciprofloxacin resistance (111). AAC(6)-Ib-cr also confers resistance to norfloxacin, but not other quinolones as they lack the unsubstituted amino nitrogen group (111). As with Qnr, AAC(6)-Ib-cr does not, by it self, cause clinical quinolone resistance. It increases the MIC by three- to four-fold in wild Figure 7: AAC(6´)-Ib-cr acetylation of the type E. coli, but more interestingly it increases the mutation amino nitrogen of the piperazinyl prevention concentration significantly. Thus, it likely plays an substituent in ciprofloxacin. important role in selecetion of higher level resistance mutations (98, 111). In addition to the seven different allelic AAC(6´)-Ib-cr variants that have been identified this far. An 24 amino acid longer variant, termed AAC(6´)-Ib-cr4, was discovered in a clinical isolate of Salmonella typhimurium (112).

QepA and OqxAB efflux pumps QepA and OqxAB are the major types of efflux pumpes that are involved in PMQR. QepA is part of the 14-transmembrane-segment family of the major facilitator superfamily transporters and was discovered in an E. coli isolate from Japan with lowered susceptibility to quinolones (113). QepA is able to actively pump out hydrophilic fluoroquinolones, especially norfloxacin and ciprofloxacin. The increase in MIC conferred by QepA varies from 2-64 fold, this wide range is most likely caused by differences in QepA expression (112). The OqxAB efflux system is a member of the resistance-nodulation-cell division family of transporters and is able to pump out a range of different antibiotics, including; quinolones, chloramphenicol and trimethroprim. OqxAB is highly associated with extended spectrum beta lactamase (ESBL) producing , where it is found on the chromosome and on plasmids. Like

21 QepA, the expression level of OqxAB varies widely, hence the change in MIC differs from strain to strain (112).

Reversing antibiotic resistance by helper drugs

In an effort to overcome the current crisis with treating infections caused by multi-drug resistant bacteria, many different treatment types have been investigated. One of them is the reversal of antibiotic resistance by combining antibiotic treatment with administration of a potentiating compound also known as a helper drug. A helper drug is a compound that does not have an antibiotic effect in itself, but is able to reverse the antibiotic resistance against a given antibiotic. In general helper drugs can act by either directly targeting resistance mechanisms or by targeting intrinsic mechanisms protecting the bacteria from the antibiotic, like; efflux pumps, cell membranes and repair systems. An example of a helper drug that have been used with great success in the clinic is the combination of clavulanic acid and the beta-lactam, amoxicillin. Clavulanic acid is a Beta-lactamase inhibitor (114), that reverses resistance by competitively binding to beta-lactamases (115), thereby blocking the inactivation of amoxicillin by the beta-lactamase. Another type of helper drugs that have been heavily investigated are efflux pump inhibitors (EPI), as many multi drug resistant pathogens have acquired mutations that elevates the expression of their endogenous efflux pump systems (116). Inhibitors of the resistance nodulation family (RND) of efflux pumps in Gram negative bacteria are especially interesting, as this family of efflux pumps is able to pump out a wide variety of antibiotics, including; fluoroquinolones (ciprofloxacin and levofloxacin), β-lactams, and oxazolidinones. Several compounds that inhibits the RND family, including AcrAB-TolC and MexAB-OprM, are described in the literature, but so far, none of these compounds have been licensed for medical use (117). In addition to the AcrAB-TolC and MexAB-OprM inhibitors mentioned above, celecoxib, a non-steroidal anti-inflammatory drug, has been shown to increase the susceptibility to ciprofloxacin in S. aureus. In silico screening of a small library of celecoxib analogues identified a compound that inhibited the NorA efflux pump and in vitro lowered the MIC for ciprofloxacin in a S. aureus strain overexpressing NorA(118). Furthermore, multiple compounds have been proposed to inhibit RecA in vitro and thereby prevent repair of quinolone mediated DSBs by HR and activation of the SOS response (119-125). However, only suramin and copper phtalocyanine-3, 4´, 4´´,4´´´-tetrasulfonic acid were shown to potentiate ciprofloxacin in vivo, albeit only weakly(119, 125).

22 Potential targets for potentiation of quinolones

A higher number of potential targets for quinolone helper drugs have been identified through genetic screens. Specifically, screening the entire Keio collection (126) of close to 4000 single non-essential gene deletion mutants of E. coli, revealed in excess of 25 genes, which when deleted significantly increased the susceptibility to ciprofloxacin. Unsurprisingly, genes involved in DNA replication, recombination and DNA repair counted for almost half of the total number of genes identified, though genes with a broad variety of other cellular functions were also represented (127, 128). These findings were obtained in an E. coli wild type strain, susceptible to ciprofloxacin. Paper I of this thesis addresses the question if deletion of any of the identified genes renders high- and low-level ciprofloxacin resistant E. coli strains clinically susceptible to ciprofloxacin, in an attempt to identify targets for ciprofloxacin helper drugs. During the preparation of the manuscript for paper I, it was reported that single gene deletion of more than 24 genes, identified as being involved in ciprofloxacin resistance, did not lower the MIC of a high- level ciprofloxacin resistant E. coli strain beneath the clinical break point. Conversely, deletion of acrB in combination with any of the SOS response genes recB, recC, recG or uvrD decreased the MIC of the same E. coli strain beneath the clinical breakpoint (129). Furthermore, deletion of recA was reported to render a low-level ciprofloxacin resistant strain clinically susceptible and to increase the in vivo efficacy of ciprofloxacin against the same strain in a peritoneal murine model (130).

Part II: Targeting the initiation of chromosomal DNA replication in bacteria

Potentiation of existing antibiotics is one of many methods that have been deployed to overcome antibiotic resistance. Another approach is the discovery of novel drugs that target unexploited processes essential for bacterial growth and viability. A target that is underexploited is the chromosomal DNA replication, and more specifically its initiation (131). Currently, the only antibiotics that targets the DNA replication and are used in the clinic are the quinolones and (132). The DNA replication is an attractive target for novel antibiotics for numerous reasons. The proteins that are involved in DNA replication are conserved in prokaryotes, but differ greatly with respect to their eukaryotic counterparts. Furthermore, the number of replisomes per cell is low; hence, the quantity of a given target that needs to be inhibited to block replication is correspondingly low (132). The following sections will introduce the reader to the replication initiation process and how it is regulated in E. coli, where it is best characterized.

23 Initiation of chromosomal DNA replication in E. coli

DNA replication and the cell cycle.

Early studies of the DNA content in bacteria growing at different rates showed that fast growing bacteria contained more DNA per cell, relative to slower growing ones (133). A logic explanation to this fact would be that replication forks move more rapidly during fast-growth. However, studies of DNA replication and cell cycle in balanced cultures of E. coli demonstrated that the average rate of DNA elongation is constant in cells with doubling times between 20 and 100 minutes (134). Furthermore, it was demonstrated that the duplication of the chromosome took approximately 40 minutes (the D period), independent of the growth rate, and that an additional 20 minutes was needed to complete septum formation and cell division (the C period). However, the so-called I-period, which defines the time it takes to prepare initiation of a round of chromosome replication is strictly dependent on the growth rate (135). Interestingly, the sum of C + D being equal to 60 minutes means that in cells with a doubling time below 60 minutes, replication and division is not completed before a new round of DNA replication is initiated. As replication forks in E. coli move bidirectional from a single and fixed origin of replication (136), newly divided cells, with a doubling time of less than 40 minutes, will therefore inherit branched chromosomes with multiple origins and ongoing replication forks (see Figure 8). The fact that fast growing cells contain branched chromosomes explained why they have a higher DNA content per cell relative to slower growing ones. In addition, it also revealed that timeous replication initiation was a key factor in the bacterial cell cycle (137, 138).

24 Figure 8: Replication of the E. coli chromosome during moderate growth. On top, a cell with one chromosome and four origins (green). On the right, the oldest replication forks terminates at the terminus (red) and the chromosomes are segregated. Replication is then initiated from the four origins followed by cell division. On the left, the cells have divided and now contain a single chromosome with four origins and four ongoing replication forks. Inspired by Fossum et al. 2007

Initiation of replication

In short, the chromosomal DNA replication in E. coli is initiated by binding of the initiator protein DnaA, in its active ATP-bound form, to the origin of replication, oriC. The formation of oriC-DnaAaTP nucleoprotein complex triggers the separation of the DNA double strand (139, 140). Following opening of the DNA double strand the nucleoprotein complex loads the DnaB helicase, with help from the helicase loader protein DnaC (141, 142). Loading of DnaB triggers the assembly of the remaining parts of the replication machinery necessary for DNA synthesis(143). Multiple regulatory systems have been identified, ensuring that replication initiation is triggered in a timely manner and only once per cell cycle for each origin. Following replication initiation, regulatory inactivation of DnaA (RIDA) stimulates the autohydrolysis of DnaA bound ATP to ADP, increasing the level of the inactive DnaAADP (144). A similar, but RIDA independent hydrolysis of ATP bound to DnaA is mediated by a mechanism referred to as, datA-dependent DnaAATP-hydrolysis (DDAH) (145). Furthermore, the negative initiation regulator protein SeqA sequesters hemi-methylated DNA in

25 the oriC, sterically hindering replication initiation by DnaAATP (146, 147). When time comes to reinitiate DNA replication, DnaAADP is reactivated by oligomerization at two DnaA-activating sequences (DARS1 and DARS2), which triggers the release of ADP. The nucleotide free apo-DnaA is then activated by binding of ATP and ready to initiate the DNA replication (148).

The origin of replication

In E. coli, the minimal oriC is a 245 bp DNA element containing two regions with distinct functionality; one is the DnaA oligomerization region (DOR) and the other the DNA unwinding element (DUE) (see Figure 9) (139). The DUE is defined by three 13-mer sequences (L, M and R) with the consensus sequence 5´-GATCTnTTnTTTT-3´ (149). DnaAATP complexed at the DOR unwinds the DUE region, which is susceptible to duplex unwinding due to its high AT-content. The DOR contains twelve DnaA binding sites, known as DnaA boxes, with the 9-mer consensus sequence 5´ TTATnCACA-3` (139, 150). The DOR can be divided into three sub-regions; the left- and right-halfs and a middle. The six DnaA boxes in the left-half (R1, τ1, R5M, τ2, I1 and I2) all point in the same direction, i.e. they are situated on the same strand. The remaining five DnaA boxes in the right-half (C3, C2, I3, C1 and R4) and the one in the middle (R2) share directionality, but in thee opposite direction of the DnaA boxes in the left-half (151-153). The R1, R2 and R4 DnaA boxes are moderate to high affinity DnaA boxes bound by either DnaAATP or DnaAADP throughout most of the cell cycle, While the remaining DnaA boxes are low affinity boxes (154, 155). The left-half DOR also contains a binding site (IBS) for the integration host factor (IHF) between DnaA box R1 and τ1.

Figure 9: Structure of the minimal oriC in E. coli. The twelve DnaA boxes and their directionality are shown by blue triangles. The IHF binding site is marked by a square between DnaA box R1 and τ1. Red arrows mark the three 13-mer AT-rich sequences of the DUE. Katayama et al., 2017.

26 The initiator protein DnaA

The master replication initiator DnaA is a conserved 473 amino acids (aa) long protein composed of four domains (156, 157) (see Figure 10). Domain I covers the first 87 aa in the N-terminal and is important for protein-protein interactions (158). Specifically, mutational studies have shown that Trp-6 is essential for the oligomerization of DnaA at the oriC by promoting domain I-domain I interactions between neighboring DnaAATP molecules (159-161). Furthermore, substitution of either Glu-21 or Phe-46 with alanine results in failure of DnaB helicase loading and for Phe-46 also binding of DiaA (161, 162), a stimulator of DnaAATP assembly on the oriC and DUE unwinding (163). Finally, Asn44 has been shown in vitro to be essential for RIDA, but not for initiation of replication (164).

Figure 10: Overview of the four domains of DnaA and its functions.

Domain II of DnaA is the least conserved domain and varies significantly in both length and sequence among bacterial species (157). It is usually described as flexible linker that connects domain I and domain III. Systematic deletions studies showed that having either the 21 N-terminal residues or the 27 C-terminal residues of the domain is sufficient for correct DnaA function, though replication initiation in the deletion mutants was less efficient than in wild type cells (158).

27 Domain III is the largest domain of DnaA and contains the AAA+ (ATPases associated with diverse cellular activities) region, making DnaA part of the AAA+ superfamily of proteins. The AAA+ module of DnaA can be divided into two subdomains; a αβα- nucleotide binding core and a Figure 11: The ATPase module of DnaA from Aquifex aeolicus bound by the smaller C-terminal α-helical bundle, ATP analog β,ϒ-methylene-ATP. Modified from Snider et al. 2008. known as the “lid”. The αβα-core is composed of several signature motifs holding residues that are important for ADP/ATP binding and ATP hydrolysis, while the sensor 2 motif is found in the “lid” (See Figure 11). The Walker A element forms a loop structure, important for ATP/ADP binding, while residues of the Walker B motif interacts with the magnesium ion that is crucial for ATPase activity (165). Lys-178 is an essential residue in the walker A element that is highly acetylated in stationary growth phase cells, preventing binding of DnaA to ATP and has therefore recently been proposed as novel regulatory mechanism of replication initiation (166). Asp-269 and Arg-334 of the sensor 1 and 2 motifs, respectively, are required for high affinity ADP/ATP binding (167, 168). Furthermore, Arg-334 is essential in DnaA ATP auto hydrolysis by both RIDA and DDAH, most likely due to direct interactions with the ϒ-phosphate of the bound ATP (169, 170). Box IV contains an arginine finger (Arg-285) that is exposed upon binding of ATP to DnaA. It is believed that the exposed Arg-285 is able to interact with the ATP in the neighboring DnaAATP molecule in the DnaAATP- oriC- complex, thereby facilitating the assembly of an active initiation complex (171). This kind of assembly is shared between all AAA+ oligomers that have been structurally characterized (165). In addition to Arg-285, four other residues, Lys-243, Arg-227, Arg-281 and Leu-290, are required for Domain III-Domain III interactions(139) and contribute to DnaA oligomerization and DUE unwinding (172-174). It as has recently been shown that Lys-243 can be acetylated in vivo, which blocks binding to the low affinity DnaA boxes I3, C1 and C3 in vitro, though its significance for replication initiation is still unsure (175). During DUE unwinding Val-211 and Arg-245 are believed to bind ssDUE, as in vitro assays have shown that alanine substitution mutants of any of the two residues leads to deficiency in both DUE unwinding and ssDUE binding. Domain IV in the C-terminal of DnaA contains a typical helix-turn-helix motif (HTH), which binds specifically to the DnaA box 9-mer consensus sequence. Crystallography studies of domain IV complexed with the R1 DnaA box revealed that the binding leads to a 20O degree bend in the DNA. A α-

28 helix in the HTH-motif, constituted by residue 434-451, is inserted into the major grove of the DnaA box, recognizing the 5’-TnCACA-3’ part of the consensus sequence (176). In addition, several residues of domain IV interacts with the phosphate backbone of the DnaA box, specifically mutations in Arg-407 and Lys-417 leads to DNA binding deficiencies (176, 177). A single residue of the HTH-motif, Arg-399, mediates base pair recognition by domain IV in the minor groove of the DnaA box (176). The importance of Arg-399 is emphasized by the fact that mutations in this specific residue leads to loss of sequence recognition and DNA binding (177). As for the major grove, multiple residues of the HTH-motif also interacts with the phosphate backbone of the minor groove (176). Molecular dynamic simulations and crystallography studies have shown that a short flexible loop connecting domain IV with domain III, allows for pivoting of domain IV, and indicated that this is an important feature in DnaA oligomerization (178, 179). Besides its essential function in DNA-binding, two residues of domain IV, Leu-422 and Pro- 423, contributes to binding of Hda, which is essential for RIDA activity (180).

Replication initiation by DnaAATP

The hallmarks of the replication initiation process is binding and oligomerization of DnaAATP on the oriC, DUE unwinding and DnaB helicase loading. Although the process of replication initiation has been investigated for decades, the exact structural and dynamic events that leads to replication initiation still remains to be fully elucidated, due to its complex nature.

Formation of the DnaAATP initiation complex

Studies of the orientation of DnaAATP molecules in complex with oriC have shown that structurally distinct complexes are formed on the left-half, right-half and middle DOR. As described above the DnaA boxes within each DOR are orientated in the same direction, hence the DnaAATP molecules bound to each box are also orientated in the same direction and interact in a head to tail manner (152, 173, 181). Truncation studies of the DOR regions revealed that the left-half DOR complexed with DnaAATP and bound by IHF is capable of mediating DUE unwinding, independently of the right-half and middle DOR (182). In the right half DOR, DnaAATP is believed to initially bind the high affinity box R4, which then triggers sequential binding of CI, I3, C2 and C3 (152). A similar binding order has also been proposed for the left-half DOR, where R1 binding is followed by binding at τ1, R5M, τ2, I1 and I2 (152). However, recently DnaA assembly studies on the left-half DOR revealed that deletion of R1 did not have a significant effect on DnaAATP assembly at the remaining DnaA boxes. Whereas, deletion of the low affinity box R5M severely impaired complex formation. Indicating, that R5M acts as the core assembly point in the left-half DOR (153). These findings fits well with the fact that sequential binding of DnaAATP molecules is most effective if the distance between the binding sites is 2-5 bp (152), which is the case

29 for R5M with respect to τ1 and τ2, but not for R1 that is situated 33 bp from its nearest neighbor, τ1 (E. coli, MG1655) (183). DnaAATP occupying the middle DOR DnaA box, R2, has been suggested to interact, via domain I-domain I interactions with DnaAATP occupying the I2 box, thereby stabilizing and promoting assembly of the initiation complex on the left-half DOR (153, 184).

DUE unwinding

Three different models have been proposed for the events leading to DUE unwinding (see Figure 12). The first is known as the continuous model or the two-state model. In this model, DnaAATP can take two forms; an extended dsDNA binding state and a closed ssDNA binding state. Initially, DnaAATP in its extended state binds to the DOR and a continuous DnaAATP filament is branched into the DUE, where a combination of ATP-dependent unwinding by DnaAATP and torsional stress starts to open the DNA duplex. As the DUE is unwound the conformational state of the DnaAATP molecules in the DUE shifts to the closed confirmation allowing them to bind and stabilize the ssDUE (185-187). The two other models for DUE unwinding are variants of the so-called loop back model. In the first variant, DnaAATP assembly in the right-half DOR starts at R4 and ends at C3. In the left-half DOR DnaAATP binds to R1, followed by binding of IHF. The IHF induced bend in the DNA loops back the R1-DnaAATP complex to the low affinity boxes in the left-half DOR and triggers the assembly of DnaAATP on the remaining left-half DnaA boxes. A combination of DNA bending by IHF and/or interactions with the DnaAATP oligomer on the left-half DOR unwinds the DUE, which is then bound by DnaAATP in its closed ssDNA binding state (188). The second variant of the loop back model differs from the first variant in two key points; i) R5M is proposed as the core assembly site of the DnaAATP oligomer in the left-half DOR, though the R1- DnaAATP still interacts with the DnaAATP oligomer in the left-half DOR. ii) The ssDUE directly interacts with the DnaAATP oligomer in the left-half DOR, through Val-211 and Arg-245 of domain III, known as the H/B motifs (153, 179, 182).

Figure 12: Current models for DUE unwinding. Only DnaA domain III and IV are shown. Sakiyama et al., 2017.

30 DnaB helicase loading

The next step in the replication initiation process, following DUE unwinding, is loading of the DnaB helicase. The functional DnaB helicase is a hexamer of identical DnaB monomers that forms a barrel shaped toroid structure (189-191). Replicative helicases, like DnaB, are molecular motors driven by ATP hydrolysis that are able to translocate along ssDNA and induce unwinding of duplex DNA in front of the moving replication fork (192). Loading of the DnaB hexamer onto the ssDUE is chaperoned by the DnaC helicase loader. Recent evidence suggests that the DnaB ring structure opens and closes and that binding of three to six DnaC molecules traps it in its open conformation, ready for loading onto the ssDUE (193). DnaB helicase loading has been proposed to happen independently for the left- and right- half DOR-DnaAATP complex, creating two distinct DOR-DnaAATP-DnaB complexes (173, 179, 182). The loading of the DnaB is mediated by interactions between DnaA domain I, including Glu-21 and Phe-46, and DnaB (161, 162, 194). For the second variant of the loop back model, it has been suggested that a DnaB-DnaC complex is initially loaded onto the ssDUE opposite of the DNA strand that interacts with the left-half DnaAATP-DOR complex. The loaded helicase then moves forward in the direction of the right-half DnaAATP-DOR complex, revealing a stretch of ssDNA available for DnaB loading, by the right-half DnaA- DOR complex, in the opposite direction and on the opposing strand of the other DnaB helicase (182, 195). Following loading of the DnaB helicase onto the ssDUE the DnaC molecules dissociates from the DnaB hexamer. The release of DnaC is suggested to be mediated by interactions between DnaB and the DnaG primase, stimulating the ATPase function of DnaC (196).

Regulation of the replication initiation

As mentioned above, the replication initiation is regulated to happen only once from each origin during a cell cycle. Even in rapidly dividing cells, where the oriC copy number per cell is higher than two, replication initiation at sister origins is triggered simultaneously and only once per cell cycle (139). Several regulatory systems are deployed during the cell cycle to ensure that replication initiation is

31 triggered in a timely manner (See Figure 13). These regulatory systems are described in the sections below.

Figure 13: An overview of the regulatory systems that ensures timely initiation of the DNA replication during the cell cycle. Katayama et al., 2017.

The dual role of DiaA in regulating replication initiation

DiaA is a 196 amino acid long protein that forms homo-tetramers in which each monomer holds a DnaA- binding site (163, 197). Observations that DiaA mutants initiates DNA replication asynchronously and that DiaA in vitro promotes replication of mini-chromosomes, led to the conclusion that DiaA is a DnaA associated factor that ensures timely initiation of the DNA replication process (163). A combination of mutational and crystallography studies revealed that the DiaA homo-tetramer can bind multiple DnaA molecules at once, and thereby stimulate DnaA oligomerization and DUE unwinding (197). The stimulatory effect of DiaA on replication initiation has been explained by a linker effect observed for several DNA binding proteins. By themselves the DNA binding proteins has a moderate affinity for DNA, but when they are linked, through a linker protein, their DNA affinity increases dramatically (198). Hence, DiaA linkage of DnaA molecules is suggested to increases the affinity of DnaA for DnaA boxes. Interestingly, both DiaA deletion and overproduction inhibits replication initiation in vivo, indicating that DiaA both has a positive and a negative effect on the initiation process. As described earlier, Phe-46, of DnaA domain I is both involved in binding of DiaA and the DnaB helicase, thus DiaA proposedly blocks the loading of the DnaB helicase by hindering the interaction between DnaA domain I and DnaB (162).

32 DiaA is believed to dissociate from the oriC-DnaA complex during DUE unwinding or closely after. However, the dissociation mechanism still needs to resolved (139).

Regulatory inactivation of DnaAATP (RIDA)

Following a successful round of replication initiation DnaAATP is converted to its inactive form DnaAADP. As described earlier, RIDA and DDAH are the two regulatory processes that are responsible for this conversion, though RIDA is the predominant one (199). In RIDA, the activity of the DnaA ATPase is stimulated by the DnaA homologue, Hda, in complex with the DNA loaded β sliding clamp (DnaN) of the DNA polymerase III holoenzyme (144, 200). Like DnaA, Hda is member of the AAA+ superfamily of proteins and holds a AAA+ module in its C-terminal (200), while the N-terminal is responsible for interactions with the β-clamp. The DnaA ATPase stimulatory effect of the DnaN-Hda complex is only active when Hda is bound by ADP (201). Arg-153 constituting the Arg-finger of the Hda AAA+ module is crucial for the function of RIDA (202). The binding of ADP to Hda likely triggers a conformational change in the Arg-finger, enabling interaction and activation of the ATPase region in domain III of DnaAATP (201). In addition, interactions between DnaA domain I and the C-terminal of Hda seems to stabilize the contact and promote the conversion of DnaAATP to DnaAADP (164). Recently, a crystal structure of a β- clamp-Hda complex from E. coli revealed insight into how the activation of RIDA might be regulated. Interestingly, the β-clamp- Hda complex was shown to form an octamer, where two pairs of Hda dimers were sandwiched by two β-clamp ring structures. Based on these findings, and additional biochemical and genetic evidence, it was proposed that the octameric complex negatively regulates RIDA, by encaging Hda. Additionally, it was suggested that loading the β-clamp with DNA, by the clamp loader, leads to dissociation of the octamer and formation of a DNA-β-clamp-Hda complex that is active in RIDA (203). The requirement for a DNA loaded β-clamp in activating RIDA neatly couples active DNA elongation with inhibition of the replication initiation (204). datA-dependent DnaAATP-hydrolysis (DDAH)

In 1996, Kitagawa et al. identified a novel high affinity DnaA binding region (later datA) at 94.7 min. on the E. coli chromosome, relatively close to the oriC (84.6 min.) (205). Shortly after its discovery, it was reported that deletion of the datA locus led to asynchronous initiation of replication and that a DnaA titrating plasmid suppressed the mutant phenotype. It was therefore proposed that datA repressed untimely initiation by titrating high amounts of DnaAATP following a round of replication initiation (145). In addition, an IHF binding site within the datA locus was shown to be important for maintaining a proper timing of replication initiation (206). It was first over a decade after the initial discovery of datA that the true mechanism by which datA regulates the timing in replication initiation was revealed.

33 Through a series of experiments, it was shown that datA in complex with IHF promotes hydrolysis of ATP bound to DnaA, through inter-DnaA interactions at the datA locus (170). DDAH functionality depends on a minimal datA locus of 183 bp containing the two high affinity DnaA boxes 2 and 3, the low affinity DnaA box 7 and a single IHF binding site (see Figure 14A) (145, 170, 206-208). In similarity to the individual DOR regions in the oriC, the essential DnaA boxes in datA are all orientated in the same direction, suggesting Figure 14: A) Schematic of the minimal datA locus needed for that DnaAATP bound at these sites interacts in a DDAH activity (DnaA box 4 is not essential). B) IHF induced ATP head to tail manner (208). Due to the long bending of datA leads to interactions between Dna bound to DnaA box 2 and 3, which lead to activation DDAH. From distance between DnaA box 2 and 3, DnaAATP Katayama et al., 2017. at these two sites cannot interact without binding of IHF. The binding of IHF to the IBS, which is situated between DnaA box 2 and 3, bends the DNA and brings DnaAATP bound to box 2 and 3 in close proximity, thereby enabling their interaction (see Figure 14B ) (208). The interaction between DnaAATP at box 7, 2 and 3 is mediated by domain III AAA+ Arg-finger and is further stabilized by Arg-281 and Leu-290. Furthermore, negative supercoiling of the DNA stabilizes DnaAATP- DnaAATP interactions and the binding of IHF (170, 208). It has been suggest that activation of DnaAATP hydrolysis at datA is promoted by conformational changes to the nucleotide binding pocket of the AAA+ module, induced by inter DnaAATP interactions via Arg-281 (208). Conversion of DnaAATP to DnaAADP is thought to destabilize the domain III-doamin III interaction mediated via Leu- 290, leading to release of the DnaAADP molecule from datA and loading of a new DnaAATP molecule.

ATP ADP Current evidence supports two distinct models for the conversion of DnaA to DnaA at datA. In one model, DnaAATP hydrolysis is only activated in DnaAATP bound to DnaA box 7. In the other, DnaAATP is hydrolysed at both DnaA box 7 and 2 (208). However, further research is needed to determine which of the two models that may be correct.

Regulation of DDAH activity

Binding of IHF to datA is essential for the timely activation of DDAH. Cell cycle analysis of IHF-datA complex formation indicated that IHF dissociates from datA before replication initiation and temporarily

34 binds to datA shortly after the DNA replication is initiated. This suggests that the activation of DDAH by IHF binding is tightly regulated by specific cell cycle events (139). Inhibiting transcription by treatment with is suggested to hinder dissociation of IHF from datA. As IHF is abundant in cells growing exponentially (209), the inhibitory effect of rifampicin treatment on IHF dissociation from datA, indicates that transcription in general or transcription of a specific factor is needed for inhibition of IHF- datA complex formation (139). This hypothesis is further backed by the fact that moving datA to a highly transcribed region on the chromosome inhibits the activity of DDAH (210). Hence, the timely binding of IHF to datA might be regulated by changes in transcription through datA from adjacent genes (139). Due to datAs close proximity to the oriC on the chromosome, it isreplicated shortly after replication initiation, leading to a temporary increase in copy number. The increase in datA copy number is believed to be important for repression of untimely replication initiation (205), which is in agreement with the observation that a four-fold increase in the datA copy number delays replication initiation (211). In contrast, datA deletion or transversal of the datA locus to the terminus region allows for untimely replication initiation (145, 210). As mentioned above DNA supercoiling of datA promotes the activity of DDAH (170, 208). In addition, an increase in untimely replication is observed for a datA deletion mutant grown in nutrient poor-medium, relative to rich-medium (145). Indicating that the nutritional state of the cell might influence DDAH activity. This theory was further established by analysis of the chromosomal conformation during amino acid starvation, where datA was shown to interact with the oriC (212).

SeqA, a negative regulator of the replication initiation

In addition to the conversion of DnaAATP to DnaAADP by RIDA and DDAH, replication initiation is also negatively regulated by the SeqA protein, which sequesters hemi-methylated GATC sites DNA in the oriC after DNA replication has been initiated. In E. coli the Dam adenine methylase (Dam) methylates GATC sites in the DNA. In newly replicated DNA only the parental strand is fully methylated while the daughter strand is unmethylated, referred to as hemi-methylated DNA. Based on the early findings that Dam deficient cells could not be efficiently transformed with mini-chromosomes unless it was methylated and that fully methylated plasmids were only replicated one round in dam- cells (213, 214). It became evident that the hemi-methylated state of the DNA somehow was involved in the regulation of the DNA replication (215). Interestingly, it was shown that hemi-methylated DNA could be replicated in vitro. Indicating that in vivo replication of hemi-methylated DNA was inhibited by an unknown factor, rather than the hemi-methylation itself (216, 217). The unknown factor was later identified as SeqA in screens for dam- mutants that could be transformed with, and maintain, a fully methylated mini-chromosome (146, 147).

35 SeqA sequestrates the oriC for approximately one-third of the cell cycle and ensures that a new round of replication is not triggered untimely at the newly replicated origins (217). The ratio of DnaAATP to DnaAADP reaches its maximum at initiation and gradually decreases due to RIDA and DDAH activity. In cells with a doubling time of 30 minutes, it takes approximately ten minutes to decrease the DnaAATP to DnaAADP ratio to a level that prevents replication initiation. Hence, SeqA sequestrates the oriC for a period equal to the time needed by RIDA and DDAH to prevent initiation by decreasing the DnaAATP/ DnaAADP ratio to a certain threshold value (218-220). The minimal oriC contains 11 GATC sites, which is significantly more than the average random distribution of GATC sites on the rest of the chromosome. In vitro the binding of SeqA to hemi-methylated GATC sites in the oriC blocks DnaAATP binding to DnaA box R5M, I2 and I3 that are all three overlapping with GATC sites (221, 222). Conversely, the sequestration of the oriC by SeqA does not interfere with binding of DnaAATP/ADP to the high affinity DnaA boxes R1 and R4 and the moderate affinity DnaA box R2 (222), which is in agreement with the observation that DnaA occupies these three sites throughout most of the cell cycle(154, 155). Furthermore, evidence show that the period of hemi-methylation of the oriC is reduced when the available amount of DnaA is decreased (223).Though, the molecular mechanism that links the sequestration period with the DnaA concentration is still not known. However, direct interaction between SeqA and DnaA has been proposed to stabilize the sequestration of the oriC, though such an interaction remains to be proven (223). Alternatively, DnaA binding to DnaA boxes overlapping GATC sites in newly replicated origins might protect from methylation by Dam. A DnaA-SeqA exchange at the oriC is then suggested to be mediated by an increase in allosteric DnaA binding sites due to ongoing replication (217). At high concentrations, SeqA binding to DNA inhibits the formation of negative supercoils, this inhibitory effect has been proposed to counteract unwinding of the DUE by DnaAATP (224). In addition to the sequestration of the oriC, SeqA also binds to GATC sites in the dnaA promoter following its replication (215). This binding inhibits the transcription of dnaA and thereby contributes to the accurate timing of the replication initiation (225, 226).

Rejuvenation of the cellular DnaAATP pool

When it is time for the cell to prepare a new round of replication initiation, the DnaAATP level is increased by three mechanisms. One is de novo synthesis of DnaA, which then bind ATP readily available in the cytosol. The second and third are distinct pathways that lead to dissociation of ADP from DnaAADP and subsequent binding of ATP. This process is mediated by either phospholipids in the cell membrane or a pair of specific chromosomal DNA elements known as DnaA-reactivating sequences, DARS1 and DARS2 (139).

36 DARS1 and DARS2 As mentioned, the dissociation of ADP from DnaAADP is promoted by DARS1 and DARS2, subsequently leading to regeneration of DnaAATP and stimulation of replication initiation (148, 227). Even though DARS2 is more than four times the length of DARS1, 455 bp versus 101 bp, both elements contain a similar core region of three DnaA boxes (I, II and III) that are bound mainly by DnaAADP. The regulatory region is the major factor that Figure 15: A) Schematic of the DARS2 region, In light blue DnaA box I-IV, in differentiates the two DARS green the IHF binding site (IBS) and in orange the Fis binding sites (FBSs). B) The ADP elements. DARS1 has a small ≈50 DnaA -IHF-Fis complex at DARS2. Modified from Katayama et al., 2017. bp regulatory region, in contrast to the ≈400 bp in DARS2 (139). The differences in the regulatory regions leads to an important functional difference of DARS1 and DARS2. In vitro, DARS1 is able to mediate the dissociation of ADP from DnaAADP without any additional factors, while DARS2 activity is significantly stimulated by interaction of its regulatory region with IHF and Fis (see Figure 15AB) (148, 227). In vivo, both DARS regions promotes DnaAATP production and replication initiation, as the deletion of either delays the commencement of the replication (148, 227). However, deletion of DARS1 effects the timing of the replication initiation less than deletion of DARS2, indicating that DARS2 promotes the production of DnaAATP to a higher degree than DARS1. In addition, increasing the copy number of DARS2 leads to a more severe over-initiation, than an increase in DARS1 (148, 227). The observed difference between the activity of DARS1 and DARS2 is likely, due to a promoting effect of IHF and Fis on the number of DnaAADP molecules that oligomerizes at DARS2, as shown by pull-down assays (227). Owing to the difference in the activity of DARS1 and DARS2, it is believed that DARS2 is important for timing the replication initiation, while DARS1 might act to maintain a basal level of DnaAATP in the cell (139). In both DARS1 and DARS2, DnaA box I is orientated in the opposite direction of DnaA box II and III. Therefore, DnaAADP at DnaA box I and II interacts in a head to head manner, in contrast to the head to tail interactions observed at oriC and datA. Even though the DnaA box core region of both DARS1 and DARS2 is arranged in a similar manner, the events leading to oligomerization and ADP dissociation are likely not identical. DnaA mutant analysis demonstrated that a D269N mutant was

37 deficient in both DnaAADP oligomerization and ADP dissociation at DARS1, but not at DARS2 (148, 227). This difference is likely caused by unknown functions of IHF and Fis at DARS2 (139) Conversely, the ADP dissociation via DARS1 was unaffected by a R334A mutation, while DARS2-mediated ADP dissociation was moderately impaired by this mutation (148, 227). Like for inter DnaAATP-DnaAATP interactions at the oriC Leu-290 is essential for the the oligomerization of DnaAADP at DARS2 (227). It remains to be investigated, if Leu-290 is essential for DnaAADP oligomerization at DARS1. In the current mechanistic model for DARS2-mediated ADP dissociation from DnaAADP. A DnaAADP oligomer forms at the DnaA box core region, while IHF and Fis binds to their respective sites in the regulatory region. The binding of IHF bends the DNA, which promotes the interaction between Fis and the DnaAADP oligomer at the core region. The resultant DnaAADP-IHF-Fis complex (see Figure 15B) induces conformational changes in DnaAADP leading to dissociation of ADP. The apo-DnaA then dissociates from the DARS2 complex and binds to free ATP in the cytosol (139). The activation of DARS2-mediated rejuvenation of the cellular DnaAATP pool is regulated in a cell cycle coordinated manner by binding of IHF. IHF binds DARS2 in the pre-initiation period and dissociates again just before the replication is initiated (227). Unlike IHF dissociation from datA, IHF binding and release from DARS2 is resistant to inhibition of the transcription. Furthermore, the binding of IHF is not coupled to replication initiation, as the binding and dissociation still occurs under conditions where replication initiation is blocked. In light of these observations it is suggested that IHF-DARS2 interactions are regulated by an unknown cell-cycle dependent pathway that is uncoupled from the regulation of the replication initiation (227). In contrast to IHF, Fis binds DARS2 throughout the cell cycle and proposedly couples the replication initiation with the growth phase of the cell(227). As Fis is abundant in cells growing exponentially, but scarce in stationary phase cells (209).

Phospholipid mediated reactivation of DnaAADP The first evidence that phospholipids were involved in reactivation of DnaAADP, by mediating the release of ADP, was published in 1988. Here it was shown that in vitro cardiolipin, an acidic phospholipid fund in the E. coli cell membrane, interacted with DnaA and mediated the release of both ADP and ATP (228). Subsequently it was demonstrated that mixtures of phospholipids and fluidic membranes also promoted nucleotide dissociation from DnaA (229, 230). Furthermore, DnaA/oriC independent replication, known as constitutive stable DNA replication (cSDR), suppressed the growth arrest observed in an E. coli strain depleted for acidic phospholipids (231). In the same strain, it was demonstrated that the growth arrest was also suppressed by expression of a DnaA L366K mutant (232). However, it remains unknown how DnaA L366K suppresses the growth arrest. Flow cytometry analysis revealed a simultaneous shutdown of the DNA replication and the protein synthesis in cells depleted of acidic phospholipids. Indicating,

38 that phospholipid mediated regulation of replication initiation might be part of a globular response system (233). Nonetheless, more research is required to elucidate the mechanisms that lead to regulation of replication initiation by phospholipids.

De novo synthesis of DnaA The final known mechanism that is involved in increasing the DnaAATP level is de novo synthesis of DnaA. The newly synthesized DnaA molecules bind to ATP that is abundant in the cytosol and are thereby ready to partake in initiating the DNA replication. As described above, SeqA sequestration of the dnaA promoter inhibits its transcription shortly after it has been replicated (215, 225). The sequestration of the dnaA promoter by SeqA is proposedly auto-regulated by DnaA binding to DnaA boxes in the promoter region, which stabilizes the sequestration (217). The dnaA promoter stays hemi-methylated for approximately one sixth of the cell cycle following replication initiation; permitting initiation of RIDA and DDAH activity and replication of DnaA titration sites on the chromosome (215, 234). Furthermore, translocation of dnaA further away from the oriC leads to asynchronous replication initiation, explained by an increase in the available amount of DnaA at the end of the oriC sequestration period. Hence, the coordination between the periods of SeqA sequestration of both the oriC and the dnaA promoter is crucial in timing the replication initiation (226).

The lethal action of severe over-initiation of the DNA replication

The importance of RIDA and DARS in regulating the replication initiation is emphasized by the severe growth retardation and over-initiation observed in hda mutants, deficient in RIDA, and cells carrying multiple copies of DARS2 (200). Evidence show that Hda deficient cells are viable under anaerobic conditions or if the GO repair system is impaired (235). The GO system is involved in prevention and repair of 8-oxo-dGTP incorporation in the DNA. Incorporation of 8-oxo-dGTP in the DNA is potentially mutagenic due to its ability to form base pairs with both cytosine and adenine (236).The GO repair system consists of at least three proteins, MutM, MutT and MutY. MutT acts by hydrolyzing 8-oxo-dGTP to 8-oxo-dGMP, disabling its incorporation into the DNA. The excision of 8-oxo-dGTP already incorporated into the DNA is mainly carried out by the formamidopyrimidine glycosylase, encoded by mutM. Finally, the glycosylase activity of MutY enables it to remove adenines inserted opposite incorporated 8-oxo-dGTPs (236). If 8-oxo-dGTPs are closely spaced in the DNA or encountered by replication forks during repair, they may cause DSBs in the DNA (57). Based on the observations described above and the fact that over-initiating cells have an increased number of ongoing replication forks. It was suggested that the lethal effect of over-initiating replication is due to the formation of DSBs when replication forks encounters 8-oxo-dGTP lesions that are under repair by the GO system (235).

39 Two other models have been suggested for how over-initiation leads to accumulation DSBs in the DNA. In one model, it is proposed that ongoing replication forks collide with forks that have been stalled, which leads to replication fork collapse and DSBs (237). In the second model, dNTP starvation, due to a high number of replication forks, is suggested to lead to accumulation of DSBs in the DNA (238, 239). In hda- mutants secondary mutations quickly arises. These mutations are known as hda suppressor mutations (hsm), as they suppress the over-initiating phenotype of hda- cells. Several of the identified hsm directly affects the replication initiation or the oriC itself, thereby diminishing the over- initiation of the hda- cells (125, 240-242), while others permit growth despite of over-initiation (235, 243, 244). The later includes mutations in iscU and fre encoding an iron-sulfur cluster assembly protein and the flavin reductase, respectively (244). Differential gene expression analysis of the iscU and fre mutants by micro-array demonstrated a down regulation of genes involved in the TCA cycle and the aerobic respiratory chain, while genes involved in the micro-aerobic respiratory chain were up- regulated. Indicating a rerouting of the electron flow from the aerobic respiratory chain to the micro- aerobic respiratory chain (243). The effect of such a rerouting is a decrease in the generation of ROS, which in turn leads to a decrease in 8-oxo-dGTP formation and its repair. Hence, decreasing the ROS production enables unhindered progression of replication forks in over-initiating cells (235, 243). In paper II of this thesis, we further verify the proposed model for the lethal action of over-initiating the DNA replication. As during a screen for replication initiation inhibitors, using over-initiating cells, we identify the iron chelator deferoxamine, a known inhibitor of ROS production via Fenton chemistry (245, 246), as a compound that rescues the growth over-initiating cells by enabling fork progression during hyper-replication.

Targeting the Initiation of replication

Multiple compounds have been identified that target different parts of the DNA replication machinery, including, DNA ligase (247, 248), DNA polymerase III (249, 250), the β-sliding clamp (251, 252) and single-stranded DNA-binding proteins (253). However, screening for putative inhibitors of the replication initiation process have been limited and so far unsuccessful. A single screen for replication initiation inhibitors has been published. This screen is based on a conditional lethal, cold sensitive DnaA E. coli mutant that over-initiates replication. Thus, inhibition of replication initiation, at non-permissive conditions, restores growth (254). Subjecting the screen to a library of pharmacological active compounds (LOPAC), did not lead to the discovery of any replication initiation inhibitors. However, the benzazepine derivative, (±)-6-Chloro-PB hydrobromide (S143), was identified as a novel gyrase inhibitor that rescues the growth of over-initiating cells (255). Despite the current lack of success in identifying compounds that blocks replication initiation, there is evidence that the initiation of chromosomal DNA

40 replication is a druggable process. As an inhibitor has been identified for the distinct replication initiation process of the second chromosome in the Vibrionaceae family of bacteria (256). In addition, expression of a cyclic DnaA domain I or over-expression of DnaA domain IV and I lead to inhibition of the replication initiation, most likely by interfering with DnaA oligomerization at the oriC (252, 257). In paper II and III of this thesis, we present two distinct strategies for identifying replication initiation inhibitors.

41

Paper I: Can Ciprofloxacin Resistance be Reversed by Helper Drugs?

Currently in review at: Annals of Clinical Microbiology and .

42 1 Can Ciprofloxacin Resistance be Reversed by Helper

2 Drugs?

3 Rasmus N. Klitgaard 1, Bimal Jana2, Luca Guardabassi2, Karen Leth Nielsen3 and Anders Løbner-

4 Olesen 1,*

5 1 Department of Biology, Section for Functional Genomics, University of Copenhagen, Copenhagen,

6 Denmark.

7 2 Department of Veterinary and Animal Sciences, Section for Veterinary Clinical Microbiology, University of

8 Copenhagen, Denmark.

9 3 Department of Clinical Microbiology, Center for Diagnostics, Rigshospitalet, Copenhagen, Denmark.

10 * [email protected]; Tel: +4535322068

11 Academic Editor: name

12 Received: date; Accepted: date; Published: date

13 Abstract

14 Background

15 Fluoroquinolones such as ciprofloxacin are potent antibacterial drugs that are widely used in the

16 clinic. As a consequence of their extensive use, resistance has emerged in almost all clinically

17 relevant bacterial species. In an attempt to reverse ciprofloxacin resistance, we searched for potential

18 helper drug targets in Escherichia coli strains with different levels and mechanisms of ciprofloxacin

19 resistance.

20 Methods

21 The search for ciprofloxacin helper drug targets was conducted by a combined transcriptomic and

22 genetic approach. Differential gene expression (DGE) analysis of the high-level ciprofloxacin

23 resistant E. coli sequence type (ST) 131 UR40 strain, treated with 2 µg/ml ciprofloxacin, was done by

43 24 RNA-Seq. In the genetic screen 23 single gene deletions were transduced from the Keio collection

25 into a high ciprofloxacin resistant E. coli strain LM693 (carrying gyrAS83L, gyrAD87N and parCS80I

26 mutations), followed by determination of the minimal inhibitory concentration (MIC) by Etest. The

27 seven individual gene deletions that lowered the ciprofloxacin MIC the most were subsequently

28 introduced into strains LM862/pRNK1 and LM862/pRNK9 carrying the aac(6´)-Ib-cr and qnrS genes,

29 which confer low-level ciprofloxacin resistance by drug modification and target protection,

30 respectively. The ciprofloxacin MICs were then determined for these two strains by broth micro-

31 dilution.

32 Results

33 Differential gene expression analysis of ST131 UR40 treated with ciprofloxacin, showed that the

34 transcriptome was similar to that of untreated samples, i.e. no genes were found to be significantly

35 upregulated. The genetic screen of the 23 single gene deletions in LM693 identified a number of

36 genes that significantly lowered the ciprofloxacin MIC, including genes encoding the AcrAB-TolC

37 efflux pump, SOS-response genes and the global regulator fis. However, none of the deletions

38 lowered the MIC beneath the clinical breakpoint. In the low-level resistant strains carrying aac(6´)-

39 Ib-cr and qnrS, respectively, deletion of acrA, tolC, recC or recA all rendered the strains clinically

40 susceptible to ciprofloxacin.

41 Conclusions

42 The results of the combined transcriptomic and genetic approach show that it is not straightforward

43 to reverse ciprofloxacin resistance in high-level ciprofloxacin resistant E. coli strains. On the other

44 hand, components of AcrAB-TolC efflux pump and the SOS response proteins, RecA and RecC were

45 identified as possible helper drug targets in E. coli strains with a MIC closer to the clinical

46 breakpoint.

47 Keywords: Antibiotic resistance, ciprofloxacin, helper drugs, RNA-Seq, transcriptomics.

44 48

49 Background

50 Fluoroquinolones are some of the most prescribed antibacterial drugs in the world [1-3], but this

51 has not always been the case. For the first two decades after the discovery of nalidixic acid in 1962,

52 and its introduction into the clinic in 1964, the quinolones were only used to treat uncomplicated

53 urinary tract infections. This changed with the release of the second generation quinolones,

54 including ciprofloxacin, which showed significant activity outside the urinary tract and against a

55 broad spectrum of both Gram-negative and Gram-positive bacteria. Ciprofloxacin acts by binding

56 to its targets, gyrase and topoisomerase IV, inhibiting the native ability of these two enzymes to re-

57 ligate double stranded DNA breaks, in turn leading to fragmentation of the chromosome. Due to its

58 mechanism of action it is sometimes referred to as topoisomerase poison[4]. Inevitably, considering its

59 extensive use and misuse, resistance towards ciprofloxacin has arisen in almost all clinically

60 relevant bacteria [5, 6]. One method to overcome antibacterial resistance is by combinatorial

61 treatment with a potentiating compound, also known as a helper drug. A helper drug is by

62 definition non-antibacterial when administered alone, but it enhances the activity of the antibiotic

63 when used in concert. The potentiating effect of a helper drug can be achieved by either direct

64 inhibition of the resistance mechanism or by targeting endogenous cellular components and

65 pathways like, cell membranes, efflux pumps and cellular repair systems. A classic example of

66 targeting the resistance mechanism is the combination of amoxicillin and the β-lactamase inhibitor

67 clavulanic acid [7]. In Gram-negative bacteria, high-level ciprofloxacin resistance is mainly

68 associated with multiple target site mutations in gyrA and parC, encoding subunits of the DNA

69 gyrase and topoisomerase IV, respectively. Since 1998 three different plasmid-mediated

70 ciprofloxacin resistance mechanisms have been identified; i) target protection (Qnr proteins), ii)

71 efflux pumps (QepA and OqxAB) and iii) drug modification (AAC(6´)-Ib-cr acetyltransferase)[8].

45 72 Here, we used a combined transcriptomic and genetic approach to identify potential helper drug

73 targets in Escherichia coli strains with different levels and mechanisms of ciprofloxacin resistance.

74 Methods

75 Bacterial strains and plasmids

76 Strains LM693 and LM862 were obtained from Diarmaid Hughes from Uppsala University.

77 LM693 is isogenic to MG1655 besides two gyrA mutations, S83L and D87N, and one parC mutation,

78 S80I. LM862 is also isogenic to MG1655, but with one gyrA S83L mutation and one parC S80I

79 mutation. ST131 UR40 has two gyrA mutations, S83L and D87, and two parC mutations, S80I and

80 E84V, and carries aac-6´-Ib-cr on a plasmid[9]. The aac-6´-Ib-cr carrying plasmid pRNK1 was

81 constructed as follows: aac-6´-Ib-cr gene was amplified by PCR from ST131 UR40, using the

82 following primers: GATCGGATCCATGAGCAACGCAAAAACAAAGTTAGGC and

83 CATCGAATTCTTAGGCATCACTGCGTGTTCGC, and cloned into pMW119 (Nippon Gene,

84 Toyama, Japan) using BamHI and EcoRI. The qnrS-carrying plasmid pRNK9 was constructed as

85 follows: qnrS was amplified by PCR from the clinical E. coli isolate EC38 using the following

86 primers: GATCGGATCCATGGAAACCTACAATCATACATATCGGC and

87 GATCAAGCTTTTAGTCAGGATAAACAACAATACCCAGTGC, and cloned into pMG25 using

88 BamHI and HindIII (M. Mikkelsen and K. Gerdes, unpublished). Strain EC38 was isolated from a

89 patient with a urinary tract at Hvidovre Hospital, Denmark.

90 Genetic screening and MIC tests

91 For the genetic screen, P1 phage lysates were prepared from the relevant Keio collection strains

92 [10] and used for transduction into LM693 and LM862. All the transduced strains were verified by

93 PCR. Theciprofloxacin MICs for LM693 and derived strains were determined using E-tests (0.002-32

94 µg/ml, BioMerieux) and according to the manufactures guidelines. The MICs for LM862 and derived

46 95 strains were determined by broth micro-dilution using cation adjusted Mueller Hinton broth II with

96 1mM and 10 µM IPTG for pRNK1 and pRNK9 respectively. The reference E. coli strain ATCC 25922

97 was used as standard in all MIC tests and the susceptibility was evaluated according to CLSI

98 breakpoints.

99 Checkerboard assay

100 All wells in a micro-titter plate were filled with 100 µl cation adjusted Mueller Hinton broth II (200

101 µL in the negative control wells). Copper phtalocyanine-3, 4´, 4´´,4´´´-tetrasulfonic acid, was added

102 to the first row, followed by serial dilution along the abscissa, leading to a start concentration of 100

103 µM. Hereafter ciprofloxacin was serial diluted along the ordinate, giving a start concentration of 2

104 µg/ml and 64 µg/ml for LM862 and LM693, respectively. 100 µl diluted culture with an OD600 of

105 0.001 was then inoculated in each well and the plates were incubated at 370C for 24 hours.

106 RNA-sequencing

107 ciprofloxacin was added to a balanced ST131 UR40 culture to a final concentration of 2 µg/ml.

108 Samples for RNA isolation were taken at 0 minutes (prior to ciprofloxacin addition) and 30 and 90

109 minutes after ciprofloxacin addition. Total RNA was isolated using a Thermo Scientific GeneJET

110 RNA isolation kit. Dnase treated with TURBO DNA-free kit from Ambion. rRNA was depleted using

111 an Illumina Ribo-zero rRNA removal kit, followed by RNA-Seq library prep using an Illumina

112 TruSeq Stranded mRNA Library Prep Kit. Sequencing was performed on an Illumina Miseq with a

113 Miseq reagent kit v3. (75bp paired-end) from Illumina. Data analysis was performed in Rockhopper

114 ver.2.03[11]. E. coli NA114 (ST131) (accession number: NC_017644) was used as reference genome[12].

115 Results

116 Identification of helper drug targets by genetic screening

47 117 More than 25 single gene knockouts have already been shown to increase ciprofloxacin

118 susceptibility in wildtype E. coli strains [13-16]. Here, 23 of these deletions were introduced into the

119 high ciprofloxacin resistant strain LM693 [17] and tested for hyper-susceptibility towards

120 ciprofloxacin (Table 1). LM693 is isogenic to the commonly used laboratory strain MG1655 besides

121 two gyrA mutations; S83L and D87, and one parC mutation; S80I. Even though nine of the mutant

122 strains showed a three to four fold reduction in MIC , none of them were lowered beneath the CLSI

123 clinical breakpoint of 1 µg/ml. Our results therefore indicate that none of the tested gene-knockouts

124 identify valid helper drug targets in high-level ciprofloxacin resistant E. coli strains but could

125 potentially be used as helper drug targets to reverse low-level resistance. To create low-level

126 ciprofloxacin resistant strains, we constructed plasmids pRNK1 and pRNK9 carrying the

127 ciprofloxacin resistance determinants aac-6´-Ib-cr and qnrS, respectively. AAC-6´-Ib-cr inactivates

128 ciprofloxacin by N-acetylation of the amino nitrogen of its piperazinyl substituent [18], while QnrS

129 acts as a DNA mimic, binding to and protecting the gyrase from the action of ciprofloxacin[8].

130 Introduction of pRNK1 and pRNK9 into strain LM862, which carries gyrA S83L and parC S80I

131 mutations, increased the MIC from 1 to 2 µg/ml, i.e. above the clinical breakpoint. We then evaluated

132 the ability of seven of the most promising of the 23 gene deletions described above to reduce

133 ciprofloxacin resistance. Four of the deletions (acrA, tolC, recA and recC) lowered the MIC beneath the

134 clinical break point. (Table 2). To assess whether inhibition of RecA was an amenable strategy for

135 potentiation of ciprofloxacin, synergy between ciprofloxacin and a RecA inhibitor, copper

136 phtalocyanine-3, 4´, 4´´,4´´´-tetrasulfonic acid[19], was tested by a checkerboard assay. No reductions

137 of the ciprofloxacin MICs were observed for either of the high- (LM693) -or low-level (LM862)

138 resistant strains.

48 Strain/Single MIC(µg/ml) MIC (µg/ml) deletions Strain pRNK1 pRNK9 LM862 (No plasmid) 1 1 LM693 24-32 LM862/Empty vector 1 1 tolC 1.5 LM862 2 2 acrA, acrB and fis 2 tolC 0.25 0.5 recC, xseA, xseB, uvrD 4 and recA acrA 0.25 0.5 ruvC and dksA 6 recA 0.5 0.5 recG and hlpA 8 recC 0.5 0.5 pgm, ybgF and ybgC 12 uvrD 2 1 deoR, ydcS, yciT, ybjQ 16 xseA 1 1 ygcO and nlpC 24 rimK 24-32 fis 2 4

Table 1. MIC values for the single gene Table 2. MIC values for the single gene 139 deletions in LM693. deletions in LM862/pRNK1 and LM862/pRNK9

140 Identification of helper drug targets by RNA sequencing

141 The E. coli clonal group sequence type (ST) 131 has become the predominant E. coli lineage

142 isolated from extra-intestinal infections and is currently regarded a global problem in hospitals and

143 clinical practices. Two independent studies have shown that more than 90% of ESBL-producing

144 ST131 isolates are also resistant to ciprofloxacin [20, 21]. Strain ST131 (UR40) is resistant to high levels

145 of ciprofloxacin due to gyrA mutations S83L and D87, and parC mutations S80I and E84V [9]. Here

146 we used RNA-Seq to map the transcriptomic changes during treatment of ST131 UR40 with a

147 clinically relevant concentration of ciprofloxacin (2 µg/ml). The rationale behind this was to identify

148 potential helper drug target genes that are over-expressed upon ciprofloxacin exposure and

149 putatively involved in ciprofloxacin resistance. In contrast to the genetic screen, the RNA-Seq analysis

150 would also reveal targets encoded by essential genes and non-coding RNA. The transcriptomic

151 analysis did not show any non-ribosomal transcripts to be significantly upregulated in the presence

152 of ciprofloxacin, i.e with a false discovery rate of <1% and more than 2-fold expression change.

49 153 Discussion

154 By utilizing a combination of “direct genetic screening “and differential gene expression analysis, we

155 have attempted to identify genes suitable as targets for ciprofloxacin potentiating compounds. We

156 did not find any genes to be significantly up-regulated by ciprofloxacin, indicating that the

157 transcriptome of ST131 UR40 was fairly unaffected by treatment with a sub-inhibitory and yet

158 clinically relevant concentration of ciprofloxacin. The ciprofloxacin is most likely not binding to its

159 target, pumped out by efflux pumps or inactivated by Aac-6’-Ib-cr. The lack of an upregulation of the

160 SOS response genes in the transcriptomic analysis clearly shows that the ciprofloxacin exposure did

161 not cause sufficient DNA damage to induce a SOS response; hence it was not necessary for ST131

162 UR40 to up-regulate any specific genes to cope with the presence of ciprofloxacin at a sub-inhibitory

163 concentration.

164 The screening of selected mutant strains revealed a number of genes, which when deleted,

165 lowered the MIC for ciprofloxacin significantly. These findings are in accordance with genes reported

166 to contribute to high-level ciprofloxacin resistance by Tran et al. [22]. Treatment of bacteria with

167 ciprofloxacin generates double stranded breaks in the DNA of the bacteria [23], which in turn

168 activates the SOS response. Seven of the tested gene deletions; recA, recC, recG, uvrD, xseAB and ruvC,

169 which all significantly reduced the MIC of LM693, are part of the SOS response and involved in DNA

170 damage repair [24-26]. The results from the MIC analyses indicate that deletion of any of these seven

171 genes lowers the ability of the bacteria to cope with ciprofloxacin induced DNA damage. Deletion of

172 genes encoding the AcrAB-TolC efflux pump, or the global regulator Fis (Factor for inversion

173 stimulation) showed the largest decreases in MIC values relative to LM693. The Fis protein has been

50 174 shown to repress the gyrA and gyrB promoters, thereby reducing the expression of the DNA gyrase

175 [27]. Deletion of fis therefore increases DNA gyrase expression and the number of ciprofloxacin

176 targets. As ciprofloxacin works as a topoisomerase poison, an increase in ciprofloxacin bound DNA

177 gyrase could potentially lead to an increase in double stranded breaks, and this could explain the

178 decrease in MIC for the fis deletion strain. The fis deletion did not have the same effect in the low-

179 resistant strains LM862/pRNK1 and LM862/pRNK9, which may be explained by the relatively higher

180 affinity of ciprofloxacin for its target in LM862, compared to that of LM693. Hence, the increase in

181 expression of the DNA gyrase might lead to an increase in ciprofloxacin-gyrase complexes, but if the

182 ciprofloxacin induced DNA damage already is at a level, where the DNA repair mechanisms cannot

183 keep up, the fis deletion does not have a dramatic effect on the MIC.

184 Individual deletions of acrA, acrB or tolC genes encoding the AcrAB-TolC efflux pump had a

185 large effect on the ciprofloxacin susceptibility of both LM693 and LM862 strains. This was not

186 surprising as overexpression of the AcrAB-TolC efflux system has been connected to ciprofloxacin

187 resistance numerous times [28]. The deletion of acrA or tolC in the LM862 strains lowered the MIC

188 beneath the clinical breakpoint indicating that AcrAB-TolC efflux system is a potential target for

189 ciprofloxacin potentiating compounds in low level resistant E. coli. A number of AcrAB-TolC

190 inhibitors have been identified [29-33], two of which have been shown to decrease the MIC of

191 ciprofloxacin in susceptible E. coli strains [29, 30], but none of them are used in clinical practice so far.

192 Inhibition of RecA and thereby of the SOS response has been proposed as a strategy to fight

193 antibiotic resistance numerous times [19, 34, 35]. Our finding; that deletion of RecA in low-level

194 resistant strains of E. coli lowers the MIC beneath the clinical break-point, is in accordance with recent

51 195 observations by Recacha et al.[36]. Combined, this indicates that RecA could be a potential

196 ciprofloxacin helper drug target.

197 Even though deletion of AcrAB-TolC or RecA rendered LM862/pRNK1 and LM862/pRNK9

198 clinically susceptible to ciprofloxacin, the respective MICs were only 2 to 4-folds lower than the

199 clinical break-point. It therefore seems reasonable to assume that a given inhibitor should completely

200 block the activity of either RecA or AcrAB-TolC in order for it to be an efficient helper drug. This

201 hypothesis is backed by the failure of lowering the ciprofloxacin MIC of LM862 and LM693 with the

202 relatively poor RecA inhibitor phtalocyanine-3, 4´, 4´´,4´´´-tetrasulfonic acid. Overall, it might

203 therefore prove difficult to reverse ciprofloxacin resistance by helper drugs targeting the proteins

204 encoded by the genes tested in this study.

205 Conclusions

206 The combined transcriptomic and genetic approach show that it may be difficult to reverse

207 ciprofloxacin resistance in high-level resistant E. coli strains. However, the components of the AcrAB-

208 TolC efflux pump along with the SOS response proteins RecA and RecC were identified as putative

209 targets for reversing resistance in low-level ciprofloxacin resistant strains. The only described RecA

210 inhibitor working in vivo, phtalocyanine-3, 4´, 4´´,4´´´-tetrasulfonic acid, was found unable to reverse

211 resistance, suggesting that it did not inhibit RecA to a degree sufficient to re-sensitize cells to

212 ciprofloxacin.

213 Abbreviations

214 MIC: Minimal inhibitory concentration, ST: Sequence type.

215 Declarations

52 216 Ethics approval and consent to participate

217 Not applicable.

218 Consent for publication

219 Not applicable.

220 Availability of data and materials

221 The RNAseq datasets generated and analyzed during the current study are available in the Gene

222 expression Omnibus, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE89507

223 Competing interests

224 The authors declare that they have no competing interests.

225 Funding

226 Study was funded with financial support from the University of Copenhagen Centre for Control of

227 Antibiotic Resistance (UC-Care) and by the Center for Bacterial Stress Response and Persistence

228 (BASP) funded by a grant from the Danish National Research Foundation (DNRF120).

229 Authors´ contributions

230 RNK carried out all experimental work, designed the study, analysed the data and prepared the

231 final manuscript. ALO supervised all aspects of the study and helped prepare the final manuscript.

232 BJ assisted and supervised the experimental part of the RNAseq. LG supervised and delivered the

233 ST131 UR40 strain. KLN performed genomic analyses and delivered the EC38 strain carrying the qnrS

234 gene. All authors read and approved the final manuscript

235 Acknowledgments

53 236 We acknowledge the financial support from the University of Copenhagen Centre for Control of

237 Antibiotic Resistance (UC-Care) and by the Center for Bacterial Stress Response and Persistence

238 (BASP) funded by a grant from the Danish National Research Foundation (DNRF120).

239 References 240 1. Mitscher LA: Bacterial topoisomerase inhibitors: quinolone and pyridone 241 antibacterial agents. Chemical reviews 2005, 105(2):559-592. 242 2. Linder JA, Huang ES, Steinman MA, Gonzales R, Stafford RS: Fluoroquinolone 243 prescribing in the United States: 1995 to 2002. The American Journal of Medicine, 244 118(3):259-268. 245 3. Emmerson AM, Jones AM: The quinolones: decades of development and use. The 246 Journal of antimicrobial 2003, 51 Suppl 1:13-20. 247 4. Aldred KJ, Kerns RJ, Osheroff N: Mechanism of quinolone action and resistance. 248 Biochemistry 2014, 53(10):1565-1574. 249 5. Werner NL, Hecker MT, Sethi AK, Donskey CJ: Unnecessary use of 250 fluoroquinolone antibiotics in hospitalized patients. BMC Infectious Diseases 251 2011, 11:187-187. 252 6. Dalhoff A: Global Fluoroquinolone Resistance Epidemiology and Implictions for 253 Clinical Use. Interdisciplinary Perspectives on Infectious Diseases 2012, 2012:37. 254 7. White AR, Kaye C, Poupard J, Pypstra R, Woodnutt G, Wynne B: Augmentin 255 (amoxicillin/clavulanate) in the treatment of community-acquired respiratory 256 tract infection: a review of the continuing development of an innovative 257 antimicrobial agent. The Journal of antimicrobial chemotherapy 2004, 53 Suppl 258 1:i3-20. 259 8. Rodríguez-Martínez JM, Machuca J, Cano ME, Calvo J, Martínez-Martínez L, 260 Pascual A: Plasmid-mediated quinolone resistance: Two decades on. Drug 261 Resistance Updates 2016, 29:13-29. 262 9. Cerquetti M, Giufrè M, García-Fernández A, Accogli M, Fortini D, Luzzi I, Carattoli 263 A: Ciprofloxacin-resistant, CTX-M-15-producing Escherichia coli ST131 clone 264 in extraintestinal infections in Italy. Clinical Microbiology and Infection, 265 16(10):1555-1558. 266 10. Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, Datsenko KA, Tomita 267 M, Wanner BL, Mori H: Construction of Escherichia coli K‐12 in‐frame, single‐ 268 gene knockout mutants: the Keio collection. Molecular systems biology 2006, 2(1). 269 11. McClure R, Balasubramanian D, Sun Y, Bobrovskyy M, Sumby P, Genco CA, 270 Vanderpool CK, Tjaden B: Computational analysis of bacterial RNA-Seq data. 271 Nucleic Acids Res 2013, 41(14):e140. 272 12. Avasthi TS, Kumar N, Baddam R, Hussain A, Nandanwar N, Jadhav S, Ahmed N: 273 Genome of Multidrug-Resistant Uropathogenic Escherichia coli Strain NA114 274 from India. Journal of bacteriology 2011, 193(16):4272-4273.

54 275 13. Cirz RT, Chin JK, Andes DR, de Crécy-Lagard V, Craig WA, Romesberg FE: 276 Inhibition of Mutation and Combating the Evolution of Antibiotic Resistance. 277 PLoS Biol 2005, 3(6):e176. 278 14. Tamae C, Liu A, Kim K, Sitz D, Hong J, Becket E, Bui A, Solaimani P, Tran KP, 279 Yang H et al: Determination of Antibiotic Hypersensitivity among 4,000 Single- 280 Gene-Knockout Mutants of Escherichia coli. Journal of bacteriology 2008, 281 190(17):5981-5988. 282 15. Yamada J, Yamasaki S, Hirakawa H, Hayashi-Nishino M, Yamaguchi A, Nishino K: 283 Impact of the RNA chaperone Hfq on multidrug resistance in Escherichia coli. 284 The Journal of antimicrobial chemotherapy 2010, 65(5):853-858. 285 16. Liu A, Tran L, Becket E, Lee K, Chinn L, Park E, Tran K, Miller JH: Antibiotic 286 sensitivity profiles determined with an Escherichia coli gene knockout 287 collection: generating an antibiotic bar code. Antimicrobial agents and 288 chemotherapy 2010, 54(4):1393-1403. 289 17. Marcusson LL, Frimodt-Møller N, Hughes D: Interplay in the Selection of 290 Fluoroquinolone Resistance and Bacterial Fitness. PLoS Pathog 2009, 291 5(8):e1000541. 292 18. Robicsek A, Strahilevitz J, Jacoby GA, Macielag M, Abbanat D, Park CH, Bush K, 293 Hooper DC: Fluoroquinolone-modifying enzyme: a new adaptation of a common 294 aminoglycoside acetyltransferase. Nature medicine 2006, 12(1):83-88. 295 19. Alam MK, Alhhazmi A, DeCoteau JF, Luo Y, Geyer CR: RecA Inhibitors 296 Potentiate Antibiotic Activity and Block Evolution of Antibiotic Resistance. Cell 297 chemical biology 2016, 23(3):381-391. 298 20. Brisse S, Diancourt L, Laouénan C, Vigan M, Caro V, Arlet G, Drieux L, Leflon- 299 Guibout V, Mentré F, Jarlier V et al: Phylogenetic Distribution of CTX-M- and 300 Non-Extended-Spectrum-β-Lactamase-Producing Escherichia coli Isolates: 301 Group B2 Isolates, Except Clone ST131, Rarely Produce CTX-M Enzymes. 302 Journal of Clinical Microbiology 2012, 50(9):2974-2981. 303 21. López-Cerero L, Navarro MD, Bellido M, Martín-Peña A, Viñas L, Cisneros JM, 304 Gómez-Langley SL, Sánchez-Monteseirín H, Morales I, Pascual A et al: Escherichia 305 coli belonging to the worldwide emerging epidemic clonal group O25b/ST131: 306 risk factors and clinical implications. Journal of Antimicrobial Chemotherapy 307 2014, 69(3):809-814. 308 22. Tran T, Ran Q, Ostrer L, Khodursky A: De Novo Characterization of Genes That 309 Contribute to High-Level Ciprofloxacin Resistance in Escherichia coli. 310 Antimicrobial agents and chemotherapy 2016, 60(10):6353-6355. 311 23. Drlica K, Malik M, Kerns RJ, Zhao X: Quinolone-mediated bacterial death. 312 Antimicrobial agents and chemotherapy 2008, 52(2):385-392. 313 24. Michel B: After 30 Years of Study, the Bacterial SOS Response Still Surprises 314 Us. PLoS Biology 2005, 3(7):e255. 315 25. Kuzminov A: RuvA, RuvB and RuvC proteins: cleaning-up after 316 recombinational repairs in E. coli. BioEssays : news and reviews in molecular, 317 cellular and developmental biology 1993, 15(5):355-358.

55 318 26. Chase JW, Richardson CC: Exonuclease VII of Escherichia coli : MECHANISM 319 OF ACTION. Journal of Biological Chemistry 1974, 249(14):4553-4561. 320 27. Schneider R, Travers A, Kutateladze T, Muskhelishvili G: A DNA architectural 321 protein couples cellular physiology and DNA topology in Escherichia coli. 322 Molecular microbiology 1999, 34(5):953-964. 323 28. Mazzariol A, Tokue Y, Kanegawa TM, Cornaglia G, Nikaido H: High-Level 324 Fluoroquinolone-Resistant Clinical Isolates of Escherichia coli Overproduce 325 Multidrug Efflux Protein AcrA. Antimicrobial agents and chemotherapy 2000, 326 44(12):3441-3443. 327 29. Yilmaz S, Altinkanat-Gelmez G, Bolelli K, Guneser-Merdan D, Ufuk Over-Hasdemir 328 M, Aki-Yalcin E, Yalcin I: Binding site feature description of 2-substituted 329 benzothiazoles as potential AcrAB-TolC efflux pump inhibitors in E. coli. SAR 330 and QSAR in Environmental Research 2015, 26(10):853-871. 331 30. Opperman TJ, Kwasny SM, Kim HS, Nguyen ST, Houseweart C, D'Souza S, Walker 332 GC, Peet NP, Nikaido H, Bowlin TL: Characterization of a novel pyranopyridine 333 inhibitor of the AcrAB efflux pump of Escherichia coli. Antimicrobial agents and 334 chemotherapy 2014, 58(2):722-733. 335 31. Aparna V, Dineshkumar K, Mohanalakshmi N, Velmurugan D, Hopper W: 336 Identification of Natural Compound Inhibitors for Multidrug Efflux Pumps of 337 Escherichia coli and Using In Silico High-Throughput 338 Virtual Screening and In Vitro Validation. PloS one 2014, 9(7):e101840. 339 32. Bohnert JA, Schuster S, Kern WV: Pimozide Inhibits the AcrAB-TolC Efflux 340 Pump in Escherichia coli. The open microbiology journal 2013, 7:83-86. 341 33. Chevalier J, Bredin J, Mahamoud A, Malléa M, Barbe J, Pagès J-M: Inhibitors of 342 Antibiotic Efflux in Resistant aerogenes and Klebsiella 343 pneumoniae Strains. Antimicrobial agents and chemotherapy 2004, 48(3):1043- 344 1046. 345 34. Culyba MJ, Mo CY, Kohli RM: Targets for Combating the Evolution of Acquired 346 Antibiotic Resistance. Biochemistry 2015, 54(23):3573-3582. 347 35. Blázquez J, Couce A, Rodríguez-Beltrán J, Rodríguez-Rojas A: Antimicrobials as 348 promoters of genetic variation. Current Opinion in Microbiology 2012, 15(5):561- 349 569. 350 36. Recacha E, Machuca J, Díaz de Alba P, Ramos-Güelfo M, Docobo-Pérez F, 351 Rodriguez-Beltrán J, Blázquez J, Pascual A, Rodríguez-Martínez JM: Quinolone 352 Resistance Reversion by Targeting the SOS Response. mBio 2017, 8(5).

353

354

56

Paper II: A strategy for finding DNA replication inhibitors in E. coli identifies iron chelators as molecules that promote survival of hyper-replicating cells.

Currently in review at: Molecular Microbiology.

57 1 A strategy for finding DNA replication inhibitors in E. coli identifies iron chelators as molecules

2 that promote survival of hyper-replicating cells

3

4 Godefroid Charbon1†, Rasmus Nielsen Klitgaard1†, Charlotte Dahlmann Liboriussen1, Peter Waaben

5 Thulstrup2, Sonia Ilaria Maffioli3, Stefano Donadio3 and Anders Løbner-Olesen1*

6

7 From the 1University of Copenhagen, Dept. of Biology, Ole Maaløes Vej 5, 2200 Copenhagen N,

8 Denmark. 2University of Copenhagen, Dept. of Chemistry, Universitetsparken 5, 2100 Copenhagen

9 Ø, Denmark. 3NAICONS Srl, Viale Ortles 22/4, 20139 Milano, Italy.

10

11 Running title: Screens for DNA replication inhibitors

12

13 † Equally contributing authors.

14

15 *To whom correspondence should be addressed: Anders Løbner-Olesen: University of

16 Copenhagen, Dept. of Biology, Ole Maaløes Vej 5, 2200 Copenhagen N, Denmark. Phone: +45

17 3532 2068 .

18

19 Keywords: Escherichia coli, anti-replication initiation compounds, deferoxamine, iron ,

20 oxidative stress, DNA damage, DNA replication, drug screening.

58 21

22

23 Summary

24 DNA replication is often considered an attractive target for new antibacterial compounds. Here we

25 present a strategy to select molecules that inhibit initiation of chromosome replication. We made

26 use of two Escherichia coli strains that display hyper-initiation of replication by keeping the DnaA

27 initiator protein in its active ATP bound state. While viable under anaerobic growth or when grown

28 on poor media, theses strains become inviable when grown in rich media. Our strategy relies on the

29 ability of putative anti-replication initiation molecules to restore their survival. Extracts from

30 actinomyces strains were screened, leading to the identification of deferoxamine (DFO) as the

31 active compound in one of them. However, rather than inhibit replication initiation, we suggest that

32 DFO chelates cellular iron to limit the formation of reactive oxygen species and promote

33 processivity of DNA replication. We also argue that the benzazepine derivate (±)-6-Chloro-PB

34 hydrobromide acts in a similar manner.

35 Introduction

36 Duplication of the genetic material is essential for bacterial proliferation. Targeting DNA

37 replication for inhibition by new antimicrobials is attractive because the many factors contributing

38 to this process are conserved between prokaryotes, but differ significantly from their eukaryotic

39 counterparts (Robinson et al., 2012). Yet, only DNA topoisomerase inhibitors such as quinolones

40 are currently used in the clinic. Other molecules have been found to directly target components of

41 the DNA replication machinery as reviewed in (Robinson et al., 2012) but status for clinical use is

42 uncertain at this stage.

59 43 In Escherichia coli, like most bacteria, the commencement of DNA replication is controlled by

44 DnaA. DnaA is a conserved protein that binds to the chromosomal origin of replication, oriC,

45 promotes strand opening and loads the replication machinery (for recent reviews see (Leonard &

46 Grimwade, 2015, Riber et al., 2016, Skarstad & Katayama, 2013)). In E. coli, DnaA activity is

47 controlled by multiple regulatory pathways to ensure that it starts DNA replication only once per

48 cell cycle and at a defined cellular mass (Donachie, 1968, Cooper & Helmstetter, 1968). Deviations

49 from this once-and-only-once rule has fatal consequences for cell survival (Kellenberger-Gujer et

50 al., 1978, Hirota et al., 1970) . An increased frequency of initiations, such as provoked by hyper-

51 activation of DnaA, leads to accumulation of strand breaks and cell death in a manner somewhat

52 resembling the mode of action of quinolones (Simmons et al., 2004, Charbon et al., 2014).

53 Inactivating DnaA on the other hand leads to an arrest in cell proliferation due to the absence of

54 duplication of the genetic material (Hirota et al., 1970). Slight deviations in the timing of initiation

55 that are seemingly inconsequent for bacterial growth in a laboratory setting affect competitiveness

56 in the host digestive tract (Frimodt-Moller et al., 2015). Thus compounds that affect DnaA function

57 and/or the replication initiation frequency holds promise for therapeutic use.

58 DnaA is composed of four domains performing distinct functions in the initiation process (Messer

59 et al., 1999), and domain I, III and IV functions could serve as putative targets for inhibition.

60 Domain I interacts with the DNA helicase to commence the assembly of the DNA replication

61 machine at the origin of replication and is involved in oligomerization of the protein. Domain II is a

62 flexible linker region that shows little conservation between DnaA proteins from different bacterial

63 species (Messer, 2002) . Domain III is an AAA+ ATPase domain which is often found in initiator

64 proteins. Domain III has a crucial function in promoting formation of a DNA bound DnaA polymer

65 necessary to induce DNA duplex opening and to interact with single stranded DNA (Erzberger et

66 al., 2006). Finally, binding of DnaA to oriC is ensured by a helix-turn-helix motif in Domain IV.

60 67 The regulation of DnaA is quite complex, but in essence, DnaA bound to ATP is the active form

68 that accumulates prior to initiation when a DnaAATP-oriC nucleoprotein complex is formed at the

69 origin. This complex triggers strand opening, helicase loading and assembly of the DNA replication

70 machinery to commence DNA replication. This multimeric DnaAATP assembly on oriC is regulated

71 by binding and hydrolysis of ATP in the AAA+ domain and is the key regulatory feature that

72 ensures proper timing of initiation (Sekimizu et al., 1987, Kurokawa et al., 1999). Following

73 initiation, and to prevent a new cycle of initiation, DnaAATP is inactivated, i.e. converted to

74 DnaAADP. This inactivation is triggered by regulatory inactivation of DnaA (RIDA) (Kato &

75 Katayama, 2001) and datA-dependent DnaA-ATP hydrolysis (DDAH) (Kasho & Katayama, 2013)

76 process. RIDA is performed by the Hda protein in complex with the -clamp loaded on the

77 chromosome. In this complex, Hda directly stimulates the ATPase activity of the DnaAATP

78 complex; DnaA now bound to ADP is inactive. Inactivation of DnaA by DDAH is achieved by the

79 formation of a DnaAATP nucleo-protein complex on the non-coding DNA element datA, which

80 stimulate ATP hydrolysis. Several factors stimulate the DnaA dependent initiation process without

81 being essential. These include DiaA, H-NS, IHF etc. (For review see (Riber et al., 2016, Skarstad &

82 Katayama, 2013)).

83 Prior to a new initiation event the pool of active DnaA molecules is increased by de novo synthesis

84 of DnaA and by rejuvenation of DnaAADP into DnaAATP. This rejuvenation is controlled by the

85 binding of DnaAADP to two DNA elements called DARS1 and DARS2 (Fujimitsu et al., 2009).

86 DnaAADP binding to DARS promotes the release of ADP which permits DnaA to rebind ATP and

87 be active for initiation.

88 Cells deficient in Hda and cells carrying a multi-copy DARS2 plasmid, have an increased

89 DnaAATP/DnaAADP ratio (Kato & Katayama, 2001, Fujimitsu et al., 2009) . This results in hyper-

90 initiation of replication, also called over-initiation, and in most conditions loss of viability or

61 91 selection of compensatory mutations (Kato & Katayama, 2001, Fujimitsu et al., 2009, Charbon et

92 al., 2011, Riber et al., 2006). The high number of replication forks present in these cells makes

93 them hypersensitive to DNA damages such as those provoked by reactive oxygen species (ROS)

94 (Charbon et al., 2014, Simmons et al., 2004) (for review (Charbon et al., 2017b)). However, hyper-

95 initiating cells are viable under growth conditions that reduce conflicts between the elevated

96 number of replication forks and DNA repair processes. These conditions include anaerobic growth

97 to lower oxidative damage to the DNA or slow growth to increase spacing between replication

98 forks. (Charbon et al., 2014, Charbon et al., 2017a).

99 Here we present a screen for inhibitors that target the initial step in the duplication of the bacterial

100 chromosome, i.e. replication initiation at oriC. The screen is based on shifting hyper-initiating cells

101 from permissive conditions to non-permissive conditions, the latter being aerobic growth on rich

102 medium. In principle, a compound that reduces initiations to a level that sustains growth can be

103 selected as it will provide viability to the cells. Such anti-replication initiation compounds are

104 expected to lower DNA replication and thereby viability in wild-type cells (Fig. 1A). 400 extracts

105 of filamentous actinomycetes were screened for containing putative replication initiation inhibitors,

106 a strategy that led to the discovery of -clamp targeting griselimycins antibiotics (Kling et al., 2015,

107 Terlain & Thomas, 1971). We identified deferoxamine (DFO) as being able to restore growth of

108 over-initiating cells. A detailed characterization of its mode of action however points to titration of

109 the cellular iron pool to reduce the Fenton reaction and thereby also ROS inflicted DNA damage.

110 Rather than being a replication inhibitor, DFO thus works by promoting replication elongation in

111 over-initiating cells. The benzazepine derivate (±)-6-Chloro-PB hydrobromide (S143) that was

112 previously identified in a similar screen (Johnsen et al., 2010, Fossum et al., 2008) and proposed to

113 target the DNA gyrase was found to act in a manner similar to DFO.

114 Results

62 115 Screening microbial extracts for inhibitors of the initiation of DNA replication

116 Cells deficient in Hda and cells carrying a pBR322 type plasmid with DARS2 have an increased

117 DnaAATP/DnaAADP ratio and hence over-initiate chromosomal replication, albeit to different extent,

118 with the degree of over-initiation being strongest in the presence of pBR322-DARS2 (Charbon et

119 al., 2014). Both cell types are viable during anaerobic growth or growth on a poor carbon source

120 such as glycerol (referred to as minimal poor medium) i.e. permissive conditions, while inviability

121 is observed during aerobic growth on rich medium, i.e. non-permissive conditions (Charbon et al.,

122 2014). When over-initiating cells were plated on minimal medium plates supplemented with

123 glucose and casamino acids (referred to as minimal rich medium) and incubated aerobically no

124 growth was observed. In order to screen for inhibitors of DNA replication initiation, cells plated on

125 minimal rich medium agar plates were exposed to bioactive natural products supplied in small holes

126 punctured in the agar plate. Following overnight incubation at 37 C° the presence or absence of

127 cellular growth can be determined by visual inspection (Fig. 1B).

128 To search for compounds targeting chromosomal replication initiation, 400 microbial extracts

129 derived from a collection of filamentous actinomycetes were screened using the pBR322-DARS2

130 setup. Seven extracts rescued the growth of the pBR322-DARS2 strain on minimal rich medium.

131 These seven hits were then tested in the hda-screen, giving six strong hits and one weaker; judged

132 from the diameter of the growth zone at non-permissive conditions (Fig. 2A). Extract 18C9 derived

133 from a Streptomyces sp. ID. 62762 gave a strong response in both the pBR322-DARS2- and the

134 hda-screen, and was therefore chosen for further characterization and fractionated into 24 fractions

135 by high performance liquid chromatography (HPLC).

136

137 Identifying the active compound of extract 18C9

63 138 To identify the active compound in extract 18C9, the 24 HPLC fractions were screened using the

139 hda-screen. Only fraction five and six rescued the growth of hda mutant cells, indicating that these

140 contained the active compound (Fig. 2B). These two fractions were then analyzed by HPLC and

141 mass spectrometry (MS). Figure 2C depicts the HPLC chromatogram and MS results for fraction

142 five. In the HPLC chromatogram, there is a distinctive peak between five and six minutes that was

143 only abundant in in these two fractions. MS analysis of the HPLC peak revealed a peak at 585 m/z

144 585 [M-2H+Al], with a clear MS fragmentation pattern. Submission of the MS data in the GNPS

145 database identified the compound as deferoxamine (DFO), a known iron-chelator.

146

147 Deferoxamine rescues the growth of the pBR322-DARS2 strain and the hda mutant

148 Iron plays a key role for many important processes in microorganisms, including reduction of

149 oxygen for ATP synthesis and amino acid synthesis (Roosenberg et al., 2000). Although iron is one

150 of the most abundant elements, the most common oxidation state iron (III) is very insoluble under

151 physiological conditions. Therefore, many microorganisms secrete iron-chelators, also known as

152 siderophores, to scavenge and solubilize iron from their environment to be transported across the

153 cell membrane (Hider & Kong, 2010). DFO, the presumed active compound in extract 18C9, is a

154 siderophore that is produced and secreted by different Streptomyces species (Barona-Gomez et al.,

155 2004). To assess whether DFO is indeed the active compound that can rescue growth of over-

156 initiating cells, five different DFO concentrations were tested in both the pBR322-DARS2 and hda

157 screen. All five DFO concentrations resulted in growth rescue at non-permissive conditions for both

158 types of over-initiating cell types (Fig. 2D). Note that a higher level of DFO was needed to rescue

159 cells carrying a pBR322- DARS2 plasmid in agreement with these cells having the strongest over-

160 initiation phenotype. We estimated the minimal hda rescuing concentration of DFO to be at ~8µg

161 ml-1 (Fig. S1).

64 162

163 Deferoxamine does not prevent bacterial growth

164 In cells grown under permissive conditions, i.e. on minimal poor medium, a clearing zone was

165 observed around the point of DFO addition (Fig. 2D), suggesting that DFO can interfere with E.

166 coli growth. To evaluate the antimicrobial activity of DFO against wild-type E. coli, we attempted

167 to determine the minimal inhibitory concentration (MIC) for DFO with 512 µg ml-1 as the highest

168 concentration. Consistent with previous reports (Thompson et al., 2012) we did not observe

169 complete growth inhibition even at concentrations as high as 512 µg ml-1. However, we observed a

170 ~20 pct reduction in doubling time of wild-type cells at DFO concentrations ranging from 100 µg

171 ml-1 to 10 µg ml-1 (Fig. S2). This Indicates that the clearing zone observed around the point of DFO

172 addition most likely reflects growth retardation due to iron depletion.

173

174 Deferoxamine does not inhibit initiation of chromosome replication

175 When wild-type cells were grown in minimal poor medium and treated with rifampicin and

176 cephalexin prior to flow cytometric analysis, they were found to contain mainly one, two or four

177 fully replicated chromosomes indicating the same number of origins prior to drug addition (Fig. 3).

178 When shifted to minimal rich medium, the doubling time decreased from 90 minutes to 35 minutes

179 and cells contained mainly two and four replication origins in accordance with the increased growth

180 rate (Cooper & Helmstetter, 1968). The addition of 150 M DFO to the minimal rich medium

181 increased the doubling time from 35 minutes to 43 minutes and the number of origins per cell

182 decreased somewhat consistent with the reduced growth rate. The origin concentration did not

183 change in the presence of DFO suggesting that it does not affect initiation of replication in wild-

184 type cells.

65 185 Cells deficient in Hda and cells containing the pBR322-DARS2 plasmid had an increased number

186 of origins per cell when grown in minimal poor medium and over-initiated replication as

187 demonstrated by an increased origin concentration. When these cells were shifted to minimal rich

188 medium for four hours the number of origins per cell increased from an average of 2.9 and 3.0 to >7

189 and >8 for hda mutant and pBR322-DARS2 carrying cells, respectively (Fig. 3). Note that the

190 replication run out following treatment with rifampicin and cephalexin was incomplete and the

191 cellular number of origins is therefore underestimated (Fig. 3). When the same cells were shifted to

192 minimal rich medium in the presence of DFO the situation was different. The number of origins per

193 cell increased somewhat due to the increased growth rate, replication runout was complete and the

194 origin concentration remained the same or was only slightly elevated (Fig. 3).

195 Altogether this suggests that DFO does not reduce initiations from oriC and that this is not the

196 mechanism behind the rescue of over-initiating cells.

197

198 Deferoxamine does not rescue over-initiating cells by reducing their growth rate.

199 We have previously shown that lethal over-initiation in hda mutant cells can be suppressed by slow

200 growth (Charbon et al., 2017a). Because the presence of DFO was found to slow down the growth

201 of wild-type cells we wondered whether this could be the mechanism behind the rescue of hda

202 mutant and pBR322-DARS2 carrying cells.

203 We therefore tested the ability of DFO to rescue the hda mutant in the richer LB medium. Wild-

204 type and hda mutant cells were grown exponentially in presence of 150 M DFO for more than 12

205 generations and had doubling times of 28 and 31 minutes, respectively (Fig. 4A). In minimal

206 medium with DFO hda mutant cells had a doubling time of 60 minutes (Fig. 3).

207 We proceeded to shift cells from DFO containing to DFO free medium. During such a shift the

208 doubling time of the hda mutant increased (Fig. 4A insert) and eventually ceased altogether. The

66 209 number of origins per cell increased from ~15 to more than 30, while the origin concentration could

210 not be determined precisely due to an incomplete run-out. This demonstrates that the presence of

211 DFO ensures viability of hda mutant cells even at doubling times as fast as 31 minutes, where cells

212 over-initiate dramatically. The aggravation of the growth and replication phenotypes after DFO

213 removal also indicates that the DFO rescue was not due to accumulation of suppressor mutations

214 (Kato & Katayama, 2001, Fujimitsu et al., 2009, Charbon et al., 2011, Riber et al., 2006). We

215 therefore conclude that DFO does not rescue hda mutant cells by merely reducing their growth rate.

216

217 Deferoxamine increases processivity of replication forks in over-initiating cells

218 The flow cytometry histograms of over-initiating cells at non-permissive conditions indicated that

219 these failed to complete replication in the presence of rifampicin and cephalexin (Figs. 3 and 4). We

220 therefore determined the origin to terminus ratio (ori/ter) for wild-type, hda mutant cells and cells

221 carrying a pBR322-DARS2 plasmid during growth on minimal poor medium and four hours

222 following a shift to minimal rich medium (Fig. 4B). As expected the ori/ter ratio for wild-type cells

223 only increased from 1.2 to 2.4 when shifted from minimal poor to minimal rich medium, as

224 expected from the increase in growth rate (Fig. 4B). On the other hand the ori/ter ratio for hda

225 mutant cells and cells carrying a pBR322-DARS2 plasmid increased from 1.6 and 1.7 to >25 and

226 >75, respectively, following the same shift (Fig. 4B) suggesting that many replication forks initiated

227 at oriC never reach the terminus in these cells. Again this is in agreement with the strongest over-

228 initiation phenotype elicited by the pBR322-DARS2 plasmid. The presence of 150 M DFO in

229 minimal rich medium reduced the ori/ter ratio of hda mutant cells and cells carrying a pBR322-

230 DARS2 plasmid from >25 and >75 to 2.0 and 5.0 relative to cells without DFO, respectively (Fig.

231 4B). Altogether, this indicates that DFO helps the DNA replication elongation process in over-

232 initiating cells.

67 233

234 Optimization of the pBR322-DARS2 and hda screens by addition of excess iron.

235 The ability of DFO to ensure viability of over-initiating cells by promoting replication elongation

236 was not surprising as it is known that oxidative damage to DNA is a main reason for inviability

237 (Charbon et al., 2011, Charbon et al., 2017a, Babu et al., 2017, Charbon et al., 2014). A major

238 source of ROS species that can cause oxidative damage is the iron dependent Fenton reactions

239 which are inhibited by DFO (Imlay et al., 1988, Liu et al., 2011), most likely by its ability to bind

240 iron.

241 In order to reduce the risk of false positives such as iron chelators and reducing agents that lower

242 ROS formation in our screens, we added excess iron (II) or (III), in the form of Fe(ClO4)2 or FeCl3,

243 when performing the hda based screen (Fig. 5). The rationale behind adding excess iron to the

244 plates, was to ensure that a given iron chelator would not deplete iron in the plates to a level that

245 limit the generation of ROS, and rescue the over-initiating cells in this manner. To test the

246 hypothesis, 5 l of 10 mM of the four iron chelators; DFO, phenanthroline, bipyridyl, EDTA and

247 the reducing agent dithiothreitol (DTT) were tested, with iron (II) or (III) at a final concentration of

248 3 or 200 µM in the plates. DFO, phenanthroline and EDTA all rescued the growth at the standard

249 iron (II) or (III) concentration of 3 µM, while DTT and bipyridyl only did at higher concentration

250 (Fig. 5 A). When iron (II) is at a final concentration of 200 µM, the rescuing effect of EDTA, DFO

251 and phenanthroline was no longer observed (Fig. 5A). As expected the DFO effect was also

252 counteracted by iron supplementation in the pBR322-DARS2 screen (Fig. S3). These results are

253 also consistent with the recovery of growth rate observed when wild-type cells treated with DFO

254 are provided with excess iron in the liquid medium (Fig. S4), i.e. DFO treated cells are depleted for

255 iron. This demonstrates that a high level of iron (II) in the agar plates removes falls positives from

256 iron chelators and reducing agents in the screens.

68 257 To verify that a high level of iron (II) in the screen did not negatively interfere with

258 the detection of DNA replication inhibitors, we assessed the IPTG dependent expression of either

259 the negative initiation regulator SeqA (Lu et al., 1994, Campbell & Kleckner, 1990, von

260 Freiesleben et al., 2000, Charbon et al., 2011) or a cyclic DnaA domain I derived peptide inhibiting

261 DnaA activity (Kjelstrup et al., 2013) in the hda based screen (Fig. 5B). Production of either the

262 cyclic peptide or SeqA was able to rescue the hda mutant cells in presence of 200 µM iron (II). A

263 high level of iron therefore did not have a negative effect on the screen. (Fig. 5 A,B).

264 Finally, we determined whether the remaining six positive hits from the initial screen (Fig. 2A)

265 were false positives by subjecting them to the hda screen with 200 µM iron in the agar plates. This

266 time the six extracts did not rescue the growth of the hda mutant, indicating that they were false

267 positives, most probably preventing ROS formation one way or another.

268

269 (±)-6-Chloro-PB hydrobromide (S143) rescues the growth of the hda mutant

270 Previously, Johnsen et al. reported that the benzazepine derivate (±)-6-Chloro-PB hydrobromide

271 (S143) rescued the growth of over-initiating cells (Johnsen et al., 2010). The rescuing effect of

272 S143 was assigned to a partial inhibition of the DNA gyrase, demonstrated by a supercoiling assay

273 and by countering growth inhibition caused by gyrase overproduction (Johnsen et al., 2010). When

274 tested as a 10 mM solution in our hda based screen, S143 rescued growth of the mutant on plates

275 with iron (II) at a final concentration of 3 µM but not 200 µM (Fig. 6A). S143 also gave rise to a

276 clearing zone when tested on wild-type cells (Fig. 6B) suggesting that the compound interfere with

277 bacterial growth. The growth inhibition could be overcome by addition of Iron (II) at final

278 concentration of 200 µM (Fig. 6 B). Taken together these results indicate that S143 affects iron

279 homeostasis. Note that in presence of excess iron in the plate, S143 changes color (Fig. 6 B).

280

69 281 S143 chelates iron

282 The structure of S143 indicates that it may have a catechol type iron chelation activity (Fig. 7).

283 Catechol groups are found in many siderophores such as E.coli’s enterobactin that contains three

284 catechol groups and has an extremely high affinity for chelating iron(III)(Raymond et al., 2003).

285 We first tested the ability of S143 to outcompete the chelation of iron II by phenanthroline using

286 DFO as a control. Phenanthroline complexes with iron (II) (3:1) and absorbs light at 510 nm. We

287 measured absorbance at 510 nm when a limiting amount of iron (II) was mixed with increasing

288 amount of S143 or DFO prior to addition of a fixed amount of phenanthroline (Fig. 6 C). It was

289 clear that both DFO and S143 outcompete phenanthroline, with DFO being more efficient,

290 indicating that both compounds here are able to bind iron (Fig. 6 C and Fig. S5), although both

291 preferably bind iron(III) over iron(II). Because our assay is performed aerobically in unbuffered

292 ddH2O, the assay likely shows in all or in part, binding of S143 to iron (III) due to iron (II)

293 oxidation. When S143 was mixed with iron (II) perchlorate or iron (III) nitrate, the mixture became

294 green. We therefore measured the absorption spectrum of S143 mixed with iron (III) nitrate. The

295 absorption spectrum indicates that iron (III) and S143 forms complexes absorbing at ~450 nm and

296 ~700 nm (Fig. 6 D). Altogether, these data indicate that S143 binds iron as expected for a catechol-

297 containing ligand, however at tested conditions the mono-complex is formed rather than the bis- or

298 tris-complex (Sever & Wilker, 2004).

299

300 Discussion

301 We designed screens to identify inhibitors of initiation of chromosome replication. We made use of

302 the fact that hda mutants or cells carrying a pBR322-DARS2 plasmid accumulate DnaAATP, hyper-

303 initiate replication, accumulate strand breaks and eventually die. Consequently, compounds that

304 reduce the initiation frequency are expected to restore viability. These screens leave the DnaA

70 305 protein intact as opposed to a related screen employing DnaA mutated in the AAA+ domain

306 (Johnsen et al., 2010, Fossum et al., 2008). Our approach uses a dual sensitivity assay, with

307 pBR322-DARS2 cells being the most selective (Fig. 2). Testing hda and DARS2 assays also has the

308 advantage of discarding certain type of compounds that would be false positives in the pBR322-

309 DARS2 screen. These include a molecule that inhibits plasmid replication of the pBR322, as this is

310 expected to restore growth of pBR322-DARS2 transformed cells but not hda mutant cells (not seen

311 with the collection of extracts tested insofar).

312 Deferoxamine was identified from an extract of Streptomyces sp. ID. 62762 as a molecule that

313 restores viability of both types of hyper-initiating cells. Deferoxamine is a siderophore produced by

314 actinomycetes that has been in use as therapeutic agent for iron or aluminum poisoning (Barata et

315 al., 1996) . Because of its iron chelation properties, it has also been tested as a bacteriostatic agent,

316 albeit with poor outcome (Thompson et al., 2012). Successful anti-microbial use of siderophores

317 was previously reported (Saha et al., 2016) but only for species unable to use the siderophore in

318 question. For bacteria capable of using DFO as a siderophore, the situation is reversed (D'Onofrio et

319 al., 2010) and DFO enhance the growth of Klebsiella pneumoniae and increase the susceptibility of

320 mice to infections caused by Yersenia enterocolitica (Chan et al., 2009, Robins-Browne & Prpic,

321 1985).

322 We found that although DFO reduced the growth rate of E. coli, the minimal inhibitory

323 concentration was in above 512 µg ml-1 indicating that it failed to display bacteriostatic or

324 bactericidal effects below this concentration in agreement with previous data (Thompson et al.,

325 2012). We found no indication that DFO affects DNA replication in wild-type cells since its

326 presence affect neither origin concentration nor initiation synchrony.

327 The reason for identifying DFO in our screens may solely come from its ability to chelate iron and

328 thus inhibit the Fenton reaction such as described previously in vitro (Imlay et al., 1988) and in

71 329 vivo (Liu et al., 2011). This results in reduced generation of ROS and hence a reduced level of

330 oxidative damage in cells treated with DFO. While there is a narrow time window to repair

331 oxidative damage prior to passage of the next replication fork in wild-type cells (Takahashi et al.,

332 2017, Charbon et al., 2014, Foti et al., 2012), forks are more frequent and closely spaced in hyper-

333 initiating cells where they occasionally encounter a single stranded region resulting from repair of

334 oxidized bases. This results in double stranded DNA breaks, the ultimate reason for cell death

335 (Charbon et al., 2014). Overall, we therefore suggest that DFO acts by binding iron to reduce ROS

336 generated by the Fenton reactions. This results in a reduced level of oxidative DNA damage, which

337 in turn permits closely spaced replication forks to proceed unimpeded in hyper-initiating cells. This

338 explains why hda mutant cells and cells carrying a pBR322-DARS2 plasmid have a close to wild-

339 type ori/ter ratio when treated with DFO despite of continued over-initiation. This is also in

340 agreement with data showing that hyper-initiating cells that generate less or no ROS, due to

341 anaerobic growth or due to having their energy metabolism shifted towards fermentation are viable,

342 as these cells has less or no ROS inflicted DNA damage that need repair (Charbon et al., 2017a,

343 Charbon et al., 2014). The iron chelator bipyridyl and other reducing agents were found to have the

344 same effects as DFO 32.

345 Finally, we tested the benzazepine derivate (±)-6-Chloro-PB hydrobromide (S143) that rescued the

346 growth of cells carrying the conditional hyperactive DnaA219 protein at non-permissive conditions

347 (Johnsen et al., 2010, Fossum et al., 2008) and found it capable of rescuing the growth of hda cells.

348 S143 was initially described as a selective agonist of the dopamine D1-like receptor (Weed et al.,

349 1993) but has also been proposed to be a partial inhibitor of E. coli DNA gyrase (Johnsen et al.,

350 2010, Fossum et al., 2008). However, this seems unlikely as the ability of S143 to rescue the

351 growth of hda cells was counteracted by addition of excess iron. We predict that S143 is able to

352 bind iron up to a 3:1 stoichiometry (Fig. 7) through its catechol group and demonstrated that it

72 353 forms complexes with iron (II or III). We therefore suggest that the S143 mode of action is, like for

354 DFO, explained by its iron chelating properties. This is also in agreement with S143 being selected

355 as a molecule capable to promote survival of myocardial cells exposed to toxic level of H2O2 (Gero

356 et al., 2007). Here it was concluded that S143 is an indirect inhibitor of cellular PARP activity.

357 Viewing our results, another likely explanation can be found in the chelation of iron and thereby

358 reducing the Fenton reactions. We suggest that S143 chelates iron (and likely other metals) and that

359 this activity is responsible in all or in part for the effects previously observed with this drug.

360 It seems clear that a drawback of our screens is that they will identify molecules that limit reactive

361 oxygen species mediated DNA damage. DFO and other siderophores are often co-produced with

362 other metabolites by actinomycetes. As DFO clearly did not represent the type of molecules we

363 (and other) originally pursued, we adapted our screens to avoid “Fenton reaction moderators” by

364 adding iron in excess in the growth medium. This still allowed identification of replication initiation

365 inhibitors because overproduction of SeqA or a DnaA Domain I derived peptide came out positive

366 in the modified screens. With these modified screens we retested all of our original hit extracts and

367 found none of them to be positive, suggesting that naturally occurring replication initiation

368 inhibitors isolated from actinomycetes strains are found at a much lower frequency in natural

369 extracts than iron chelators such as DFO.

370

371 Experimental procedures

372 Medium

373 Cells were grown in Lysogeny Broth (LB) medium or AB minimal medium (Clark & Maaløe,

374 1967) supplemented with 10 µg ml-1 thiamine and either 0.2% glycerol (minimal poor medium) or

375 0.2% glucose and 0.5% casamino acids (minimal rich medium).

376

73 377 Bacterial strains and plasmids

378 All strains used are derivatives of the E. coli strain MG1655 (F- λ- rph-1) (Guyer et al., 1981). The

379 deletion of hda was performed by P1 mediated transduction (Miller, 1972) as described previously

380 (Riber et al., 2006) and plated on minimal poor medium. The pBR322-DARS2 plasmid is described

381 in (Charbon et al., 2014) (Bolivar et al., 1977). pRNK4 is derived from pSC116 (Kjelstrup et al.,

382 2013) by digestion with PvuI followed by re-ligation, thereby removing the chloramphenicol

383 resistance gene and reconstituting the ampicillin resistance gene. Plasmid pMAK7 was described

384 previously (von Freiesleben et al., 2000).

385

386 Chemicals and reagents

387 Deferoxamine mesylate salt (CAS:138-14-7), (±)-6-chloro-PB hydrobromide (S143, CAS:71636-

388 61-8), 2,2´-Bipyridyl (CAS:366-18-7), 1,10-Phenanthroline (CAS:66-71-7), Iron (III) chloride

389 hexahydrate (CAS:100025-77-1), Iron (III) nitrate nonahydrate (CAS:7782-61-8) and Iron (II)

390 perchlorate hydrate (CAS:335159-18-7) were all purchased from Sigma-Aldrich. While EDTA

391 disodium salt (CAS:6381-92-6) and DL-Dithiothreitol (CAS:3483-12-3) was purchased from

392 Chemsolute and VWR Life science, respectively.

393

394 hda screen

395 MG1655 hda::cat was grown overnight in minimal poor medium containing 20 µg ml-1

5 396 chloramphenicol. The overnight culture was diluted to OD600 = 0.0004 (approximately 2x10 cfu

397 ml-1) in 0.9% NaCl. 100 µl of the diluted culture was then plated on minimal poor medium and

398 minimal rich medium plates. Holes were punched in the agar plates using a glass-pipette, and the

399 extracts or compounds to be tested were dispensed into these holes. Following overnight incubation

74 400 at 37oC, the minimal rich medium plates were inspected for growth rescue and the minimal poor

401 medium plates for inhibition zones.

402

403 Multi-copy DARS2 Screen

404 MG1655/pBR322DARS2 and MG1655/pBR322 were grown overnight in minimal poor medium

-1 405 containing 150 µg ml ampicillin. The overnight cultures were then diluted to OD600 = 0.0004

406 (approximately 2x105 cfu ml-1) in 0.9% NaCl. 100 µl of the diluted culture was then spread on

407 minimal rich medium and minimal poor medium plates, containing 150 µg ml-1 ampicillin. A glass-

408 pipette was used to punch holes in the agar, and the extracts or compounds to be tested were

409 dispensed into these holes. Following overnight incubation at 37oC, the minimal rich medium plates

410 were inspected for growth rescue and the minimal poor medium plates for inhibition zones. For

411 screening the 400 microbial extracts, 5µl of extract was dispensed in the holes of the plates, and 5µl

412 10% DMSO was used as a negative control.

413

414 Preparation of microbial extracts

415 The 400 microbial extracts were prepared by Naicons srl., Milan, Italy. 10 ml cultures of

416 filamentous actinomycetes were centrifuged at 3000 rpm for 10 minutes to separate the cells from

417 the supernatant. 4 ml ethanol was added to the pellet and incubated for 1h at room temperature with

418 shaking. 0.2 ml aliquots of the ethanolic extracts were distributed in 96-well microtiter plates, dried

419 under vacuum, and stored at 4°C. HP20 resin (Mitsubishi Chemical Co., 1 ml) was added to the

420 supernatant and incubated for 2 hours at room temperature with shaking. The resin was washed with

421 6 ml H2O, and eluted with 5 ml 80% MeOH. 0.25 ml aliquots were distributed in 96-well microtiter

422 plates, dried under vacuum, and stored at 4°C.

423

75 424 HPLC fractionation of extract 18C9

425 Extract, from plate-well 18-C9, was dissolved in 100 l of 80% MeOH. 90 l were fractionated by

426 HPLC on a Shimadzu LC 2010A-HT with the following settings, Column: Merck LiChrosphere

427 RP-18, LiChrocart 5 μm 4.6 x 125mm, phase A: 0.01M HCOONH4 (ammonium formate), phase B:

428 MeCN, flow: 1 ml min-1 at 50°C, UV detection: 230 nm. Linear gradient of phase B: 10 to 95% in

429 18 minutes followed by 5 minutes at 95%. 24 fractions (1 ml each) were collected. 100 l of each

430 fraction were stored for LC/MS analysis while the remaining was dried in a speedvac at 40°C

431 overnight and re-dissolved in 100 µl 10% dmso for the screening.

432

433 Identification of Deferoxamine from extract 18C9 by LC-MS.

434 LC-MS analyses was carried out using a Dionex UltiMate 3000 coupled with an LCQ Fleet mass

435 spectrometer equipped with an electrospray interface (ESI) and a tridimensional ion trap. The

436 following settings were used for liquid chromatography: 1 minute of pre-concentration at 10%, a 7

437 minutes linear gradient from 10 to 95%, followed by an isocratic step at 95% of 2 minutes and 1

438 minute of re-equilibration at 10% of CH3CN with an aqueous phase of 0.05% formic acid. The

439 column was an Atlantis T3 C18 5 μm x 4.6 mm x 50 mm at a flow rate of 0.8 ml min -1. The m/z

440 range (120-2000) and the ESI conditions were as follows: spray voltage of 3500 V, capillary

441 temperature of 275 °C, sheat gas flow rate at 35 and auxiliary gas flow rate at 15. The mass data

442 (.RAW files) from Xcalibur were converted to .mzXML file format, followed by submission to the

443 Global Natural Products Social Molecular Networking (Wang et al., 2016) database for de-

444 replication.

445

446 Marker frequency analysis by qPCR

76 447 Cells centrifuged 5 minute 8000x g the supernatant discarded and the cells resuspended in 100 l of

448 cold 10 mM Tris pH7.5. The cells were then fixed by adding 1 ml of 77% ethanol and stored at 4 °C

449 until use. For the qPCR analysis, 100 l of ethanol fixed cells were centrifuged 7 minutes at 17000

450 x g, the supernatant discarded and the samples centrifuged again for 30 seconds at 17000 x g,

451 followed by removal of the remaining ethanol. The cell pellet was resuspended in 1ml cold water

452 and 2 µl was used as template for qPCR analysis. The Quantitative-PCR was performed using a

453 Takara SYBR Premix Ex Taq II (RR820A) in a BioRAD CFX96. All ori/ter ratios were

454 normalized to the ori/ter ratio of MG1655 treated with rifampicin for 2h. The origin and terminus

455 was quantified using primers 5′-TTCGATCACCCCTGCGTACA-3′ and 5′-

456 CGCAACAGCATGGCGATAAC-3′ for the origin and 5′-TTGAGCTGCGCCTCATCAAG-3′ and

457 5′-TCAACGTGCGAGCGATGAAT-3′ for the terminus.

458

459 Flow cytometry

460 Preparation of samples for determination of number of origin per cell: 1 ml of cell culture was

461 incubated at 37°C for 2 to 4 hours with 300 µg ml-1 rifampicin and 36 µg ml-1 cephalexin. Cells

462 were fixed in 70% ethanol and stored at 4°C, as described for the marker frequency analysis by

463 qPCR.

464 Preparation of samples for determination of cell size: 1 ml of cell culture was placed on ice and

465 fixed as described for the marker frequency analysis by qPCR.

466 DNA Staining; 100-300 µl of fixed cells were centrifugated at 15,000 x g for 15 min. The

467 supernatant was discarded and the pellet resuspended in 130 µl “Staining solution” (90 µg ml-1

468 mithramycin, 20 µg ml-1 ethidium bromide, 10 mM MgCl2, 10 mM Tris pH 7.5). Samples were

469 then kept on ice for a minimum of 10 min. prior to flow cytometric analysis. Flow cytometry was

77 470 performed using an Apogee A10 Bryte instrument. For each sample, 30 000 to 200 000 cells were

471 analyzed.

472

473 Minimal inhibitory concentration

474 The MIC of DFO was determined by micro-dilution in a 96-well plate. MG1655 was grown to an

475 OD600 of 0.5 in minimal rich medium. The culture was diluted to an OD600 of 0.001 in minimal rich

476 medium. 100 µl diluted culture was added to each well of a 96-well plate containing a dilution

477 series of DFO in minimal rich medium, giving a final concentration range of 512 to 0.5 µg ml-1 of

478 DFO. The 96-well plate was incubated at 37 °C for 24 hours and inspected for visible growth

479 inhibition.

480

481 Minimal rescuing concentration

482 MG1655 hda::cat was grown overnight in minimal poor medium at 37°C. The overnight culture

483 was diluted 100x in minimal rich medium and grown for four hours at 37°C. The culture was

484 diluted to an OD600 of 0.001 in minimal rich medium and 100 µl culture was added to each well of a

485 96-well plate containing a dilution series of DFO in minimal rich medium, giving a final

486 concentration range of 512 to 0.5 µg ml-1 of DFO. The growth at 37 °C during continuous shaking

487 was monitored for sixteen hours, using a Biotek Synergy H1 plate reader.

488 Acknowledgments

489 The authors were funded by grants from the Danish National Research Foundation (DNRF120)

490 through the Center for Bacterial Stress Response and Persistence (BASP) and by the University of

491 Copenhagen Centre for Control of Antibiotic Resistance (UC-Care).

78 492 References

493 Babu, V.M.P., M. Itsko, J.C. Baxter, R.M. Schaaper & M.D. Sutton, (2017) Insufficient levels of the nrdAB- 494 encoded ribonucleotide reductase underlie the severe growth defect of the Δhda E. coli strain. 495 Molecular microbiology 104: 377-399. 496 Barata, J.D., P.C. D'Haese, C. Pires, L.V. Lamberts, J. Simoes & M.E. De Broe, (1996) Low- (5 mg/kg) 497 desferrioxamine treatment in acutely aluminium-intoxicated haemodialysis patients using two drug 498 administration schedules. Nephrol Dial Transplant 11: 125-132. 499 Barona-Gomez, F., U. Wong, A.E. Giannakopulos, P.J. Derrick & G.L. Challis, (2004) Identification of a cluster 500 of genes that directs desferrioxamine biosynthesis in Streptomyces coelicolor M145. J Am Chem 501 Soc 126: 16282-16283. 502 Bolivar, F., R.L. Rodriguez, P.J. Greene, M.C. Betlach, H.L. Heyneker, H.W. Boyer, J.H. Crosa & S. Falkow, 503 (1977) Construction and characterization of new cloning vehicles. II. A multipurpose cloning system. 504 Gene 2: 95-113. 505 Campbell, J.L. & N. Kleckner, (1990) E. coli oriC and the dnaA gene promoter are sequestered from dam 506 methyltransferase following the passage of the chromosomal replication fork. Cell 62: 967-979. 507 Chan, G.C., S. Chan, P.L. Ho & S.Y. Ha, (2009) Effects of chelators (deferoxamine, deferiprone and 508 deferasirox) on the growth of Klebsiella pneumoniae and Aeromonas hydrophila isolated from 509 transfusion-dependent thalassemia patients. Hemoglobin 33: 352-360. 510 Charbon, G., L. Bjorn, B. Mendoza-Chamizo, J. Frimodt-Moller & A. Lobner-Olesen, (2014) Oxidative DNA 511 damage is instrumental in hyperreplication stress-induced inviability of Escherichia coli. Nucleic 512 Acids Res 42: 13228-13241. 513 Charbon, G., C. Campion, S.H. Chan, L. Bjorn, A. Weimann, L.C. da Silva, P.R. Jensen & A. Lobner-Olesen, 514 (2017a) Re-wiring of energy metabolism promotes viability during hyperreplication stress in E. coli. 515 PLoS Genet 13: e1006590. 516 Charbon, G., L. Riber, M. Cohen, O. Skovgaard, K. Fujimitsu, T. Katayama & A. Lobner-Olesen, (2011) 517 Suppressors of DnaA(ATP) imposed overinitiation in Escherichia coli. Mol Microbiol 79: 914-928. 518 Charbon, G., L. Riber & A. Lobner-Olesen, (2017b) Countermeasures to survive excessive chromosome 519 replication in Escherichia coli. Curr Genet. 520 Clark, D.J. & O. Maaløe, (1967) DNA replication and the division cycle in Escherichia coli. J.Mol.Biol. 23: 99- 521 112. 522 Cooper, S. & C.E. Helmstetter, (1968) Chromosome replication and the division cycle of Escherichia coli B/r. 523 J Mol Biol 31: 519-540. 524 D'Onofrio, A., J.M. Crawford, E.J. Stewart, K. Witt, E. Gavrish, S. Epstein, J. Clardy & K. Lewis, (2010) 525 Siderophores from neighboring organisms promote the growth of uncultured bacteria. Chem Biol 526 17: 254-264. 527 Donachie, W.D., (1968) Relationship between cell size and time of initiation of DNA replication. Nature 219: 528 1077-1079. 529 Erzberger, J.P., M.L. Mott & J.M. Berger, (2006) Structural basis for ATP-dependent DnaA assembly and 530 replication-origin remodeling. Nat Struct Mol Biol 13: 676-683. 531 Fossum, S., G. De Pascale, C. Weigel, W. Messer, S. Donadio & K. Skarstad, (2008) A robust screen for novel 532 antibiotics: specific knockout of the initiator of bacterial DNA replication. FEMS Microbiol Lett 281: 533 210-214. 534 Foti, J.J., B. Devadoss, J.A. Winkler, J.J. Collins & G.C. Walker, (2012) Oxidation of the guanine nucleotide 535 pool underlies cell death by bactericidal antibiotics. Science 336: 315-319. 536 Frimodt-Moller, J., G. Charbon, K.A. Krogfelt & A. Lobner-Olesen, (2015) Control regions for chromosome 537 replication are conserved with respect to sequence and location among Escherichia coli strains. 538 Front Microbiol 6: 1011.

79 539 Fujimitsu, K., T. Senriuchi & T. Katayama, (2009) Specific genomic sequences of E. coli promote replicational 540 initiation by directly reactivating ADP-DnaA. Genes Dev 23: 1221-1233. 541 Gero, D., K. Modis, N. Nagy, P. Szoleczky, Z.D. Toth, G. Dorman & C. Szabo, (2007) Oxidant-induced 542 cardiomyocyte injury: identification of the cytoprotective effect of a dopamine 1 receptor agonist 543 using a cell-based high-throughput assay. Int J Mol Med 20: 749-761. 544 Guyer, M.S., R.R. Reed, J.A. Steitz & K.B. Low, (1981) Identification of a sex-factor-affinity site in E. coli as 545 gamma delta. Cold Spring Harb.Symp.Quant.Biol. 45: 135-140. 546 Hider, R.C. & X. Kong, (2010) Chemistry and biology of siderophores. Natural product reports 27: 637-657. 547 Hirota, Y., J. Mordoh & F. Jacob, (1970) On the process of cellular division in Escherichia coli. 3. 548 Thermosensitive mutants of Escherichia coli altered in the process of DNA initiation. J Mol Biol 53: 549 369-387. 550 Imlay, J.A., S.M. Chin & S. Linn, (1988) Toxic DNA damage by hydrogen peroxide through the Fenton 551 reaction in vivo and in vitro. Science 240: 640-642. 552 Johnsen, L., C. Weigel, J. von Kries, M. Moller & K. Skarstad, (2010) A novel DNA gyrase inhibitor rescues 553 Escherichia coli dnaAcos mutant cells from lethal hyperinitiation. J Antimicrob Chemother 65: 924- 554 930. 555 Kasho, K. & T. Katayama, (2013) DnaA binding locus datA promotes DnaA-ATP hydrolysis to enable cell 556 cycle-coordinated replication initiation. Proc Natl Acad Sci U S A 110: 936-941. 557 Kato, J. & T. Katayama, (2001) Hda, a novel DnaA-related protein, regulates the replication cycle in 558 Escherichia coli. EMBO J 20: 4253-4262. 559 Kellenberger-Gujer, G., A.J. Podhajska & L. Caro, (1978) A cold sensitive dnaA mutant of E. coli which 560 overinitiates chromosome replication at low temperature. Mol Gen Genet 162: 9-16. 561 Kjelstrup, S., P.M.P. Hansen, L.E. Thomsen, P.R. Hansen & A. Løbner-Olesen, (2013) Cyclic Peptide Inhibitors 562 of the β-Sliding Clamp in Staphylococcus aureus. PloS one 8: e72273. 563 Kling, A., P. Lukat, D.V. Almeida, A. Bauer, E. Fontaine, S. Sordello, N. Zaburannyi, J. Herrmann, S.C. Wenzel, 564 C. Konig, N.C. Ammerman, M.B. Barrio, K. Borchers, F. Bordon-Pallier, M. Bronstrup, G. 565 Courtemanche, M. Gerlitz, M. Geslin, P. Hammann, D.W. Heinz, H. Hoffmann, S. Klieber, M. 566 Kohlmann, M. Kurz, C. Lair, H. Matter, E. Nuermberger, S. Tyagi, L. Fraisse, J.H. Grosset, S. Lagrange 567 & R. Muller, (2015) Antibiotics. Targeting DnaN for tuberculosis therapy using novel griselimycins. 568 Science 348: 1106-1112. 569 Kurokawa, K., S. Nishida, A. Emoto, K. Sekimizu & T. Katayama, (1999) Replication cycle-coordinated change 570 of the adenine nucleotide-bound forms of DnaA protein in Escherichia coli. EMBO J 18: 6642-6652. 571 Leonard, A.C. & J.E. Grimwade, (2015) The orisome: structure and function. Front Microbiol 6: 545. 572 Liu, Y., S.C. Bauer & J.A. Imlay, (2011) The YaaA protein of the Escherichia coli OxyR regulon lessens 573 hydrogen peroxide toxicity by diminishing the amount of intracellular unincorporated iron. J 574 Bacteriol 193: 2186-2196. 575 Lu, M., J.L. Campbell, E. Boye & N. Kleckner, (1994) SeqA: a negative modulator of replication initiation in E. 576 coli. Cell 77: 413-426. 577 Messer, W., (2002) The bacterial replication initiator DnaA. DnaA and oriC, the bacterial mode to initiate 578 DNA replication. FEMS Microbiol Rev 26: 355-374. 579 Messer, W., F. Blaesing, J. Majka, J. Nardmann, S. Schaper, A. Schmidt, H. Seitz, C. Speck, D. Tungler, G. 580 Wegrzyn, C. Weigel, M. Welzeck & J. Zakrzewska-Czerwinska, (1999) Functional domains of DnaA 581 proteins. Biochimie 81: 819-825. 582 Miller, J.H., (1972) Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, 583 NY. 584 Raymond, K.N., E.A. Dertz & S.S. Kim, (2003) Enterobactin: an archetype for microbial iron transport. Proc 585 Natl Acad Sci U S A 100: 3584-3588. 586 Riber, L., J. Frimodt-Moller, G. Charbon & A. Lobner-Olesen, (2016) Multiple DNA Binding Proteins 587 Contribute to Timing of Chromosome Replication in E. coli. Front Mol Biosci 3: 29.

80 588 Riber, L., J.A. Olsson, R.B. Jensen, O. Skovgaard, S. Dasgupta, M.G. Marinus & A. Lobner-Olesen, (2006) Hda- 589 mediated inactivation of the DnaA protein and dnaA gene autoregulation act in concert to ensure 590 homeostatic maintenance of the Escherichia coli chromosome. Genes Dev 20: 2121-2134. 591 Robins-Browne, R.M. & J.K. Prpic, (1985) Effects of iron and desferrioxamine on infections with Yersinia 592 enterocolitica. Infection and Immunity 47: 774-779. 593 Robinson, A., R.J. Causer & N.E. Dixon, (2012) Architecture and conservation of the bacterial DNA 594 replication machinery, an underexploited drug target. Curr Drug Targets 13: 352-372. 595 Roosenberg, J.M., 2nd, Y.M. Lin, Y. Lu & M.J. Miller, (2000) Studies and syntheses of siderophores, microbial 596 iron chelators, and analogs as potential drug delivery agents. Current medicinal chemistry 7: 159- 597 197. 598 Saha, M., S. Sarkar, B. Sarkar, B.K. Sharma, S. Bhattacharjee & P. Tribedi, (2016) Microbial siderophores and 599 their potential applications: a review. Environ Sci Pollut Res Int 23: 3984-3999. 600 Sekimizu, K., D. Bramhill & A. Kornberg, (1987) ATP activates dnaA protein in initiating replication of 601 plasmids bearing the origin of the E. coli chromosome. Cell 50: 259-265. 602 Sever, M.J. & J.J. Wilker, (2004) Visible absorption spectra of metal-catecholate and metal-tironate 603 complexes. Dalton Trans: 1061-1072. 604 Simmons, L.A., A.M. Breier, N.R. Cozzarelli & J.M. Kaguni, (2004) Hyperinitiation of DNA replication in 605 Escherichia coli leads to replication fork collapse and inviability. Mol Microbiol 51: 349-358. 606 Skarstad, K. & T. Katayama, (2013) Regulating DNA replication in bacteria. Cold Spring Harb Perspect Biol 5: 607 a012922. 608 Takahashi, N., C.C. Gruber, J.H. Yang, X. Liu, D. Braff, C.N. Yashaswini, S. Bhubhanil, Y. Furuta, S. Andreescu, 609 J.J. Collins & G.C. Walker, (2017) Lethality of MalE-LacZ hybrid protein shares mechanistic attributes 610 with oxidative component of antibiotic lethality. Proc Natl Acad Sci U S A. 611 Terlain, B. & J.P. Thomas, (1971) [Structure of griselimycin, extracted Streptomyces 612 cultures. I. Identification of the products liberated by hydrolysis]. Bull Soc Chim Fr 6: 2349-2356. 613 Thompson, M.G., B.W. Corey, Y. Si, D.W. Craft & D.V. Zurawski, (2012) Antibacterial Activities of Iron 614 Chelators against Common Nosocomial Pathogens. Antimicrobial agents and chemotherapy 56: 615 5419-5421. 616 von Freiesleben, U., M.A. Krekling, F.G. Hansen & A. Lobner-Olesen, (2000) The eclipse period of Escherichia 617 coli. EMBO J 19: 6240-6248. 618 Wang, M., J.J. Carver, V.V. Phelan, L.M. Sanchez, N. Garg, Y. Peng, D.D. Nguyen, J. Watrous, C.A. Kapono, T. 619 Luzzatto-Knaan, C. Porto, A. Bouslimani, A.V. Melnik, M.J. Meehan, W.T. Liu, M. Crusemann, P.D. 620 Boudreau, E. Esquenazi, M. Sandoval-Calderon, R.D. Kersten, L.A. Pace, R.A. Quinn, K.R. Duncan, 621 C.C. Hsu, D.J. Floros, R.G. Gavilan, K. Kleigrewe, T. Northen, R.J. Dutton, D. Parrot, E.E. Carlson, B. 622 Aigle, C.F. Michelsen, L. Jelsbak, C. Sohlenkamp, P. Pevzner, A. Edlund, J. McLean, J. Piel, B.T. 623 Murphy, L. Gerwick, C.C. Liaw, Y.L. Yang, H.U. Humpf, M. Maansson, R.A. Keyzers, A.C. Sims, A.R. 624 Johnson, A.M. Sidebottom, B.E. Sedio, A. Klitgaard, C.B. Larson, C.A.B. P, D. Torres-Mendoza, D.J. 625 Gonzalez, D.B. Silva, L.M. Marques, D.P. Demarque, E. Pociute, E.C. O'Neill, E. Briand, E.J.N. Helfrich, 626 E.A. Granatosky, E. Glukhov, F. Ryffel, H. Houson, H. Mohimani, J.J. Kharbush, Y. Zeng, J.A. Vorholt, 627 K.L. Kurita, P. Charusanti, K.L. McPhail, K.F. Nielsen, L. Vuong, M. Elfeki, M.F. Traxler, N. Engene, N. 628 Koyama, O.B. Vining, R. Baric, R.R. Silva, S.J. Mascuch, S. Tomasi, S. Jenkins, V. Macherla, T. 629 Hoffman, V. Agarwal, P.G. Williams, J. Dai, R. Neupane, J. Gurr, A.M.C. Rodriguez, A. Lamsa, C. 630 Zhang, K. Dorrestein, B.M. Duggan, J. Almaliti, P.M. Allard, P. Phapale, et al., (2016) Sharing and 631 community curation of mass spectrometry data with Global Natural Products Social Molecular 632 Networking. Nature biotechnology 34: 828-837. 633 Weed, M.R., K.E. Vanover & W.L. Woolverton, (1993) Reinforcing effect of the D1 dopamine agonist SKF 634 81297 in rhesus monkeys. Psychopharmacology (Berl) 113: 51-52.

81 Fig. 1. Concept of the screen. A) Principle of the screen. In absence of Hda or in presence of multiple copies of DARS2, DNA replication commences too soon and/ or too often resulting in inviability. An anti-DnaA molecule that reduces DnaA activity reestablishes the initiation frequency to a level that restores viability. Such an anti-DnaA molecule reduces DnaA activity in wild-type cells to a level that no longer sustains viability. B) Schematic representation of the screening method. Hda deficient cells or wild-type cells containing a pBR322-DARS2 plasmid are propagated under permissive growth conditions, i.e. either anaerobic or in minimal poor medium. An estimated twenty thousand cells are spread on two types of agar plates: minimal poor (permissive conditions) and minimal rich (non-permissive conditions) medium. A diffusion assay is performed by punching holes in the agar and introducing 5 ml bioactive extract into each. The plates are incubated aerobically at 37oC for 16h and visually inspected. On the non-permissive conditions plates, positive “hits” are depicted by a small clearing area separating a zone of growth encircling the hole from which the specific extract has been diffusing. The same extract on permissive conditions is depicted by a small clearing area encircling the hole from which the extract has been diffusing.

82 Fig 2. Identification of deferoxamine as a hit. A) Seven extracts rescue the growth hda mutant cells. Hda deficient cells spread on minimal rich medium plates were tested against seven extracts (19H5, 19C8, 19A6, 18C2, 18H6, 18F7and 18C9). A zone of growth is visible around the holes where the 5 ml of extracts have been introduced. B) Hda deficient cells spread on minimal rich medium plates tested against HPLC separated fractions of extract 18C9. Rescuing activity is seen with fraction 5 and 6. C) LC-MS analysis of fraction 5 identifying deferoxamine as the active compound. D) Hda deficient cells or cells carrying a multi-copy DARS2 plasmid were spread on the indicated plates and tested against varying concentration of deferoxamine. 5 ml of 76, 38, 19, 9.5 and 4.25 mM deferoxamine was dispensed in separated wells.

83 Fig. 3. Deferoxamine does not affect initiation of DNA replication. The indicated cells were grown exponentially at 37 °C in minimal medium supplemented with minimal poor medium (blue) and then diluted into minimal rich medium and incubated for 4 hours at 37 °C in absence of DFO (green) or presence of DFO at a final concentration of 150mM (orange). Cells were treated with rifampicin and cephalexin prior to flow cytometric analysis. Each panel represents a minimum of 30000 cells. When relevant, the average ori/cell (O/C), ori/mass (O/M) relative to the wild-type (wt) and mass doubling time (t) are shown in the histograms. Inserts show growth of culture where no meaningful doubling time could be obtained. N.D. – Not Determined, N.R. – Not Relevant.

84 Fig. 4. Deferoxamine restores growth hda mutant cells during fast growth. A. Wild-type and Hda deficient cells were grown in LB supplemented with 150 mM DFO. Cells were diluted 5 times in LB without DFO and maintained by dilution in fresh medium for two hours. When indicated Iron II perchlorate was added at a final concentration of 200 mM to titrate DFO. Insert: Growth of hda mutant cells was followed by measuring OD600. Cells were treated with rifampicin and cephalexin prior to flow cytometric analysis. Each panel represents a minimum of 30000 cells. When relevant, the average ori/cell (O/C), ori/mass (O/M) relative to the wild-type (wt) and mass doubling time (t) are shown in the histograms. Orange histograms represent cells grown in the presence of DFO whereas green histograms were derived from cells grown/incubated without DFO for the indicated time. N.D. – Not Determined, N.R. – Not Relevant. B. DFO promotes replication fork progression in Hda deficient cells or wild-type cells containing a pBR322-DARS2 plasmid. The indicated cells were grown exponentially at 37 °C in minimal poor medium (blue), shifted to minimal rich medium and incubated for 4 hours at 37 °C in absence of DFO or presence of DFO at a final concentration of 150 mM. The ori/ter ratios were determined by qPCR analysis. Shown is the mean ± s.d. (n=3).

85 Fig. 5. The effect of iron chelators and reducing agent can be eliminated by addition of excess iron. A. Hda deficient cells were spread on minimal rich medium agar plates containing iron (III) chloride at a final concentration of 3 mM, iron (II) perchlorate at a final concentration of 3 mM or iron (II) perchlorate at a final concentration of 200 mM and tested against metal chelators and antioxidant. 5 ml of 10 mM DFO, 10 mM phenanthroline, 10 mM EDTA, 300 mM bipyridil or 650 mM DTT was dispensed in separated wells. B. Hda deficient cells capable of producing SeqA or a cyclic DnaA domain I derived peptide were spread on minimal rich medium agar plates containing 3 mM iron (III) chloride , 3 mM iron (II) perchlorate or 200 mM iron (II) perchlorate. 5 ml of 100 mM IPTG was dispensed in separated wells to induce the overexpression of SeqA or a cyclic DnaA domain I. 5 ml of 10 mM DFO was used as control.

86 Fig. 6. S143 chelates iron. A. Hda mutant cells were plated on minimal rich medium agar plates containing 3 mM iron (III), 3 mM iron (II) perchlorate or 200 mM iron (II) perchlorate were tested against 5 ml of 10 mM S143. B. Wild-type cells were plated on minimal poor medium agar plates containing either 3 mM iron (II) perchlorate or 3 mM or 200 mM iron (II) perchlorate and tested against 5 ml of 10 mM S143. C. Iron binding of S143 and DFO was assayed by monitoring the absorbance at 510nm of the Fe (II)-Phenanthroline complex. Increasing amounts of DFO or S143 was mixed with iron (II) perchlorate (0.015 mM final concentration) and absorbance at 510nm was measured following addition of phenanthroline (1mM final concentration). The absorbance relative to Fe (II)- Phenanthroline is plotted. D. Absorption spectrum of 1mM S143 in ddH2O alone or complexed with 0.2 mM or 0.4 mM iron (III) nitrate. 87 Fig. 7. Model structure for DFO and S143 chelating iron A single DFO molecule forms six bonds with iron (III) while up to three S143 molecules can interact with one iron (III) through chelation by their catechol moieties.

88 Figure S1

0.8

0.7 32 g DFO ml-1 0.6 16 g DFO ml-1 0.5 8 g DFO ml-1 -1 0.4 4 g DFO ml -1

O.D. 600 nm 2 g DFO ml 0.3 1 g DFO ml-1 0.2 0.5 g DFO ml-1 0.1 no DFO

0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Time (hours)

Fig. S1 DFO Minimal Recovery Concentration. Hda cells pre-grown in minimal poor medium were shifted to minimal rich medium at 37 °C in the presence of DFO at different concentration (see experimental procedures). The growth was moni- tored by measuring optical density in a microplate reader. Hda deficient cells grown with 32 to 0.5 μg ml-1 of DFO are shown. Cells grown with 8 μg DFO ml-1 and above started growth earlier than those grown with 4 μg DFO ml-1 and below. The late grown cells may contain mutations suppress- ing hda.

89 Figure S2

100 g DFO ml-1 50 g DFO ml-1 25 g DFO ml-1 10 g DFO ml-1 O.D.450 nm (log) no DFO

50 60 70 80 90 100 110 120 130 140 150

time (min)

Fig. S2 DFO Minimal Recovery Concentration.

The effect of DFO on wild type growth. Wild type cells were grown in minimal rich medium and maintained exponentially growing in absence or in presence of 10, 25, 50 or 100 μg ml-1 DFO.

Growth is monitored by optical density measurement.

90 Figure S3

glu +casa

3M Iron(III) Chloride

200M Iron(II) perchlorate

DFO DMSO

Fig. S3 Excess iron counteracts the effect of DFO in the pBR322-DARS2 screen.

Cells carrying a multi-copy DARS2 plasmid were spread on minimal rich medium agar plates containing iron (III) chloride at a final concentration of 3 μM or iron (II) perchlorate at a final concentration of 200 μM and tested against DFO. 5 μl of 10 mM DFO or DMSO was dispensed in separated wells.

91 Figure S4

wt wt 150M DFO wt 150M DFO 200 M iron (II) perchlorate O.D. 600 nm (log)

010 30 50 time (min)

Fig. S4 The effect of DFO on wild type growth is counteracted by iron.

Wild type cells were grown in minimal rich medium and maintained exponentially growing in absence of DFO, in presence of 150μM DFO or 150μM DFO and excess iron (II). Growth was monitored by optical density measurement.

92 mM finalconcentr Absorption spectrumofi F i g. S5S143chelatesir ation) on. and phenanth n creasin

g amountsofS143 Absorbance (AU) 0.02 0.04 0.06 0.08 0.12 0.14 0.16 0.18 roline 0.1 0.2 0 35 0 Figure S5 (1mM f 45  (nm) 0 93 55 0 i nal conc in ddH2Omi 65 0 0.05 mMS143 No S143 10 mMS143 1.2 mMS143 0.3 mMS143 entration) x e d wi .

th iron( I I ) pe rchlor a te (0.0 20

Paper III: A Novel Fluorescence Based Screen for Inhibitors of the Initiation of DNA Replication in Bacteria.

Currently in review at: Current Drug Discovery Technologies.

94 Send Orders for Reprints to [email protected] Journal Name, Year, Volume 1 A Novel Fluorescence Based Screen for Inhibitors of the Initiation of DNA Replication in Bacteria

Rasmus N. Klitgaarda and Anders Løbner-Olesena* aDepartment of Biology, Faculty of SCIENCE, University of Copenhagen, Copenhagen, Denmark.

Abstract: Background: One of many strategies to overcome antibiotic resistance is the discovery of compounds targeting cellular processes, which have not yet been exploited. Methods and materials: Using various genetic tools, we constructed a novel high throughput, cell based, fluorescence screen for inhibitors of chromosome replication initiation in bacteria. Results: The screen was validated by expression of an intra-cellular cyclic peptide interfering with the initiator protein DnaA and by over- expression of the negative initiation regulator SeqA. We also demonstrate that neither nor ciprofloxacin triggers a false positive result. Finally, 400 extracts isolated mainly from filamentous actinomycetes were subjected to the screen. Conclusion: We conclude that the presented screen is applicable for identifying putative inhibitors of DNA replication initiation in a high throughput setup. Keywords: DNA replication initiation, inhibitors, DnaA, high throughput screen, fluorescence, microbial extracts. 1. INTRODUCTION strand[12]. Following duplex opening the nucleoprotein Antibiotic resistance is one of the major health care complex loads the DnaB helicase, with help from the problems in the world; therefore, it is important to discover helicase loader protein DnaC [13, 14]. Loading of DnaB then compounds targeting unexploited processes essential to the triggers the assembly of the remaining parts of the growth or viability of bacteria. One such process is replication machinery[12]. The initiation process, including replication of the bacterial chromosome. Targeting the DNA the DnaC assisted loading of the DnaB helicase, is highly replication is attractive for a number of reasons; i) The conserved across bacterial species [15], and is therefore an replisome number per cell is low, ii) The replication complex interesting target for novel antibiotics. is a multi-protein machinery and therefore contains a large E. coli rnhA mutants, lacking RNase HI activity, are number of potential targets, iii) Key components of the able to initiate the DNA replication by a protein synthesis replication machinery is well conserved and has low and DnaA/oriC independent pathway called constitutive sequence homology with human replication proteins and iiii) Stable DNA Replication (cSDR)[16, 17]. cSDR is initiated The DNA replication is an under exploited target, this far at a number of alternative sites on the chromosome, termed only type-II topoisomerase inhibitors are used in the oriK[18]. It is believed that lack of RNase HI activity leads clinic[1]. A number of different compounds have been to stabilization of nascent RNA transcripts, which anneals to identified targeting components of the replication machinery the DNA template behind the moving RNA polymerase, including; DNA ligase (LigA)[2, 3], DNA polymerase III[4, creating an R-loop. The RNA is thought to act as a primer 5], the sliding clamp[6, 7] and single-stranded DNA-binding for extension by DNA polymerase I, creating a D-loop like proteins[8]. In contrast, only a few efforts have been made to structure. This structure is then bound by PriA, initiating discover inhibitors of the initiation process of the bacterial assembly of the PriA dependent primosome, loading of the DNA replication [9-11]. DnaB helicase and assembly of the replisome [19, 20]. The In most bacteria, replication of the chromosomal DnaA/oriC independency of cSDR makes it a valuable tool DNA is initiated from a single origin of replication, termed when searching for inhibitors targeting the initiation of DNA oriC. The initiation process is best characterized in replication. Escherichia coli, where DNA replication is initiated by We here present a high throughput fluorescence based screen, which can be used to specifically screen for binding of the initiator protein DnaA, in its active ATP- inhibitors that targets DNA replication initiation. bound form, to the oriC. When sufficient DnaAATP molecules are bound to the oriC it forms a nucleoprotein *Address correspondence to this author at the Department of Biology, Faculty of SCIENCE, University of Copenhagen, Copenhagen, Denmark; complex responsible for separation of the DNA double E-mail: [email protected]

XXX-XXX/14 $58.00+.00 © 2014 Bentham Science Publishers

95 2 Journal Name, 2014, Vol. 0, No. 0 Klitgaard et al.

pALO17 (FW) ACAGCTTATCATCG Amplification of sopABC fragment from GATCGAATTCCTCGACAG 2. MATERIALS AND METHOD pALO17 (RV) CGACACACTT Amplification of GFPmut2 from pKEN GATCGAATTCGCGTGTTG

2.1 Bacterial strains and plasmids GFP mut2 and adding the PR-promoter ACTATTTTACCTCTGGCGG R MG1655ΔrnhA::kan (unpublished, ALO4523) and FW TGATAATGGTTGCATGTAC MG1655ΔrnhA::KanR, ΔoriC::CamR (unpublished, TAAGGAGGTTGTATGAGT ALO4524) were obtained from the laboratories strain AAAGGAGAAGAACTTTTC ACTGGAG collection. The construction of pMAK7 is described in ref. Amplification of GFPmut2 from pKEN CTTACTCGAGTTATTTGTA

[21]. pRNK4 was constructed by PvuI digest of pSC116[7] GFP mut2 and adding the PR-promoter TAGTTCATCCATGCCATGT followed by re-ligation, thereby removing the RV GTAATCC chloramphenicol resistance gene and reconstituting the Deletion of oriC FW TTGCCTGGTAAGCGGGTG ampicillin resistance gene. CTTACCAGGCATTTTTAAT GCGGTGTAGGCTGGAGCT 2.2 Construction of the mini-chromosome pRNK6 GCTTC A fragment from the mini-chromosome pRNK3 Deletion of oriC RV GCCTACAGGATGTCGGTG (unpublished) containing the oriC and cI was PCR CACAGATTCGCCAGGCAC AACAATGGGAATTAGCCA amplified and ligated using MunI and EcoRI into the TGGTCC OriR6K-dependent vector pSW25T[22], creating pRNK5. The expression of cI is controlled by a constitutive synthetic promoter (J23101) from the Anderson promoter 2.5 Fluorescence screen collection[23]. The promoter region was inserted upstream Overnight cultures were grown at 370C with the of cI by PCR amplification of cI from pSB4293[24]. To appropriate antibiotics (40 µg/mL kanamycin, 20 µg/mL stabilize the mini- chromosome a PCR fragment, from chloramphenicol, 50 µg/mL streptomycin, 150 µg/mL pALO17[25], containing the SopABC partitioning system ampicillin) in AB media supplemented with 0.2% glucose was ligated into pRNK5 using EcoRI, creating pRNK6. and 1% casamino acids (ABTG CAA). The overnight cultures were diluted to OD600 = 0.001 and 100 µL of the 2.3 Construction of the screen strain (MG1655ΔrnhA, diluted culture was added to the wells of a 96-well plate, R R ΔoriC::Cam , attB::PR-GFPmut2,Kan ) already containing 100 µL ABTG CAA, with the correct R MG1655ΔrnhA::kan was the starting point in the antibiotics and IPTG (final conc. 0.25mM). The plate was construction of the strain used in the screen. First, the KanR sealed with a Breathe-easy®sealing membrane (Diversified cassette was flipped out by expression of FLP recombinase Biotech) and incubated at 370C for 20 hours while shaking in from pCP20[26]. Hereafter, the lambda PR-promoter was a plate shaker (800 RPM). Following incubation, the plates fused to GFPmut2 by PCR amplification from pKEN GFP were centrifuged, the supernatant removed, the cells mut2[27]. The fragment was then inserted at the lambda resuspended in 200 µL 0.9% NaCl and transferred to clear attachment site (attB) on the E. coli chromosome as bottomed black sided 96-well plate. OD600 and fluorescence described in ref. [28], creating MG1655ΔrnhA, attB::PR- was detected using a BIOTEK synergy H1 plate reader. The R GFPmut2, Kan . Finally, the oriC was removed by lambda following settings were used for the fluorescence red recombination [26], deleting a region in the chromosome measurement, excitation: 485nm and emission: 528nm. The from the start codon of viaA to the stop codon of mnmG. treatment with tetracycline and ciprofloxacin was carried out as minimal inhibitory concentration assay by micro dilution 2.4 Primers in a 96-well plate, the plate was otherwise treated as stated Table 1. Primers used for construction of the screen. above. The MIC for tetracycline and ciprofloxacin was 1 and Use Sequence Amplification of cI from pSB4293 and TATAGAGCTCTTTACAGCT 0,015 µg/ml, respectively. adding the Anderson promoter J23101 AGCTCAGTCCTAGGTATTAT FW GCTAGCGCGGTGATAGATT 2.6 Preparation of microbial extracts TAACGTATGAGCA The extracts subjected to the screen were prepared Amplification of cI from pSB4293 and GATCGAGCTCTCAGCCAA by Naicons srl., Milan, Italy. Cultures of filamentous adding the Anderson promoter J23101 ACGTCTCTTCAGG RV actinomycetes (10 mL) were centrifuged (3000 rpm / 10 Amplification of cI/oriC fragment from GATCCAATTGGCCTGACG minutes) to separate the cells from the supernatant. Ethanol pRNK3 (FW) GTAGAGCACACGAT (4 mL) was added to the pellet and incubated for 1h at room Amplification of cI/oriC fragment from GTATAGAATTCCCGATCAT temperature with shaking: 0.2 ml aliquots of the ethanolic pRNK3 (RV) GCGTACCATCAAG extract were distributed in 96-well microtiter plates, dried Amplification of sopABC fragment from TATAGAATTCTCATGTTTG under vacuum, and stored at 4°C. HP20 resin (Mitsubishi

96 Short Running Title of the Article Journal Name, 2014, Vol. 0, No. 0 3

Chemical Co., 1 mL) was added to the supernatant and incubated for 2 hours at room temperature with shaking; the resin was washed with 6 mL H2O, and eluted with 5 mL MeOH: 0.25 mL aliquots were distributed in 96-well microtiter plates, dried under vacuum, and stored at 4°C.

2.7 Extract screening Dried extracts were dissolved in 99.6% DMSO and diluted with water to a final concentration of 10% DMSO. 10 µL of each extract was suspended into the wells of a 96- well polypropylene plate. An overnight culture of the screen strain was grown and diluted as described above and 190µL was added to each well containing the extracts. 190µL of the following control strains were added to wells with 10µL R 10% DMSO; i) MG1655ΔrnhA, ΔoriC::Cam , attB::PR- R R GFPmut2,Kan , ii) MG1655ΔrnhA, ΔoriC::Cam , attB::PR- GFPmut2,KanR/pRNK6, iii) MG1655ΔrnhA::KanR, ΔoriC::CamR. The incubation and measurements were performed as described in the section above. To assess the auto-fluorescence of the extracts, the screen was performed in parallel using the non-fluorescing strain; MG1655ΔrnhA::KanR, ΔoriC::CamR.

2.8 Analysis of fluorescence data The fluorescence data obtained from each well in the 96-well plate was first adjusted for background fluorescence. This was done by subtracting the fluorescence Fig. 1. Graphical representation of the screen. A) cI expressed from the mini-chromosome pRNK6, represses the of the non-GFP strain, MG1655ΔrnhA::KanR, ΔoriC::CamR. expression of GFPmut2 from the PR-promoter. B) If the The fluorescence from each well was compared to the initiation of replication by DnaA is blocked, pRNK6 is lost R fluorescence of MG1655ΔrnhA, ΔoriC::Cam , attB::PR- overtime. Hence, the repression of the PR-promoter is GFPmut2,KanR/pRNK6, grown in ABTG CAA with 0.5% released and GFP is expressed. DMSO. Following construction of the strain,

3. RESULTS MG1655ΔrnhA, ΔoriC attB::PR-GFPmut2, it was verified by 3.1 Construction and validation of the screen microscopy that the cells were fluorescing, confirming that The screen utilizes an E. coli rnhA, oriC mutant that GFP was expressed from the PR-promoter (Figure 2AB). replicates its chromosomal DNA by cSDR. As the strain Hereafter it was assessed if cI expressed from pRNK6 does not initiate from oriC it is insensitive to putative reduced expression of GFPmut2 from the lambda PR- inhibitors of the oriC dependent initiation process. The promoter. This was done by quantifying the fluorescence, fluorescence reporter system consists of GFPmut2[27] following 20 hour incubation at 370C, for MG1655ΔrnhA, expressed from a lambda phage PR-promoter inserted into ΔoriC attB::PR-GFPmut2 with and without pRNK6. The the attB site on the chromosome. Transcription from the PR- results show that the presence of pRNK6 reduced promoter is repressed by the lambda phage cI repressor, fluorescence by more than 50% relative to the strain without which is constitutively expressed on a plasmid (pRNK6) that pRNK6 (Figure 2C). Indicating that cI is repressing the only replicates from oriC, also known as a mini- expression of GFP as expected. MG1655ΔrnhA, ΔoriC chromosome. The rationale behind the screen is that attB::PR-GFPmut2/pRNK6 is from now on referred to as the Inhibition of mini-chromosome replication will lead to loss screen strain. of cI expression and to expression of GFP, which can be To validate the screen further, we introduced a detected as an increase in fluorescence in a high-throughput plasmid, pRNK4, that encodes a cyclic peptide previously setup. (Figure 1). shown to inhibit DnaA function and hence replication initiation [7]. In the screen strain expression of the DnaA inhibitor by IPTG induction, led to a 50% increase in

97 4 Journal Name, 2014, Vol. 0, No. 0 Klitgaard et al. fluorescence relative to the un-induced control (Figure 2D). antibiotics, would lead to false-positives in the screen. The screen was further verified by over-expression of SeqA Specifically, we were interested in the effect of inhibiting protein from the plasmid pMAK7. SeqA regulates the translation or DNA replication elongation. The screen was replication initiation by binding to hemi-methylated GATC therefore subjected to sub-inhibitory levels (0.5xMIC and sequences in oriC, thereby sterically hindering initiation by 0.25xMIC) of either tetracycline, an inhibitor of translation, DnaA[29]. Over-expression of SeqA, by IPTG induction, or the DNA replication elongation inhibitor, ciprofloxacin. increased the fluorescence by 30% relative to the un-induced This resulted in fluorescence signals that were lower relative control (Figure 2E). Overall these results verifies that if to the untreated sample (Figure 3), indicating that sub- replication of the oriC dependent mini-chromosome is inhibitory concentrations of antibiotics, targeting translation inhibited, repression of the PR-promoter is released and GFP or DNA replication elongation, does not lead to false is expressed, leading to an increase in the fluorescence positives. signal.

Fig. 3. Relative fluorescence of the screen strain treated with 0.25x and 0.5xMIC of tetracycline or ciprofloxacin.

3.3 Screening microbial extracts In an initial attempt to discover a novel inhibitor of the initiation of DNA replication, we screened 400 microbial extracts, mainly derived from filamentous actinomycetes. None of the extracts gave a positive hit in the screen, suggesting that initiation inhibitors are rare and that a high number of extracts needs to be screened in order to get a positive hit.

4. DISCUSSION

Fig. 2. Validation of the screen. A) Fluorescence In the battle against antibiotic resistance, it is microscopy of the strain, MG1655ΔrnhA, ΔoriC attB::PR- important to discover and develop novel compounds GFPmut2. B) Phase contrast of the same cells as in picture targeting essential cellular processes. The screen presented A. C) Verification that cI, expressed from pRNK6, reduces here could become a valuable tool in identifying inhibitors of expression of GFP by repression of the lambda PR-promoter. DNA replication initiation in bacteria. We expected that D) Verification of the screen, by expression of a DnaA expression of cI from pRNK6 would repress the expression inhibitor (pRNK4) and E) over-expression of seqA of GFP from the PR-promoter. Quantifying the relative (pMAK7) in the screen strain. fluorescence with and without pRNK6, showed a partial reduction in the fluorescence signal of around 50%. This 3.2 Impact of sub-inhibitory concentrations of antibiotics could indicate that the PR-promoter is leaky, or more likely it on the screen. is a result of instability and loss of the mini-chromosome, as Natural extracts from plants and microbes are mini-chromosomes become more unstable at slow growth frequently used when screening for novel antibiotics, these [30]. As there are no known replication initiation inhibitors extracts are usually complex and often contain several that can cross the E. coli cell membrane, the screen was compounds with antibacterial properties. It is therefore validated by expression of a cyclic peptide inhibiting the important to asses if sub-inhibitory concentrations of

98 Short Running Title of the Article Journal Name, 2014, Vol. 0, No. 0 5 function of DnaA and over-expression of the negative DiaA[34], H-NS[35], Fis[36], IHF[37] and HU[38]. Though initiation regulator SeqA. it should be noted that only DiaA, FIS and HU mutants are We wondered whether non-lethal inhibition of unable to stably maintain mini-chromosomes[34, 38, 39], translation or DNA replication elongation could affect hence inhibition of IHF or H-NS may not lead to a positive segregation of the mini-chromosome or the fluorescence hit in the screen[37, 40]. reporter system, in a manner that would lead to false positives. Subjecting the screen to sub-inhibitory CONFLICT OF INTEREST concentrations of the translation inhibitor, tetracycline, or the The authors declare no conflict of interest. DNA replication inhibitor, ciprofloxacin, confirmed that this was not the case. Overall, these results show that the screen ACKNOWLEDGEMENTS is applicable for identifying novel inhibitors of the initiation We acknowledge financial support from the University of of DNA replication. However, it should be noted that other Copenhagen Centre for Control of Antibiotic Resistance classes of inhibitors might influence the segregation of the (UC-Care) and by the Center for Bacterial Stress Response mini-chromosome, without inhibiting the growth of the and Persistence (BASP) funded by a grant from the Danish bacteria, leading to an increase in fluorescence. Especially National Research Foundation (DNRF120). Finally, we putative inhibitors of the SopABC segregation system could would like to thank NAICONS srl. for preparing and sharing lead to false positives. However, one could differentiate their microbial extracts. between a replication initiation inhibitor and a SopABC inhibitor, as the later would most likely not have an effect on REFERENCES [1] van Eijk E, Wittekoek B, Kuijper EJ, Smits WK. the growth of a wild-type E. coli strain. Furthermore, it is DNA replication proteins as potential targets for well known that transcription from the mioC gene promoter, antimicrobials in drug-resistant bacterial pathogens. Journal into the oriC, is important to maintain mini-chromosome of Antimicrobial Chemotherapy. 2017;72(5):1275-84. stability and copy number, but that it does not stimulate the [2] Surivet JP, Lange R, Hubschwerlen C, Keck W, initiation of chromosomal replication[31]. Inhibiting Specklin JL, Ritz D, et al. Structure-guided design, synthesis transcription from the mioC promoter could therefore and biological evaluation of novel DNA ligase inhibitors potentially lead to a false positive result in the screen. with in vitro and in vivo anti-staphylococcal activity. Bioorganic & medicinal chemistry letters. None of the 400 extracts that were screened gave a 2012;22(21):6705-11. positive hit. However, we gained valuable insight about the [3] Mills SD, Eakin AE, Buurman ET, Newman JV, practical procedure for the screen. Furthermore, it indicates Gao N, Huynh H, et al. Novel bacterial NAD+-dependent that the screen is specific and that a larger number extracts or DNA ligase inhibitors with broad-spectrum activity and compounds needs to be screened in order to get a positive antibacterial efficacy in vivo. Antimicrobial agents and hit. The specificity of the screen was not unexpected, as the chemotherapy. 2011;55(3):1088-96. [4] Rose Y, Ciblat S, Reddy R, Belley AC, Dietrich E, main difference between replication by cSDR and replication Lehoux D, et al. Novel non-nucleobase inhibitors of from the oriC is the initiation by DnaA[20]. Even the DnaB Staphylococcus aureus DNA polymerase IIIC. Bioorganic & helicase and its chaperone, DnaC, have been shown to be medicinal chemistry letters. 2006;16(4):891-6. required for cSDR[17]. As the initiation by DnaA is a multi- [5] Tarantino PM, Jr., Zhi C, Wright GE, Brown NC. step process, direct inhibition hereof can potentially happen Inhibitors of DNA polymerase III as novel antimicrobial at multiple points. Early in the process, the initial binding of agents against gram-positive eubacteria. Antimicrobial DnaA to the oriC could be hindered by blocking the agents and chemotherapy. 1999;43(8):1982-7. [6] Kling A, Lukat P, Almeida DV, Bauer A, Fontaine interaction between the HTH DNA-binding motif in domain E, Sordello S, et al. Antibiotics. Targeting DnaN for IV of DnaA and the oriC[32]. Furthermore, oligomerization tuberculosis therapy using novel griselimycins. Science. of DnaA at the oriC and formation of the nucleoprotein 2015;348(6239):1106-12. complex, could be inhibited by targeting either domain I or [7] Kjelstrup S, Hansen PMP, Thomsen LE, Hansen III of DnaA, that have been shown to mediate the PR, Løbner-Olesen A. Cyclic Peptide Inhibitors of the β- oligomerization [12, 15]. Finally, the process of loading the Sliding Clamp in Staphylococcus aureus. DnaB helicase, mediated by interactions between regions of PloS one. 2013;8(9):e72273. [8] Marceau AH, Bernstein DA, Walsh BW, Shapiro domain I and III of DnaA and the DnaB helicase [33], might W, Simmons LA, Keck JL. Protein interactions in genome also serve as a target for putative inhibitors of the initiation maintenance as novel antibacterial targets. PloS one. process that would be identified by the screen presented 2013;8(3):e58765. here. In addition to direct inhibition of DnaA, the initiation [9] Fossum S, De Pascale G, Weigel C, Messer W, process might also be subdued by putative inhibitors Donadio S, Skarstad K. A robust screen for novel antibiotics: targeting factors stimulating the initiation process, including; specific knockout of the initiator of bacterial DNA replication. FEMS Microbiol Lett. 2008;281(2):210-4.

99 6 Journal Name, 2014, Vol. 0, No. 0 Klitgaard et al.

[10] Johnsen L, Weigel C, von Kries J, Moller M, Proceedings of the National Academy of Sciences of the Skarstad K. A novel DNA gyrase inhibitor rescues United States of America. 2013;110(43):17386-91. Escherichia coli dnaAcos mutant cells from lethal [25] Jensen MR, Lobner-Olesen A, Rasmussen KV. hyperinitiation. The Journal of antimicrobial chemotherapy. Escherichia coli minichromosomes: random segregation and 2010;65(5):924-30. absence of copy number control. Journal of molecular [11] Yamaichi Y, Duigou S, Shakhnovich EA, Waldor biology. 1990;215(2):257-65. MK. Targeting the Replication Initiator of the Second Vibrio [26] Datsenko KA, Wanner BL. One-step inactivation of Chromosome: Towards Generation of Vibrionaceae-Specific chromosomal genes in Escherichia coli K-12 using PCR Antimicrobial Agents. PLOS Pathogens. products. Proceedings of the National Academy of Sciences 2009;5(11):e1000663. of the United States of America. 2000;97(12):6640-5. [12] Erzberger JP, Mott ML, Berger JM. Structural basis [27] Cormack BP, Valdivia RH, Falkow S. FACS- for ATP-dependent DnaA assembly and replication-origin optimized mutants of the green fluorescent protein (GFP). remodeling. Nature Structural &Amp; Molecular Biology. Gene. 1996;173(1 Spec No):33-8. 2006;13:676. [28] Atlung T, Nielsen A, Rasmussen LJ, Nellemann LJ, [13] Davey MJ, Fang L, McInerney P, Georgescu RE, Holm F. A versatile method for integration of genes and O’Donnell M. The DnaC helicase loader is a dual ATP/ADP gene fusions into the lambda attachment site of Escherichia switch protein. The EMBO journal. 2002;21(12):3148-59. coli. Gene. 1991;107(1):11-7. [14] Fang L, Davey MJ, O'Donnell M. Replisome [29] Skarstad K, Katayama T. Regulating DNA Assembly at oriC, the Replication Origin of E. coli, Reveals Replication in Bacteria. Cold Spring Harbor Perspectives in an Explanation for Initiation Sites outside an Origin. Biology. 2013;5(4):a012922. Molecular Cell. 1999;4(4):541-53. [30] Lobner-Olesen A, Atlung T, Rasmussen KV. [15] Messer W. The bacterial replication initiator DnaA. Stability and replication control of Escherichia coli DnaA and oriC, the bacterial mode to initiate DNA minichromosomes. Journal of bacteriology. replication. FEMS Microbiol Rev. 2002;26(4):355-74. 1987;169(6):2835-42. [16] Kogoma T, von Meyenburg K. The origin of [31] Løbner-Olesen A, Boye E. Different effects of replication, oriC, and the dnaA protein are dispensable in mioC transcription on initiation of chromosomal and stable DNA replication (sdrA) mutants of Escherichia coli minichromosomal replication in Escherichia coli. Nucleic K-12. The EMBO journal. 1983;2(3):463-8. Acids Research. 1992;20(12):3029-36. [17] Kogoma T. A novel Escherichia coli mutant [32] Erzberger JP, Pirruccello MM, Berger JM. The capable of DNA replication in the absence of protein structure of bacterial DnaA: implications for general synthesis. Journal of molecular biology. 1978;121(1):55-69. mechanisms underlying DNA replication initiation. The [18] de Massy B, Fayet O, Kogoma T. Multiple origin EMBO journal. 2002;21(18):4763-73. usage for DNA replication in sdrA(rnh) mutants of [33] Seitz H, Weigel C, Messer W. The interaction Escherichia coli K-12. Initiation in the absence of oriC. domains of the DnaA and DnaB replication proteins of Journal of molecular biology. 1984;178(2):227-36. Escherichia coli. Molecular microbiology. 2000;37(5):1270- [19] Martel M, Balleydier A, Sauriol A, Drolet M. 9. Constitutive stable DNA replication in Escherichia coli cells [34] Ishida T, Akimitsu N, Kashioka T, Hatano M, lacking type 1A topoisomerase activity. DNA repair. Kubota T, Ogata Y, et al. DiaA, a Novel DnaA-binding 2015;35:37-47. Protein, Ensures the Timely Initiation of Escherichia coli [20] Kogoma T. Stable DNA replication: interplay Chromosome Replication. Journal of Biological Chemistry. between DNA replication, homologous recombination, and 2004;279(44):45546-55. transcription. Microbiology and molecular biology reviews : [35] Katayama T, Takata M, Sekimizu K. The nucleoid MMBR. 1997;61(2):212-38. protein H-NS facilitates chromosome DNA replication in [21] von Freiesleben U, Krekling MA, Hansen FG, Escherichia coli dnaA mutants. Journal of bacteriology. Lobner-Olesen A. The eclipse period of Escherichia coli. 1996;178(19):5790-2. The EMBO journal. 2000;19(22):6240-8. [36] Filutowicz M, Ross W, Wild J, Gourse RL. [22] Demarre G, Guérout A-M, Matsumoto-Mashimo C, Involvement of Fis protein in replication of the Escherichia Rowe-Magnus DA, Marlière P, Mazel D. A new family of coli chromosome. Journal of bacteriology. 1992;174(2):398- mobilizable suicide plasmids based on broad host range 407. R388 plasmid (IncW) and RP4 plasmid (IncPα) conjugative [37] Filutowicz M, Roll J. The requirement of IHF machineries and their cognate Escherichia coli host strains. protein for extrachromosomal replication of the Escherichia Research in microbiology. 2005;156(2):245-55. coli oriC in a mutant deficient in DNA polymerase I activity. [23] Anderson JC. Anderson promoter collection, The New biologist. 1990;2(9):818-27. available at: [38] Ogawa T, Wada M, Kano Y, Imamoto F, Okazaki http://parts.igem.org/Promoters/Catalog/Anderson T. DNA replication in Escherichia coli mutants that lack protein HU. Journal of bacteriology. 1989;171(10):5672-9. [24] Norregaard K, Andersson M, Sneppen K, Nielsen [39] Gille H, Egan JB, Roth A, Messer W. The FIS PE, Brown S, Oddershede LB. DNA supercoiling enhances protein binds and bends the origin of chromosomal DNA cooperativity and efficiency of an epigenetic switch. replication, oriC, of Escherichia coli. Nucleic Acids Research. 1991;19(15):4167-72.

100 Short Running Title of the Article Journal Name, 2014, Vol. 0, No. 0 7

[40] Langer U, Richter S, Roth A, Weigel C, Messer W. replication origin, oriC, of Escherichia coli. Molecular A comprehensive set of DnaA-box mutations in the microbiology. 1996;21(2):301-11.

Received: March 20, 2014 Revised: April 16, 2014 Accepted: April 20, 2014

101

Discussion

Antibiotic resistance has become an urgent problem, which not only threatens the future of health care, as we know it today, but also might turnout out to be a heavy burden for the global economy (12, 13). An effective strategy to overcome antibiotics resistance will need to be multidimensional. Thus, facilitate a sustainable use of antibiotics in the clinic as well as in the industry, diminish the spread of infectious disease, and encourage development of novel drugs and preservation of existing antibiotics. In the last 10 years over 50 national and international initiatives have been founded with the goal of encouraging research and development of antibiotics (258). Though action has already been taken; evidence and experience show that the current market for antibiotics does not foster major investments from the large pharmaceutical companies, who have set their course for more profitable markets (12). We have contributed to the fight against antibiotic resistance, by searching for potential helper drug targets to reverse quinolone resistance and thereby preserve the efficacy of one of the most important classes of antibiotics. Furthermore, we have developed and verified two distinct strategies for the discovery of novel classes of antibiotics targeting the initiation of chromosomal DNA replication in bacteria.

Potentiation of the quinolones

In paper I, we sought to identify targets for potentiation of ciprofloxacin by introduction of more than twenty separate single gene deletions in a high-level ciprofloxacin resistant strain. None of the tested gene deletions rendered the high-level resistant strain clinically susceptible. However, deletion of acrA, tolC, recA or recC decreased the MIC of a low-level ciprofloxacin resistant strain beneath the clinical break point. Indicating that inhibition of the AcrAB-tolC efflux-pump or HR repair of DNA DSBs, via RecA or RecC inhibition, is a plausible strategy for reversal of low-level ciprofloxacin resistance. These findings are in agreement with the observations made by Tran et al. and Recacha et al.(129, 130). The discovery and development of putative inhibitors of RecA, and thereby the SOS response, is attractive for several reasons. Most bactericidal antibiotics are inducers of the SOS response (55), thus inhibition of RecA might not only potentiate the quinolones but also other classes of bactericidal antibiotics. The SOS response also plays an important role in the evolution of antibiotic resistance, by inducing horizontal gene transfer (HGT) of antibiotic resistance genes (259) and by promoting mutagenesis via the error prone DNA polymerases IV and V (86). In addition, induction of the SOS response has been shown to promote HGT of pathogenicity associated genes in E. coli and S.

102 aureus. Thus, RecA inhibitors could potentially exert a dual mode of action in increasing antibiotic susceptibility and decreasing evolution of antibiotic resistance and pathogenicity. As described earlier, multiple efforts have been made to identify putative inhibitors of RecA, leading to the discovery of several compounds that blocks the ATPase activity of RecA in vitro (121, 122, 124). It is unknown if any these compounds are still under development. Copper phtalocyanine-3, 4´, 4´´,4´´´-tetrasulfonic acid (CuPTA) and suramin are the only compounds that evidently inhibits RecA in vivo (119, 125). In paper I, we report that CuPTA does not change the ciprofloxacin MIC for a high- or low-level ciprofloxacin resistant strain. Indicating that either CuPTA is a weak RecA inhibitor or that a given inhibitor needs to completely inactivate RecA for potentiating ciprofloxacin. Hence, development of an effective RecA inhibitor in regards to potentiation of the quinolones and other bactericidal antibiotics might prove difficult. In contrast to RecA, there only exist a single report of identification of putative inhibitors of the RecBCD complex, however the potential of the identified compounds to potentiate the quinolones has not been assessed (260). Inhibition of efflux pumps, like AcrAB-TolC, is a well investigated mean of potentiating antibiotics. Similar to inhibition of RecA, efflux pump inhibition generates some desirable side effects in addition to decreasing the MIC. In E. coli, deletion of acrAB delays the emergence of levofloxacin resistance (261), while the virulence of Salmonella enterica to some extend relies on the functionality of its drug efflux systems (262). A described earlier several EPIs that targets the AcrAB-TolC efflux pump have been identified. However, not a single EPI has made it through clinical trials and into the clinic (263, 264). One of the major challenges in EPI development is the broad compound specificity exerted by most efflux pumps. Consequently, it is challenging to setup guidelines for discriminating between efflux pump substrates and inhibitors, making it difficult to pick suitable compound libraries for screening (264). Current challenges in clinical development of EPIs are; toxicity, pharmacokinetics, potency and spectrum of activity (265). Hence, further insight into the structure and function of efflux pumps is essential for successful discovery and development of EPIs in the future. Tran et al. showed that combinatorial disruption of the AcrAB-TolC efflux pump and the SOS response rendered a high-level ciprofloxacin resistant strain clinically susceptible (129). However, deployment of this observation in the clinic would require a three drug combinatorial treatment. Experience from combinatorial drug therapy of cancer, has shown that treatment with multiple drugs is challenging due to overlapping toxicities and differences in pharmacological profiles (266). Indicating, that it would be difficult to develop such a treatment, though there is no doubt that combinatorial inhibition of drug efflux and the SOS response, would be a powerful weapon in overcoming antibiotic resistance.

103 Based on the findings presented in paper I, it seems that reversing high-level ciprofloxacin resistance with a single helper drug is not possible. However, during the gene deletion analysis, we were not able to delete priA in the high-level resistant strain, LM693. A similar observation was reported by Cirz et al., who were unable to construct an E. coli gyrA(S83L) ΔpriA mutant (72). PriA is the initiator protein of replication restart, a housekeeping process that facilitates the restart of stalled replication forks during normal growth (267). E. coli priA- strains suffers from severe growth retardation and are hyper-susceptible to ciprofloxacin with a MIC below 1 ng/ml (72). The above observations indicates that PriA inhibition is possibly lethal to bacterial strains with gyrA mutations conferring ciprofloxacin resistance. Thus, combinatorial treatment of a PriA inhibitor with ciprofloxacin, or cycling between the two, could be an effective mean of treating infections caused by ciprofloxacin resistant bacteria.

Targeting the commencement of DNA replication in bacteria

Duplication of the chromosome is an essential part of the bacterial cell cycle and its initiation could potentially serve as a novel antibiotic target. We have presented two novel cell based strategies for identifying DNA replication initiation inhibitors. The replication initiation process and its regulation is complex and can therefore, potentially be inhibited via multiple different targets, of which directly targeting DnaA is the most apparent one. However, the negative or positive regulation of the initiation, might also serve as potential targets. Emphasized by the fact that deletion of either datA, DARS1 or DARS2 in E. coli, results in a reduced ability to colonize the large intestine in mice (183). DnaA binding to the DnaA boxes in the oriC is a plausible target for putative inhibitors of the replication initiation. Specifically, targeting the HTH motif of DnaA domain IV could in principal block the binding of DnaA to both the high and low affinity DnaA boxes in the oriC. Interference with the assembly of the DnaAATP-OriC nucleoprotein complex is attainable in multiple ways. Binding of an inhibitor in the nucleotide-binding pocket of the AAA+ module of domain III, could block ATP binding and lock the structural conformation of DnaA in an apo-DnaA or DnaAADP-like state that is inactive in DnaA oligmerization. In addition, the cooperative binding of DnaAATP could also be inhibited by interfering with the DnaA-DnaA interactions mediated by specific residues of DnaA domain I and III. However, DnaA boxes in the oriC are not arranged similarly across bacterial species, indicating that there are many different ways of assembling the nucleoprotein complex. Thus, it is not necessarily the same residues that mediates the DnaA-DnaA interactions in different bacterial species (188). Due to their stimulatory role in the replication initiation process, factors like; IHF, DiaA, HU, and H-NS, might also serve as targets for inhibiting replication initiation. Although none of these factors are essential for replication initiation, their deletion does lead to under-initiation (163, 268-270), though it remains to be assessed if it has a lethal effect in a hostile environment like the human body.

104 Loading of the DnaB helicase by DnaA is an essential step in initiating the replication process and is therefore an obvious target. Inhibiting the binding of DnaC would likely block the loading of DnaB onto the ssDUE. As DnaB would not be locked into to the open conformation that is essential for its loading onto the ssDUE (193). In addition, hindering the interaction between DnaB and DnaA domain I should also block the loading of the DnaB helicase. As mentioned above, the processes that regulates the DnaAATP/DnaAADP ratio is also a potential target for interfering with the replication initiation. Interestingly, over-initiation seems to be more lethal than decreasing the initiation frequency (237). Thus, inhibiting the conversion of DnaAATP to DnaAADP by RIDA or DDAH, or the sequestration of the oriC by SeqA, is likely more favorable than targeting the rejuvenation of DnaAATP. Obstructing SeqA binding to the GATC in the oriC and the DnaA promoter region, would likely lead to over-initiation and increased levels of DnaAATP. However, the increase in initiation observed when seqA is deleted is not detrimental to the cell (271). DDAH is less efficient in stimulating DnaAATP hydrolysis than RIDA (170), indicating that RIDA is the favorable target. RIDA inactivation could be achieved by; i) hindering binding of ADP to Hda, ii) blocking the ATPase stimulatory interaction of the Hda Arg-finger with the DnaAATP ATPase in domain III, iii) inhibiting the stabilizing interactions between the C-terminal of Hda and DnaAATP domain I. Though the idea of inducing over-initiation is intriguing, the fact that secondary mutations arise quickly in cells that are over-initiating (240), designates that resistance would quickly develop. Furthermore, the ROS dependency of the lethal action of hyper-replication suggests, that the strategy of inducing over- initiation is not plausible at low ROS conditions i.e. anaerobsis or when free iron availability is limited. Inhibition of the regeneration of DnaAATP via DARS1 or DARS2, is likely achievable by blocking the DnaA-DnaA interactions in the DnaA box core region, or by hindering the binding of IHF or Fis to the regulatory region of DARS2. The mechanisms that lead to DnaAATP regeneration mediated by phospholipids are unknown. Consequently, inhibition of the phospholipid synthesis is the only known mean by which this process could be inhibited. Targeting the regulatory mechanisms of the replication initiation has a downside, as it currently not known how many bacterial species that uses the DnaAATP level to regulate replication initiation. Hda homologs have been identified in Caulobacter and most enterobacteria, but not in Bacillus, Staphylococcus and H. pylori (131, 272). DARS1 and DARS2 are likely conserved in proteobacteria that are closely related to E. coli, while more distantly related proteobacteria have DARS-like DnaA box clusters in other intergenic regions. Non-proteobacterial species like S. aureus, Mycobacterium tuberculosis and Bacillus subtilis have DnaA box clusters near dnaA, though with a different arrangement of the DnaA boxes than the one in E. coli (148). Consequently, compounds that target the regulation of replication initiation will likely not have a broad bacterial spectrum.

105 Despite all of the above mentioned potential targets, it is curios that not a single compound that targets the replication initiation process has been identified. This fact could indicate that the replication initiation is not an ideal target for development of novel antibiotics. Furthermore, the complexity of its regulation may enhance the occurrence of secondary compensatory mutations that counteract the inhibitory action of a given compound. Yet, it is important to keep in mind that direct efforts at identifying replication initiation inhibiters has so far been scarce. The screens presented in paper II and III have some distinct differences concerning their specificity and practical applicability. The fluorescence-based screen relies on replication by cSDR and is therefore not able to identify putative inhibitors of the DnaB helicase loading; as such compounds would kill the cells. Conversely, the screens based on over-initiating cells cannot be used to identify inhibitors that induce over-initiation. Whereas, over-initiation of the mini-chromosome in the fluorescence based screen, could potentially lead to instability and loss of the mini-chromosome and thereby expression of the green fluorescent protein (GFP). Concerning the practical applicability, the fluorescence based screen is performed in 96-well plates in a high throughput manner. In contrast, the agar plate based platform of the hda and DARS2 screens needs to undergo further optimization for use in a high throughput setup. The agar plate based platform has a significant advantage, as a concentration gradient is created when the extract or compound diffuse from the well into the agar. Thus, several drug concentrations are tested at once, in contrast to the fluorescence based screen, where only a single concentration is tested.

Why is severe over-initiation of the DNA replication lethal?

In paper II, we show that the iron chelator deferoxamine rescues the growth of over-initiating cells by promoting the processivity of replication forks. Deferoxamine has been shown to inhibit the Fenton reaction and thereby the production of ROS both in vivo and in vitro. Our findings therefore support the proposed model that the lethal action of over-initiating the chromosomal DNA replication is caused by formation of DSBs in the DNA, when replication forks encounters 8-oxo-dGTP lesions that are under repair by the GO system (235). Since overproduction of ribonucleotide reductase (RNR), restores the growth of over-initiating cells, it has been suggested that during replication over-initiation the cells are starved for dNTPs because of the increased number of replication forks (238, 239, 241). The dNTP starvation is proposedly responsible for the severe growth retardation of cells deficient in Hda (239). However, several observations contradict this model; i) the reduction in the dNTP pool of the hda mutant is not significant, ii) the inviability of an hda mutant does not resemble the inviability observed for cells starved in dNTPs, as dNTP starvation leads to obliteration of the oriC, in contrast to the observed increase in oriC copy number for hda mutants, iii) an increase in all four dNTPs is not observed

106 when RNR is overexpressed, specifically the dGTP level remains more or less unchanged, thus a hda mutant overexpressing RNR is still starved in dGTP (271). Conversely, the observed suppression of the hda phenotype by overproduction of RNR is more likely caused by a reduction in replication initiation, as hda mutants overproducing RNR has an origin concentration resembling that of a wild-type (241, 271). Hence, over-expression of RNR in an hda mutant lowers the initiation frequency, giving time for efficient repair of 8-oxo-dGTP lesions.

Conclusions

Utilizing genetic screens and differential gene-expression analysis, we have shown that reversing ciprofloxacin resistance in a high-level ciprofloxacin resistant E. coli strain is likely not possible. However, our genetic screen revealed the AcrAB-tolC efflux pump, and the SOS response proteins RecA and RecC, as plausible targets for ciprofloxacin helper drugs in E. coli strains with a MIC just above clinical breakpoint. We have constructed and verified three screens for identifying inhibitors of the initiation of chromosomal DNA replication in bacteria. One screen relies on replication inhibition of an OriC dependent mini-chromosome, leading to expression of GFP and thereby a detectable increase in fluorescence. The two other screens are based on growth rescue of cells that exerts lethal over-initiation of the DNA replication, by either harboring multiple copies of DARS2 or being deficient in Hda. As a pilot screen for inhibitors of the replication initiation process, we subjected our novel screens to a library of 400 actinomycetes extracts. Even though we did not identify any initiation inhibitors, the iron chelator deferoxamine, a known inhibitor of the Fenton reaction (245), was identified as a compound that rescues the growth of over-initiating cells. Corroborating the model that the lethality of over-initiating the chromosomal DNA replication is caused by formation of DSBs in the DNA, when replication forks encounters 8-oxo-dGTP lesions that are under repair by the GO system (235). In addition, we also showed that the growth rescue of over-initiating cells exerted by the suggested gyrase inhibitor (±)-6-Chloro-PB hydrobromide (S143) is, at least in part, due to its ability to chelate iron.

Future perspectives

Antibiotic resistance will be a remaining threat to global health care. However, by heavily investing in antibiotic drug development and regulating the use of antibiotics globally. It may be possible to halt or slow the current negative development. Based on our findings that the AcrAB-TolC efflux pump and the

107 SOS response genes RecA and RecC, might serve as ciprofloxacin helper drug targets, in treating low- level resistant strains. In addition to the fact that such helper drugs have the potential to potentiate other known antibiotics and decrease the evolution of antibiotic resistance, it would be interesting to setup screens for identifying inhibitors of these three targets. Moreover, it would be attractive to assess if PriA deletion is truly lethal for bacterial strains carrying gyrA mutations that confer ciprofloxacin resistance. As we now have two distinct strategies for identifying replication inhibitors, the next logical step is to obtain several chemical or natural extract libraries that could be subjected to the screens. The 400 extracts that were screened in paper II and III, are part of a large library of more than 4000 microbial extracts, owned by our collaborator Naicons srl., which has proven to be a source of novel antibiotic compounds (273, 274). Hopefully, future funding will give us the opportunity to screen the remainder of this vast library of bioactive natural extracts.

108 Bibliography

1. Fleming A. On the Antibacterial Action of Cultures of a Penicillium, with Special Reference to their Use in the Isolation of B. influenzæ. British journal of experimental pathology. 1929;10(3):226-36. 2. Ehrlich P, and Hata, S. Die Experimentelle Chemotherapie der Spirilosen1910. 3. Aminov R. A Brief History of the Antibiotic Era: Lessons Learned and Challenges for the Future. Frontiers in Microbiology. 2010;1(134). 4. Achari A, Somers D, Champness JN, Bryant PK, Rosemond J, Stammers DK. Crystal structure of the anti-bacterial drug target dihydropteroate synthase. Nature structural biology. 1997;4:490. 5. Gould K. Antibiotics: from prehistory to the present day. Journal of Antimicrobial Chemotherapy. 2016;71(3):572-5. 6. Clatworthy AE, Pierson E, Hung DT. Targeting virulence: a new paradigm for antimicrobial therapy. Nature Chemical Biology. 2007;3:541. 7. Fleming A. Penicillin, Nobel lecture,1945. 8. Leach KL, Brickner SJ, Noe MC, Miller PF. Linezolid, the first oxazolidinone antibacterial agent. Annals of the New York Academy of Sciences. 2011;1222:49-54. 9. Long KS, Vester B. Resistance to Linezolid Caused by Modifications at Its Binding Site on the Ribosome. Antimicrobial agents and chemotherapy. 2012;56(2):603-12. 10. Eliopoulos GM, Meka VG, Gold HS. Antimicrobial Resistance to Linezolid. Clinical Infectious Diseases. 2004;39(7):1010-5. 11. The PEW Charitable Trusts. Antibiotics currently in global clinical development 2017 [Available from: http://www.pewtrusts.org/~/media/assets/2017/12/antibiotics_currently_in_clinical_development_09_20 17.pdf?la=en. 12. Simpkin VL, Renwick MJ, Kelly R, Mossialos E. Incentivising innovation in antibiotic drug discovery and development: progress, challenges and next steps. The Journal of antibiotics. 2017;70:1087. 13. Review on Antimicrobial Resistance. Resistance: Tackling a Crisis for the Health and Wealth of Nations 2014 [Available from: https://amr-review.org/sites/default/files/AMR%20Review%20Paper%20- %20Tackling%20a%20crisis%20for%20the%20health%20and%20wealth%20of%20nations_1.pdf. 14. Mitscher LA. Bacterial topoisomerase inhibitors: quinolone and pyridone antibacterial agents. Chemical reviews. 2005;105(2):559-92. 15. Linder JA, Huang ES, Steinman MA, Gonzales R, Stafford RS. Fluoroquinolone prescribing in the United States: 1995 to 2002. The American Journal of Medicine.118(3):259-68. 16. Emmerson AM, Jones AM. The quinolones: decades of development and use. The Journal of antimicrobial chemotherapy. 2003;51 Suppl 1:13-20. 17. Werner NL, Hecker MT, Sethi AK, Donskey CJ. Unnecessary use of fluoroquinolone antibiotics in hospitalized patients. BMC Infectious Diseases. 2011;11:187-. 18. Dalhoff A. Global Fluoroquinolone Resistance Epidemiology and Implictions for Clinical Use. Interdisciplinary Perspectives on Infectious Diseases. 2012;2012:37. 19. Bisacchi GS. Origins of the Quinolone Class of Antibacterials: An Expanded “Discovery Story”. Journal of Medicinal Chemistry. 2015;58(12):4874-82. 20. Lesher GY, Froelich EJ, Gruett MD, Bailey JH, Brundage RP. 1,8-Naphthyridine Derivatives. A New Class of Chemotherapeutic Agents. Journal of Medicinal and Pharmaceutical Chemistry. 1962;5(5):1063-5. 21. Aldred KJ, Kerns RJ, Osheroff N. Mechanism of quinolone action and resistance. Biochemistry. 2014;53(10):1565-74. 22. Owens RC, Ambrose, Paul G. CLINICAL USE OF THE FLUOROQUINOLONES. Medical Clinics of North America. 2000;84(6):1447-69.

109 23. Goss WA, Deitz WH, Cook TM. Mechanism of Action of Nalidixic Acid on Escherichia coli II. Inhibition of Deoxyribonucleic Acid Synthesis. Journal of bacteriology. 1965;89(4):1068-74. 24. Hane MW, Wood TH. Escherichia coli K-12 mutants resistant to nalidixic acid: genetic mapping and dominance studies. Journal of bacteriology. 1969;99(1):238-41. 25. Gellert M, Mizuuchi K, O'Dea MH, Itoh T, Tomizawa JI. Nalidixic acid resistance: a second genetic character involved in DNA gyrase activity. Proceedings of the National Academy of Sciences of the United States of America. 1977;74(11):4772-6. 26. Sugino A, Peebles CL, Kreuzer KN, Cozzarelli NR. Mechanism of action of nalidixic acid: purification of Escherichia coli nalA gene product and its relationship to DNA gyrase and a novel nicking- closing enzyme. Proceedings of the National Academy of Sciences of the United States of America. 1977;74(11):4767-71. 27. Kato J, Nishimura Y, Imamura R, Niki H, Hiraga S, Suzuki H. New topoisomerase essential for chromosome segregation in E. coli. Cell. 1990;63(2):393-404. 28. Hooper DC. Mode of action of fluoroquinolones. Drugs. 1999;58 Suppl 2:6-10. 29. Aldred KJ, McPherson SA, Turnbough CL, Jr., Kerns RJ, Osheroff N. Topoisomerase IV- quinolone interactions are mediated through a water-metal ion bridge: mechanistic basis of quinolone resistance. Nucleic Acids Res. 2013;41(8):4628-39. 30. Aldred KJ, McPherson SA, Wang P, Kerns RJ, Graves DE, Turnbough CL, Jr., et al. Drug interactions with Bacillus anthracis topoisomerase IV: biochemical basis for quinolone action and resistance. Biochemistry. 2012;51(1):370-81. 31. Wohlkonig A, Chan PF, Fosberry AP, Homes P, Huang J, Kranz M, et al. Structural basis of quinolone inhibition of type IIA topoisomerases and target-mediated resistance. Nat Struct Mol Biol. 2010;17(99):1152-3. 32. Drlica K, Zhao X. DNA gyrase, topoisomerase IV, and the 4-quinolones. Microbiology and molecular biology reviews : MMBR. 1997;61(3):377-92. 33. Khodursky AB, Zechiedrich EL, Cozzarelli NR. Topoisomerase IV is a target of quinolones in Escherichia coli. Proceedings of the National Academy of Sciences of the United States of America. 1995;92(25):11801-5. 34. Ferrero L, Cameron B, Crouzet J. Analysis of gyrA and grlA mutations in stepwise-selected ciprofloxacin-resistant mutants of Staphylococcus aureus. Antimicrobial agents and chemotherapy. 1995;39(7):1554-8. 35. Gillespie SH, Voelker LL, Ambler JE, Traini C, Dickens A. Fluoroquinolone resistance in Streptococcus pneumoniae: evidence that gyrA mutations arise at a lower rate and that mutation in gyrA or parC predisposes to further mutation. Microbial (Larchmont, NY). 2003;9(1):17-24. 36. Fournier B, Zhao X, Lu T, Drlica K, Hooper DC. Selective targeting of topoisomerase IV and DNA gyrase in Staphylococcus aureus: different patterns of quinolone-induced inhibition of DNA synthesis. Antimicrobial agents and chemotherapy. 2000;44(8):2160-5. 37. Kreuzer KN, Cozzarelli NR. Escherichia coli mutants thermosensitive for deoxyribonucleic acid gyrase subunit A: effects on deoxyribonucleic acid replication, transcription, and bacteriophage growth. Journal of bacteriology. 1979;140(2):424-35. 38. D'Arpa P, Beardmore C, Liu LF. Involvement of nucleic acid synthesis in cell killing mechanisms of topoisomerase poisons. Cancer research. 1990;50(21):6919-24. 39. Malik M, Zhao X, Drlica K. Lethal fragmentation of bacterial chromosomes mediated by DNA gyrase and quinolones. Molecular microbiology. 2006;61(3):810-25. 40. Hiasa H, Yousef DO, Marians KJ. DNA strand cleavage is required for replication fork arrest by a frozen topoisomerase-quinolone-DNA ternary complex. The Journal of biological chemistry. 1996;271(42):26424-9. 41. Shea ME, Hiasa H. Interactions between DNA helicases and frozen topoisomerase IV- quinolone-DNA ternary complexes. The Journal of biological chemistry. 1999;274(32):22747-54.

110 42. Shea ME, Hiasa H. Distinct effects of the UvrD helicase on topoisomerase-quinolone-DNA ternary complexes. The Journal of biological chemistry. 2000;275(19):14649-58. 43. Shea ME, Hiasa H. The RuvAB branch migration complex can displace topoisomerase IV.quinolone.DNA ternary complexes. The Journal of biological chemistry. 2003;278(48):48485-90. 44. Pohlhaus JR, Kreuzer KN. Norfloxacin-induced DNA gyrase cleavage complexes block Escherichia coli replication forks, causing double-stranded breaks in vivo. Molecular microbiology. 2005;56(6):1416-29. 45. Zhao X, Malik M, Chan N, Drlica-Wagner A, Wang JY, Li X, et al. Lethal action of quinolones against a temperature-sensitive dnaB replication mutant of Escherichia coli. Antimicrobial agents and chemotherapy. 2006;50(1):362-4. 46. Deitz WH, Cook TM, Goss WA. Mechanism of action of nalidixic acid on Escherichia coli. 3. Conditions required for lethality. Journal of bacteriology. 1966;91(2):768-73. 47. Drlica K, Hiasa H, Kerns R, Malik M, Mustaev A, Zhao X. Quinolones: action and resistance updated. Current topics in medicinal chemistry. 2009;9(11):981-98. 48. Imlay JA, Fridovich I. Superoxide Production by Respiring Membranes of Escherichia Coli. Free Radical Research Communications. 1991;12(1):59-66. 49. Imlay JA. Diagnosing oxidative stress in bacteria: not as easy as you might think. Current opinion in microbiology. 2015;24:124-31. 50. Belenky P, Ye JD, Porter CB, Cohen NR, Lobritz MA, Ferrante T, et al. Bactericidal Antibiotics Induce Toxic Metabolic Perturbations that Lead to Cellular Damage. Cell Rep. 2015;13(5):968-80. 51. Becerra MC, Albesa I. Oxidative stress induced by ciprofloxacin in Staphylococcus aureus. Biochemical and Biophysical Research Communications. 2002;297(4):1003-7. 52. Albesa I, Becerra MC, Battan PC, Paez PL. Oxidative stress involved in the antibacterial action of different antibiotics. Biochem Biophys Res Commun. 2004;317(2):605-9. 53. Becerra MC, Páez PL, Laróvere LE, Albesa I. Lipids and DNA oxidation in Staphylococcus aureus as a consequence of oxidative stress generated by ciprofloxacin. Molecular and Cellular Biochemistry. 2006;285(1):29-34. 54. Goswami M, Mangoli SH, Jawali N. Involvement of Reactive Oxygen Species in the Action of Ciprofloxacin against Escherichia coli. Antimicrobial agents and chemotherapy. 2006;50(3):949-54. 55. Kohanski MA, Dwyer DJ, Hayete B, Lawrence CA, Collins JJ. A Common Mechanism of Cellular Death Induced by Bactericidal Antibiotics. Cell. 2007;130(5):797-810. 56. Dwyer DJ, Kohanski MA, Hayete B, Collins JJ. Gyrase inhibitors induce an oxidative damage cellular death pathway in Escherichia coli. Molecular systems biology. 2007;3:91. 57. Foti JJ, Devadoss B, Winkler JA, Collins JJ, Walker GC. Oxidation of the Guanine Nucleotide Pool Underlies Cell Death by Bactericidal Antibiotics. Science. 2012;336(6079):315-9. 58. Kottur J, Nair DT. Reactive Oxygen Species Play an Important Role in the Bactericidal Activity of Quinolone Antibiotics. Angewandte Chemie (International ed in English). 2016;55(7):2397-400. 59. Hassan HM, Moody CS. Induction of the manganese-containing superoxide dismutase in Escherichia coli by nalidixic acid and by iron chelators. FEMS Microbiology Letters. 1984;25(2-3):233-6. 60. Greenberg JT, Monach P, Chou JH, Josephy PD, Demple B. Positive control of a global antioxidant defense regulon activated by superoxide-generating agents in Escherichia coli. Proceedings of the National Academy of Sciences. 1990;87(16):6181-5. 61. Tsaneva IR, Weiss, B. soxR, a locus governing a superoxide response regulon in Escherichia coli K-12. Journal of bacteriology. 1990;172(8):4197-205. 62. Seo Sang W, Kim D, Szubin R, Palsson Bernhard O. Genome-wide Reconstruction of OxyR and SoxRS Transcriptional Regulatory Networks under Oxidative Stress in Escherichia coli K-12 MG1655. Cell Reports.12(8):1289-99. 63. Nunoshiba T, Hidalgo E, Amábile Cuevas CF, Demple B. Two-stage control of an oxidative stress regulon: the Escherichia coli SoxR protein triggers redox-inducible expression of the soxS regulatory gene. Journal of bacteriology. 1992;174(19):6054-60.

111 64. Chou JH, Greenberg JT, Demple B. Posttranscriptional repression of Escherichia coli OmpF protein in response to redox stress: positive control of the micF antisense RNA by the soxRS locus. Journal of bacteriology. 1993;175(4):1026-31. 65. Li X-Z, Nikaido H. Efflux-Mediated Drug Resistance in Bacteria. Drugs. 2004;64(2):159-204. 66. Wang X, Zhao X, Malik M, Drlica K. Contribution of reactive oxygen species to pathways of quinolone-mediated bacterial cell death. The Journal of antimicrobial chemotherapy. 2010;65(3):520-4. 67. Malik M, Hussain S, Drlica K. Effect of anaerobic growth on quinolone lethality with Escherichia coli. Antimicrobial agents and chemotherapy. 2007;51(1):28-34. 68. Liu Y, Imlay JA. Cell Death from Antibiotics Without the Involvement of Reactive Oxygen Species. Science. 2013;339(6124):1210-3. 69. Keren I, Wu Y, Inocencio J, Mulcahy LR, Lewis K. Killing by Bactericidal Antibiotics Does Not Depend on Reactive Oxygen Species. Science. 2013;339(6124):1213. 70. Radman M. SOS repair hypothesis: phenomenology of an inducible DNA repair which is accompanied by mutagenesis. Basic life sciences. 1975;5a:355-67. 71. Lewin CS, Howard BM, Ratcliffe NT, Smith JT. 4-quinolones and the SOS response. Journal of medical microbiology. 1989;29(2):139-44. 72. Cirz RT, Chin JK, Andes DR, de Crécy-Lagard V, Craig WA, Romesberg FE. Inhibition of Mutation and Combating the Evolution of Antibiotic Resistance. PLoS Biol. 2005;3(6):e176. 73. Little JW, Edmiston SH, Pacelli LZ, Mount DW. Cleavage of the Escherichia coli lexA protein by the recA protease. Proceedings of the National Academy of Sciences of the United States of America. 1980;77(6):3225-9. 74. Schlacher K, Goodman MF. Lessons from 50 years of SOS DNA-damage-induced mutagenesis. Nat Rev Mol Cell Biol. 2007;8(7):587-94. 75. Little JW. Mechanism of specific LexA cleavage: autodigestion and the role of RecA coprotease. Biochimie. 1991;73(4):411-21. 76. Hiom K. DNA Repair: Common Approaches to Fixing Double-Strand Breaks. Current Biology. 2009;19(13):R523-R5. 77. Singleton MR, Dillingham MS, Gaudier M, Kowalczykowski SC, Wigley DB. Crystal structure of RecBCD enzyme reveals a machine for processing DNA breaks. Nature. 2004;432(7014):187-93. 78. Wyman C, Ristic D, Kanaar R. Homologous recombination-mediated double-strand break repair. DNA repair. 2004;3(8–9):827-33. 79. Morgan-Linnell SK, Becnel Boyd L, Steffen D, Zechiedrich L. Mechanisms Accounting for Fluoroquinolone Resistance in Escherichia coli Clinical Isolates. Antimicrobial agents and chemotherapy. 2009;53(1):235-41. 80. Lindgren PK, Marcusson LL, Sandvang D, Frimodt-Møller N, Hughes D. Biological Cost of Single and Multiple Norfloxacin Resistance Mutations in Escherichia coli Implicated in Urinary Tract Infections. Antimicrobial agents and chemotherapy. 2005;49(6):2343-51. 81. Marcusson LL, Frimodt-Møller N, Hughes D. Interplay in the Selection of Fluoroquinolone Resistance and Bacterial Fitness. PLoS Pathog. 2009;5(8):e1000541. 82. CLSI. Performance Standards for Antimicrobial Susceptibility Testing. 27th ed. CLSI supplement M100. Wayne, PA: Clinical and Laboratory Standards Institute; 2017. 83. Andersson DI. The biological cost of mutational antibiotic resistance: any practical conclusions? Current Opinion in Microbiology. 2006;9(5):461-5. 84. Andersson DI, Hughes D. Effects of Antibiotic Resistance on Bacterial Fitness, Virulence, and Transmission. Evolutionary Biology of Bacterial and Fungal Pathogens: American Society of Microbiology; 2008. 85. Komp Lindgren P, Karlsson Å, Hughes D. Mutation Rate and Evolution of Fluoroquinolone Resistance in Escherichia coli Isolates from Patients with Urinary Tract Infections. Antimicrobial agents and chemotherapy. 2003;47(10):3222-32.

112 86. Fuchs RP, Fujii S, Wagner J. Properties and functions of Escherichia coli: Pol IV and Pol V. Advances in protein chemistry. 2004;69:229-64. 87. Grkovic S, Brown MH, Skurray RA. Regulation of Bacterial Drug Export Systems. Microbiology and Molecular Biology Reviews. 2002;66(4):671-701. 88. Ma D, Cook DN, Alberti M, Pon NG, Nikaido H, Hearst JE. Genes acrA and acrB encode a stress-induced efflux system of Escherichia coli. Molecular microbiology. 1995;16(1):45-55. 89. Delihas N, Forst S. MicF: an antisense RNA gene involved in response of Escherichia coli to global stress factors11Edited by D. Draper. Journal of molecular biology. 2001;313(1):1-12. 90. Cama J, Bajaj H, Pagliara S, Maier T, Braun Y, Winterhalter M, et al. Quantification of Fluoroquinolone Uptake through the Outer Membrane Channel OmpF of Escherichia coli. J Am Chem Soc. 2015;137(43):13836-43. 91. Chapman JS, Georgopapadakou NH. Routes of quinolone permeation in Escherichia coli. Antimicrobial agents and chemotherapy. 1988;32(4):438-42. 92. Poole K, Tetro K, Zhao Q, Neshat S, Heinrichs DE, Bianco N. Expression of the multidrug resistance operon mexA-mexB-oprM in Pseudomonas aeruginosa: mexR encodes a regulator of operon expression. Antimicrobial agents and chemotherapy. 1996;40(9):2021-8. 93. Ziha-Zarifi I, Llanes C, Köhler T, Pechere J-C, Plesiat P. In Vivo Emergence of Multidrug- Resistant Mutants of Pseudomonas aeruginosa Overexpressing the Active Efflux System MexA-MexB-OprM. Antimicrobial agents and chemotherapy. 1999;43(2):287-91. 94. Llanes C, Hocquet D, Vogne C, Benali-Baitich D, Neuwirth C, Plésiat P. Clinical Strains of Pseudomonas aeruginosa Overproducing MexAB-OprM and MexXY Efflux Pumps Simultaneously. Antimicrobial agents and chemotherapy. 2004;48(5):1797-802. 95. Hooper DC, Jacoby GA. Mechanisms of drug resistance: quinolone resistance. Annals of the New York Academy of Sciences. 2015;1354(1):12-31. 96. Martinez-Martinez L, Pascual A, Jacoby GA. Quinolone resistance from a transferable plasmid. Lancet (London, England). 1998;351(9105):797-9. 97. Jacoby GA. Mechanisms of Resistance to Quinolones. Clinical Infectious Diseases. 2005;41(Supplement_2):S120-S6. 98. Robicsek A, Jacoby GA, Hooper DC. The worldwide emergence of plasmid-mediated quinolone resistance. The Lancet Infectious diseases. 2006;6(10):629-40. 99. Munshi MH, Sack DA, Haider K, Ahmed ZU, Rahaman MM, Morshed MG. Plasmid-mediated resistance to nalidixic acid in Shigella dysenteriae type 1. Lancet (London, England). 1987;2(8556):419-21. 100. Jacoby GA, Strahilevitz J, Hooper DC. Plasmid-mediated quinolone resistance. Microbiology spectrum. 2014;2(2):10.1128/microbiolspec.PLAS-0006-2013. 101. Vetting MW, Hegde SS, Fajardo JE, Fiser A, Roderick SL, Takiff HE, et al. Pentapeptide repeat proteins. Biochemistry. 2006;45(1):1-10. 102. Tran JH, Jacoby GA. Mechanism of plasmid-mediated quinolone resistance. Proceedings of the National Academy of Sciences. 2002;99(8):5638-42. 103. Hata M, Suzuki M, Matsumoto M, Takahashi M, Sato K, Ibe S, et al. Cloning of a Novel Gene for Quinolone Resistance from a Transferable Plasmid in Shigella flexneri 2b. Antimicrobial agents and chemotherapy. 2005;49(2):801-3. 104. Jacoby GA, Walsh KE, Mills DM, Walker VJ, Oh H, Robicsek A, et al. qnrB, another plasmid- mediated gene for quinolone resistance. Antimicrobial agents and chemotherapy. 2006;50(4):1178-82. 105. Wang M, Guo Q, Xu X, Wang X, Ye X, Wu S, et al. New plasmid-mediated quinolone resistance gene, qnrC, found in a clinical isolate of Proteus mirabilis. Antimicrobial agents and chemotherapy. 2009;53(5):1892-7. 106. Cavaco LM, Hasman H, Xia S, Aarestrup FM. qnrD, a Novel Gene Conferring Transferable Quinolone Resistance in Salmonella enterica Serovar Kentucky and Bovismorbificans Strains of Human Origin. Antimicrobial agents and chemotherapy. 2009;53(2):603-8.

113 107. Albornoz E, Tijet N, De Belder D, Gomez S, Martino F, Corso A, et al. qnrE1, a Member of a New Family of Plasmid-Located Quinolone Resistance Genes, Originated from the Chromosome of Enterobacter Species. Antimicrobial agents and chemotherapy. 2017;61(5). 108. Pons MJ, Gomes C, Ruiz J. QnrVC, a new transferable Qnr-like family. Enfermedades Infecciosas y Microbiología Clínica. 2013;31(3):191-2. 109. Xiong X, Bromley EH, Oelschlaeger P, Woolfson DN, Spencer J. Structural insights into quinolone antibiotic resistance mediated by pentapeptide repeat proteins: conserved surface loops direct the activity of a Qnr protein from a gram-negative bacterium. Nucleic Acids Res. 2011;39(9):3917-27. 110. Vetting MW, Hegde SS, Wang M, Jacoby GA, Hooper DC, Blanchard JS. Structure of QnrB1, a plasmid-mediated fluoroquinolone resistance factor. The Journal of biological chemistry. 2011;286(28):25265-73. 111. Robicsek A, Strahilevitz J, Jacoby GA, Macielag M, Abbanat D, Park CH, et al. Fluoroquinolone-modifying enzyme: a new adaptation of a common aminoglycoside acetyltransferase. Nature medicine. 2006;12(1):83-8. 112. Rodríguez-Martínez JM, Machuca J, Cano ME, Calvo J, Martínez-Martínez L, Pascual A. Plasmid-mediated quinolone resistance: Two decades on. Drug Resistance Updates. 2016;29:13-29. 113. Yamane K, Wachino J-i, Suzuki S, Kimura K, Shibata N, Kato H, et al. New Plasmid-Mediated Fluoroquinolone Efflux Pump, QepA, Found in an Escherichia coli Clinical Isolate. Antimicrobial agents and chemotherapy. 2007;51(9):3354-60. 114. Reading C, Cole M. Clavulanic Acid: a Beta-Lactamase-Inhibiting Beta-Lactam from Streptomyces clavuligerus. Antimicrobial agents and chemotherapy. 1977;11(5):852-7. 115. White AR, Kaye C, Poupard J, Pypstra R, Woodnutt G, Wynne B. Augmentin (amoxicillin/clavulanate) in the treatment of community-acquired respiratory tract infection: a review of the continuing development of an innovative antimicrobial agent. The Journal of antimicrobial chemotherapy. 2004;53 Suppl 1:i3-20. 116. Worthington RJ, Melander C. Combination Approaches to Combat Multi-Drug Resistant Bacteria. Trends in biotechnology. 2013;31(3):177-84. 117. Van Bambeke F, Pages JM, Lee VJ. Inhibitors of bacterial efflux pumps as adjuvants in antibiotic treatments and diagnostic tools for detection of resistance by efflux. Recent patents on anti- infective drug discovery. 2006;1(2):157-75. 118. Sabatini S, Gosetto F, Serritella S, Manfroni G, Tabarrini O, Iraci N, et al. Pyrazolo[4,3- c][1,2]benzothiazines 5,5-dioxide: a promising new class of Staphylococcus aureus NorA efflux pump inhibitors. J Med Chem. 2012;55(7):3568-72. 119. Nautiyal A, Patil KN, Muniyappa K. Suramin is a potent and selective inhibitor of Mycobacterium tuberculosis RecA protein and the SOS response: RecA as a potential target for antibacterial drug discovery. The Journal of antimicrobial chemotherapy. 2014;69(7):1834-43. 120. Bellio P, Brisdelli F, Perilli M, Sabatini A, Bottoni C, Segatore B, et al. Curcumin inhibits the SOS response induced by levofloxacin in Escherichia coli. Phytomedicine. 2014;21(4):430-4. 121. Sexton JZ, Wigle TJ, He Q, Hughes MA, Smith GR, Singleton SF, et al. Novel Inhibitors of E. coli RecA ATPase Activity. Current Chemical Genomics. 2010;4:34-42. 122. Lee AM, Ross CT, Zeng B-B, Singleton SF. A Molecular Target for Suppression of the Evolution of Antibiotic Resistance: Inhibition of the Escherichia coli RecA Protein by N6-(1-Naphthyl)-ADP. Journal of Medicinal Chemistry. 2005;48(17):5408-11. 123. Lee AM, Singleton SF. Inhibition of the Escherichia coli RecA protein: zinc(II), copper(II) and mercury(II) trap RecA as inactive aggregates. Journal of inorganic biochemistry. 2004;98(11):1981-6. 124. Cline DJ, Holt SL, Singleton SF. Inhibition of Escherichia coli RecA by rationally redesigned N- terminal helix. Organic & biomolecular chemistry. 2007;5(10):1525-8. 125. Alam MK, Alhhazmi A, DeCoteau JF, Luo Y, Geyer CR. RecA Inhibitors Potentiate Antibiotic Activity and Block Evolution of Antibiotic Resistance. Cell chemical biology. 2016;23(3):381-91.

114 126. Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, et al. Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Molecular systems biology. 2006;2(1). 127. Liu A, Tran L, Becket E, Lee K, Chinn L, Park E, et al. Antibiotic sensitivity profiles determined with an Escherichia coli gene knockout collection: generating an antibiotic bar code. Antimicrobial agents and chemotherapy. 2010;54(4):1393-403. 128. Tamae C, Liu A, Kim K, Sitz D, Hong J, Becket E, et al. Determination of Antibiotic Hypersensitivity among 4,000 Single-Gene-Knockout Mutants of Escherichia coli. Journal of bacteriology. 2008;190(17):5981-8. 129. Tran T, Ran Q, Ostrer L, Khodursky A. De Novo Characterization of Genes That Contribute to High-Level Ciprofloxacin Resistance in Escherichia coli. Antimicrobial agents and chemotherapy. 2016;60(10):6353-5. 130. Recacha E, Machuca J, Díaz de Alba P, Ramos-Güelfo M, Docobo-Pérez F, Rodriguez-Beltrán J, et al. Quinolone Resistance Reversion by Targeting the SOS Response. mBio. 2017;8(5). 131. Grimwade JE, Leonard AC. Targeting the Bacterial Orisome in the Search for New Antibiotics. Frontiers in Microbiology. 2017;8:2352. 132. van Eijk E, Wittekoek B, Kuijper EJ, Smits WK. DNA replication proteins as potential targets for antimicrobials in drug-resistant bacterial pathogens. Journal of Antimicrobial Chemotherapy. 2017;72(5):1275-84. 133. Schaechter M, Maaloe O, Kjeldgaard NO. Dependency on medium and temperature of cell size and chemical composition during balanced grown of Salmonella typhimurium. Journal of general microbiology. 1958;19(3):592-606. 134. Cooper S, Helmstetter CE. Chromosome replication and the division cycle of Escherichia coli B/r. Journal of molecular biology. 1968;31(3):519-40. 135. Helmstetter C, Cooper S, Pierucci O, Revelas E. On the bacterial life sequence. Cold Spring Harbor symposia on quantitative biology. 1968;33:809-22. 136. Masters M, Broda P. Evidence for the bidirectional replications of the Escherichia coli chromosome. Nature: New biology. 1971;232(31):137-40. 137. Fossum S, Crooke E, Skarstad K. Organization of sister origins and replisomes during multifork DNA replication in <em>Escherichia coli</em>. The EMBO journal. 2007;26(21):4514. 138. Skarstad K, Boye E, Steen HB. Timing of initiation of chromosome replication in individual Escherichia coli cells. The EMBO journal. 1986;5(7):1711-7. 139. Katayama T, Kasho K, Kawakami H. The DnaA Cycle in Escherichia coli: Activation, Function and Inactivation of the Initiator Protein. Frontiers in Microbiology. 2017;8(2496). 140. Riber L, Frimodt-Møller J, Charbon G, Løbner-Olesen A. Multiple DNA Binding Proteins Contribute to Timing of Chromosome Replication in E. coli. Frontiers in Molecular Biosciences. 2016;3:29. 141. Davey MJ, Fang L, McInerney P, Georgescu RE, O’Donnell M. The DnaC helicase loader is a dual ATP/ADP switch protein. The EMBO journal. 2002;21(12):3148-59. 142. Fang L, Davey MJ, O'Donnell M. Replisome Assembly at oriC, the Replication Origin of E. coli, Reveals an Explanation for Initiation Sites outside an Origin. Molecular Cell. 1999;4(4):541-53. 143. Robinson A, van Oijen AM. Bacterial replication, transcription and translation: mechanistic insights from single-molecule biochemical studies. Nat Rev Micro. 2013;11(5):303-15. 144. Katayama T, Kubota T, Kurokawa K, Crooke E, Sekimizu K. The initiator function of DnaA protein is negatively regulated by the sliding clamp of the E. coli chromosomal replicase. Cell. 1998;94(1):61-71. 145. Kitagawa R, Ozaki T, Moriya S, Ogawa T. Negative control of replication initiation by a novel chromosomal locus exhibiting exceptional affinity for Escherichia coli DnaA protein. Genes & development. 1998;12(19):3032-43. 146. Lu M, Campbell JL, Boye E, Kleckner N. SeqA: a negative modulator of replication initiation in E. coli. Cell. 1994;77(3):413-26.

115 147. von Freiesleben U, Rasmussen KV, Schaechter M. SeqA limits DnaA activity in replication from oriC in Escherichia coli. Molecular microbiology. 1994;14(4):763-72. 148. Fujimitsu K, Senriuchi T, Katayama T. Specific genomic sequences of E. coli promote replicational initiation by directly reactivating ADP-DnaA. Genes & development. 2009;23(10):1221-33. 149. Bramhill D, Kornberg A. A model for initiation at origins of DNA replication. Cell. 1988;54(7):915-8. 150. Schaper S, Messer W. Interaction of the initiator protein DnaA of Escherichia coli with its DNA target. The Journal of biological chemistry. 1995;270(29):17622-6. 151. McGarry KC, Ryan VT, Grimwade JE, Leonard AC. Two discriminatory binding sites in the Escherichia coli replication origin are required for DNA strand opening by initiator DnaA-ATP. Proceedings of the National Academy of Sciences of the United States of America. 2004;101(9):2811-6. 152. Rozgaja TA, Grimwade JE, Iqbal M, Czerwonka C, Vora M, Leonard AC. Two oppositely- oriented arrays of low affinity recognition sites in oriC guide progressive binding of DnaA during E. coli pre- RC assembly. Molecular microbiology. 2011;82(2):475-88. 153. Sakiyama Y, Kasho K, Noguchi Y, Kawakami H, Katayama T. Regulatory dynamics in the ternary DnaA complex for initiation of chromosomal replication in Escherichia coli. Nucleic Acids Research. 2017;45(21):12354-73. 154. Samitt CE, Hansen FG, Miller JF, Schaechter M. In vivo studies of DnaA binding to the origin of replication of Escherichia coli. The EMBO journal. 1989;8(3):989-93. 155. Cassler MR, Grimwade JE, Leonard AC. Cell cycle-specific changes in nucleoprotein complexes at a chromosomal replication origin. The EMBO journal. 1995;14(23):5833-41. 156. Fujita MQ, Yoshikawa H, Ogasawara N. Structure of the dnaA region of Micrococcus luteus: conservation and variations among eubacteria. Gene. 1990;93(1):73-8. 157. Messer W, Blaesing F, Majka J, Nardmann J, Schaper S, Schmidt A, et al. Functional domains of DnaA proteins. Biochimie. 1999;81(8):819-25. 158. Nozaki S, Ogawa T. Determination of the minimum domain II size of Escherichia coli DnaA protein essential for cell viability. Microbiology (Reading, England). 2008;154(Pt 11):3379-84. 159. Felczak MM, Simmons LA, Kaguni JM. An essential tryptophan of Escherichia coli DnaA protein functions in oligomerization at the E. coli replication origin. The Journal of biological chemistry. 2005;280(26):24627-33. 160. Simmons LA, Felczak M, Kaguni JM. DnaA Protein of Escherichia coli: oligomerization at the E. coli chromosomal origin is required for initiation and involves specific N-terminal amino acids. Molecular microbiology. 2003;49(3):849-58. 161. Abe Y, Jo T, Matsuda Y, Matsunaga C, Katayama T, Ueda T. Structure and function of DnaA N- terminal domains: specific sites and mechanisms in inter-DnaA interaction and in DnaB helicase loading on oriC. The Journal of biological chemistry. 2007;282(24):17816-27. 162. Keyamura K, Abe Y, Higashi M, Ueda T, Katayama T. DiaA dynamics are coupled with changes in initial origin complexes leading to helicase loading. The Journal of biological chemistry. 2009;284(37):25038-50. 163. Ishida T, Akimitsu N, Kashioka T, Hatano M, Kubota T, Ogata Y, et al. DiaA, a Novel DnaA- binding Protein, Ensures the Timely Initiation of Escherichia coli Chromosome Replication. Journal of Biological Chemistry. 2004;279(44):45546-55. 164. Su'etsugu M, Harada Y, Keyamura K, Matsunaga C, Kasho K, Abe Y, et al. The DnaA N- terminal domain interacts with Hda to facilitate replicase clamp-mediated inactivation of DnaA. Environmental microbiology. 2013;15(12):3183-95. 165. Mott ML, Berger JM. DNA replication initiation: mechanisms and regulation in bacteria. Nature reviews Microbiology. 2007;5(5):343-54. 166. Zhang Q, Zhou A, Li S, Ni J, Tao J, Lu J, et al. Reversible lysine acetylation is involved in DNA replication initiation by regulating activities of initiator DnaA in Escherichia coli. Scientific reports. 2016;6:30837.

116 167. Kawakami H, Ozaki S, Suzuki S, Nakamura K, Senriuchi T, Su'etsugu M, et al. The exceptionally tight affinity of DnaA for ATP/ADP requires a unique aspartic acid residue in the AAA+ sensor 1 motif. Molecular microbiology. 2006;62(5):1310-24. 168. Nishida S, Fujimitsu K, Sekimizu K, Ohmura T, Ueda T, Katayama T. A Nucleotide Switch in the Escherichia coli DnaA Protein Initiates Chromosomal Replication: EVIDENCE FROM A MUTANT DnaA PROTEIN DEFECTIVE IN REGULATORY ATP HYDROLYSIS IN VITRO AND IN VIVO. Journal of Biological Chemistry. 2002;277(17):14986-95. 169. Su'Etsugu M, Kawakami H, Kurokawa K, Kubota T, Takata M, Katayama T. DNA replication- coupled inactivation of DnaA protein in vitro: a role for DnaA arginine-334 of the AAA+ Box VIII motif in ATP hydrolysis. Molecular microbiology. 2001;40(2):376-86. 170. Kasho K, Katayama T. DnaA binding locus datA promotes DnaA-ATP hydrolysis to enable cell cycle-coordinated replication initiation. Proceedings of the National Academy of Sciences of the United States of America. 2013;110(3):936-41. 171. Kawakami H, Keyamura K, Katayama T. Formation of an ATP-DnaA-specific initiation complex requires DnaA Arginine 285, a conserved motif in the AAA+ protein family. The Journal of biological chemistry. 2005;280(29):27420-30. 172. Felczak MM, Kaguni JM. The box VII motif of Escherichia coli DnaA protein is required for DnaA oligomerization at the E. coli replication origin. The Journal of biological chemistry. 2004;279(49):51156-62. 173. Ozaki S, Noguchi Y, Hayashi Y, Miyazaki E, Katayama T. Differentiation of the DnaA-oriC subcomplex for DNA unwinding in a replication initiation complex. The Journal of biological chemistry. 2012;287(44):37458-71. 174. Ozaki S, Kawakami H, Nakamura K, Fujikawa N, Kagawa W, Park SY, et al. A common mechanism for the ATP-DnaA-dependent formation of open complexes at the replication origin. The Journal of biological chemistry. 2008;283(13):8351-62. 175. Li S, Zhang Q, Xu Z, Yao Y-F. Acetylation of Lysine 243 Inhibits the oriC Binding Ability of DnaA in Escherichia coli. Frontiers in Microbiology. 2017;8(699). 176. Fujikawa N, Kurumizaka H, Nureki O, Terada T, Shirouzu M, Katayama T, et al. Structural basis of replication origin recognition by the DnaA protein. Nucleic Acids Res. 2003;31(8):2077-86. 177. Blaesing F, Weigel C, Welzeck M, Messer W. Analysis of the DNA-binding domain of Escherichia coli DnaA protein. Molecular microbiology. 2000;36(3):557-69. 178. Erzberger JP, Pirruccello MM, Berger JM. The structure of bacterial DnaA: implications for general mechanisms underlying DNA replication initiation. The EMBO journal. 2002;21(18):4763-73. 179. Shimizu M, Noguchi Y, Sakiyama Y, Kawakami H, Katayama T, Takada S. Near-atomic structural model for bacterial DNA replication initiation complex and its functional insights. Proceedings of the National Academy of Sciences. 2016;113(50):E8021-E30. 180. Keyamura K, Katayama T. DnaA protein DNA-binding domain binds to Hda protein to promote inter-AAA+ domain interaction involved in regulatory inactivation of DnaA. The Journal of biological chemistry. 2011;286(33):29336-46. 181. Noguchi Y, Sakiyama Y, Kawakami H, Katayama T. The Arg Fingers of Key DnaA Protomers Are Oriented Inward within the Replication Origin oriC and Stimulate DnaA Subcomplexes in the Initiation Complex. The Journal of biological chemistry. 2015;290(33):20295-312. 182. Ozaki S, Katayama T. Highly organized DnaA-oriC complexes recruit the single-stranded DNA for replication initiation. Nucleic Acids Res. 2012;40(4):1648-65. 183. Frimodt-Møller J, Charbon G, Krogfelt KA, Løbner-Olesen A. Control regions for chromosome replication are conserved with respect to sequence and location among Escherichia coli strains. Frontiers in Microbiology. 2015;6:1011. 184. Kaur G, Vora MP, Czerwonka CA, Rozgaja TA, Grimwade JE, Leonard AC. Building the bacterial orisome: high-affinity DnaA recognition plays a role in setting the conformation of oriC DNA. Molecular microbiology. 2014;91(6):1148-63.

117 185. Duderstadt KE, Chuang K, Berger JM. DNA stretching by bacterial initiators promotes replication origin opening. Nature. 2011;478:209. 186. Duderstadt KE, Mott ML, Crisona NJ, Chuang K, Yang H, Berger JM. Origin remodeling and opening in bacteria rely on distinct assembly states of the DnaA initiator. The Journal of biological chemistry. 2010;285(36):28229-39. 187. Costa A, Hood IV, Berger JM. Mechanisms for initiating cellular DNA replication. Annual review of biochemistry. 2013;82:25-54. 188. Leonard AC, Grimwade JE. The orisome: structure and function. Frontiers in Microbiology. 2015;6:545. 189. Arai K, Yasuda S, Kornberg A. Mechanism of dnaB protein action. I. Crystallization and properties of dnaB protein, an essential replication protein in Escherichia coli. The Journal of biological chemistry. 1981;256(10):5247-52. 190. Reha-Krantz LJ, Hurwitz J. The dnaB gene product of Escherichia coli. II. Single stranded DNA- dependent ribonucleoside triphosphatase activity. The Journal of biological chemistry. 1978;253(11):4051- 7. 191. Bujalowski W, Klonowska MM, Jezewska MJ. Oligomeric structure of Escherichia coli primary replicative helicase DnaB protein. The Journal of biological chemistry. 1994;269(50):31350-8. 192. Donate L-E, Llorca Ó, Bárcena M, Brown SE, Dixon NE, Carazo J-Ma. pH-controlled quaternary states of hexameric DnaB helicase. Journal of molecular biology. 2000;303(3):383-93. 193. Chodavarapu S, Jones AD, Feig M, Kaguni JM. DnaC traps DnaB as an open ring and remodels the domain that binds primase. Nucleic Acids Research. 2016;44(1):210-20. 194. Seitz H, Weigel C, Messer W. The interaction domains of the DnaA and DnaB replication proteins of Escherichia coli. Molecular microbiology. 2000;37(5):1270-9. 195. Soultanas P. Loading mechanisms of ring helicases at replication origins. Molecular microbiology. 2012;84(1):6-16. 196. Makowska-Grzyska M, Kaguni JM. Primase directs the release of DnaC from DnaB. Mol Cell. 2010;37(1):90-101. 197. Keyamura K, Fujikawa N, Ishida T, Ozaki S, Su'etsugu M, Fujimitsu K, et al. The interaction of DiaA and DnaA regulates the replication cycle in E. coli by directly promoting ATP DnaA-specific initiation complexes. Genes & development. 2007;21(16):2083-99. 198. Zhou H-X. The Affinity-Enhancing Roles of Flexible Linkers in Two-Domain DNA-Binding Proteins. Biochemistry. 2001;40(50):15069-73. 199. Camara JE, Breier AM, Brendler T, Austin S, Cozzarelli NR, Crooke E. Hda inactivation of DnaA is the predominant mechanism preventing hyperinitiation of Escherichia coli DNA replication. EMBO Rep. 2005;6(8):736-41. 200. Kato J-i, Katayama T. Hda, a novel DnaA-related protein, regulates the replication cycle in Escherichia coli. The EMBO journal. 2001;20(15):4253-62. 201. Su'etsugu M, Nakamura K, Keyamura K, Kudo Y, Katayama T. Hda monomerization by ADP binding promotes replicase clamp-mediated DnaA-ATP hydrolysis. The Journal of biological chemistry. 2008;283(52):36118-31. 202. Su'etsugu M, Shimuta TR, Ishida T, Kawakami H, Katayama T. Protein associations in DnaA- ATP hydrolysis mediated by the Hda-replicase clamp complex. The Journal of biological chemistry. 2005;280(8):6528-36. 203. Kim JS, Nanfara MT, Chodavarapu S, Jin KS, Babu VMP, Ghazy MA, et al. Dynamic assembly of Hda and the sliding clamp in the regulation of replication licensing. Nucleic Acids Res. 2017;45(7):3888-905. 204. Katayama T, Ozaki S, Keyamura K, Fujimitsu K. Regulation of the replication cycle: conserved and diverse regulatory systems for DnaA and oriC. Nature reviews Microbiology. 2010;8(3):163-70. 205. Kitagawa R, Mitsuki H, Okazaki T, Ogawa T. A novel DnaA protein-binding site at 94.7 min on the Escherichia coli chromosome. Molecular microbiology. 1996;19(5):1137-47.

118 206. Nozaki S, Yamada Y, Ogawa T. Initiator titration complex formed at datA with the aid of IHF regulates replication timing in Escherichia coli. Genes to cells : devoted to molecular & cellular mechanisms. 2009;14(3):329-41. 207. Ogawa T, Yamada Y, Kuroda T, Kishi T, Moriya S. The datA locus predominantly contributes to the initiator titration mechanism in the control of replication initiation in Escherichia coli. Molecular microbiology. 2002;44(5):1367-75. 208. Kasho K, Tanaka H, Sakai R, Katayama T. Cooperative DnaA Binding to the Negatively Supercoiled datA Locus Stimulates DnaA-ATP Hydrolysis. Journal of Biological Chemistry. 2017;292(4):1251- 66. 209. Ali Azam T, Iwata A, Nishimura A, Ueda S, Ishihama A. Growth phase-dependent variation in protein composition of the Escherichia coli nucleoid. Journal of bacteriology. 1999;181(20):6361-70. 210. Frimodt-Møller J, Charbon G, Krogfelt KA, Løbner-Olesen A. DNA Replication Control Is Linked to Genomic Positioning of Control Regions in Escherichia coli. PLoS genetics. 2016;12(9):e1006286. 211. Morigen, Lobner-Olesen A, Skarstad K. Titration of the Escherichia coli DnaA protein to excess datA sites causes destabilization of replication forks, delayed replication initiation and delayed cell division. Molecular microbiology. 2003;50(1):349-62. 212. Cagliero C, Grand RS, Jones MB, Jin DJ, O'Sullivan JM. Genome conformation capture reveals that the Escherichia coli chromosome is organized by replication and transcription. Nucleic Acids Res. 2013;41(12):6058-71. 213. Messer W, Bellekes U, Lother H. Effect of dam methylation on the activity of the E. coli replication origin, oriC. The EMBO journal. 1985;4(5):1327-32. 214. Russell DW, Zinder ND. Hemimethylation prevents DNA replication in E. coli. Cell. 1987;50(7):1071-9. 215. Campbell JL, Kleckner N. E. coli oriC and the dnaA gene promoter are sequestered from dam methyltransferase following the passage of the chromosomal replication fork. Cell. 1990;62(5):967-79. 216. Boye E. The hemimethylated replication origin of Escherichia coli can be initiated in vitro. Journal of bacteriology. 1991;173(14):4537-9. 217. Waldminghaus T, Skarstad K. The Escherichia coli SeqA protein. Plasmid. 2009;61(3):141-50. 218. Kurokawa K, Nishida S, Emoto A, Sekimizu K, Katayama T. Replication cycle-coordinated change of the adenine nucleotide-bound forms of DnaA protein in Escherichia coli. The EMBO journal. 1999;18(23):6642-52. 219. Skarstad K, Lobner-Olesen A. Stable co-existence of separate replicons in Escherichia coli is dependent on once-per-cell-cycle initiation. The EMBO journal. 2003;22(1):140-50. 220. Skarstad K, Katayama T. Regulating DNA Replication in Bacteria. Cold Spring Harbor Perspectives in Biology. 2013;5(4):a012922. 221. Taghbalout A, Landoulsi A, Kern R, Yamazoe M, Hiraga S, Holland B, et al. Competition between the replication initiator DnaA and the sequestration factor SeqA for binding to the hemimethylated chromosomal origin of E. coli in vitro. Genes to cells : devoted to molecular & cellular mechanisms. 2000;5(11):873-84. 222. Nievera C, Torgue JJ, Grimwade JE, Leonard AC. SeqA blocking of DnaA-oriC interactions ensures staged assembly of the E. coli pre-RC. Mol Cell. 2006;24(4):581-92. 223. Bach T, Morigen, Skarstad K. The initiator protein DnaA contributes to keeping new origins inactivated by promoting the presence of hemimethylated DNA. Journal of molecular biology. 2008;384(5):1076-85. 224. Torheim NK, Skarstad K. Escherichia coli SeqA protein affects DNA topology and inhibits open complex formation at oriC. The EMBO journal. 1999;18(17):4882-8. 225. Bogan JA, Helmstetter CE. DNA sequestration and transcription in the oriC region of Escherichia coli. Molecular microbiology. 1997;26(5):889-96.

119 226. Riber L, Løbner-Olesen A. Coordinated Replication and Sequestration of oriC and dnaA Are Required for Maintaining Controlled Once-per-Cell-Cycle Initiation in Escherichia coli. Journal of bacteriology. 2005;187(16):5605-13. 227. Kasho K, Fujimitsu K, Matoba T, Oshima T, Katayama T. Timely binding of IHF and Fis to DARS2 regulates ATP–DnaA production and replication initiation. Nucleic Acids Research. 2014;42(21):13134-49. 228. Sekimizu K, Kornberg A. Cardiolipin activation of dnaA protein, the initiation protein of replication in Escherichia coli. The Journal of biological chemistry. 1988;263(15):7131-5. 229. Crooke E, Castuma CE, Kornberg A. The chromosome origin of Escherichia coli stabilizes DnaA protein during rejuvenation by phospholipids. The Journal of biological chemistry. 1992;267(24):16779-82. 230. Castuma CE, Crooke E, Kornberg A. Fluid membranes with acidic domains activate DnaA, the initiator protein of replication in Escherichia coli. The Journal of biological chemistry. 1993;268(33):24665- 8. 231. Xia W, Dowhan W. In vivo evidence for the involvement of anionic phospholipids in initiation of DNA replication in Escherichia coli. Proceedings of the National Academy of Sciences of the United States of America. 1995;92(3):783-7. 232. Zheng W, Li Z, Skarstad K, Crooke E. Mutations in DnaA protein suppress the growth arrest of acidic phospholipid-deficient Escherichia coli cells. The EMBO journal. 2001;20(5):1164-72. 233. Fingland N, Flåtten I, Downey CD, Fossum-Raunehaug S, Skarstad K, Crooke E. Depletion of acidic phospholipids influences chromosomal replication in Escherichia coli. MicrobiologyOpen. 2012;1(4):450-66. 234. Hansen FG, Christensen BB, Atlung T. The initiator titration model: computer simulation of chromosome and minichromosome control. Research in microbiology. 1991;142(2):161-7. 235. Charbon G, Bjørn L, Mendoza-Chamizo B, Frimodt-Møller J, Løbner-Olesen A. Oxidative DNA damage is instrumental in hyperreplication stress-induced inviability of Escherichia coli. Nucleic Acids Research. 2014;42(21):13228-41. 236. Michaels ML, Miller JH. The GO system protects organisms from the mutagenic effect of the spontaneous lesion 8-hydroxyguanine (7,8-dihydro-8-oxoguanine). Journal of bacteriology. 1992;174(20):6321-5. 237. Simmons LA, Breier AM, Cozzarelli NR, Kaguni JM. Hyperinitiation of DNA replication in Escherichia coli leads to replication fork collapse and inviability. Molecular microbiology. 2004;51(2):349- 58. 238. Gon S, Camara JE, Klungsoyr HK, Crooke E, Skarstad K, Beckwith J. A novel regulatory mechanism couples deoxyribonucleotide synthesis and DNA replication in Escherichia coli. The EMBO journal. 2006;25(5):1137-47. 239. Babu VMP, Itsko M, Baxter JC, Schaaper RM, Sutton MD. Insufficient levels of the nrdAB- encoded ribonucleotide reductase underlie the severe growth defect of the Δhda E. coli strain. Molecular microbiology. 2017;104(3):377-99. 240. Riber L, Olsson JA, Jensen RB, Skovgaard O, Dasgupta S, Marinus MG, et al. Hda-mediated inactivation of the DnaA protein and dnaA gene autoregulation act in concert to ensure homeostatic maintenance of the Escherichia coli chromosome. Genes & development. 2006;20(15):2121-34. 241. Fujimitsu K, Su'etsugu M, Yamaguchi Y, Mazda K, Fu N, Kawakami H, et al. Modes of Overinitiation, dnaA Gene Expression, and Inhibition of Cell Division in a Novel Cold-Sensitive hda Mutant of Escherichia coli. Journal of bacteriology. 2008;190(15):5368-81. 242. Riber L, Fujimitsu K, Katayama T, Lobner-Olesen A. Loss of Hda activity stimulates replication initiation from I-box, but not R4 mutant origins in Escherichia coli. Molecular microbiology. 2009;71(1):107- 22. 243. Charbon G, Campion C, Chan SHJ, Bjørn L, Weimann A, da Silva LCN, et al. Re-wiring of energy metabolism promotes viability during hyperreplication stress in E. coli. PLoS genetics. 2017;13(1):e1006590.

120 244. Charbon G, Riber L, Cohen M, Skovgaard O, Fujimitsu K, Katayama T, et al. Suppressors of DnaA(ATP) imposed overinitiation in Escherichia coli. Molecular microbiology. 2011;79(4):914-28. 245. Imlay J, Chin S, Linn S. Toxic DNA damage by hydrogen peroxide through the Fenton reaction in vivo and in vitro. Science. 1988;240(4852):640-2. 246. Liu Y, Bauer SC, Imlay JA. The YaaA protein of the Escherichia coli OxyR regulon lessens hydrogen peroxide toxicity by diminishing the amount of intracellular unincorporated iron. Journal of bacteriology. 2011;193(9):2186-96. 247. Surivet JP, Lange R, Hubschwerlen C, Keck W, Specklin JL, Ritz D, et al. Structure-guided design, synthesis and biological evaluation of novel DNA ligase inhibitors with in vitro and in vivo anti- staphylococcal activity. Bioorganic & medicinal chemistry letters. 2012;22(21):6705-11. 248. Mills SD, Eakin AE, Buurman ET, Newman JV, Gao N, Huynh H, et al. Novel bacterial NAD+- dependent DNA ligase inhibitors with broad-spectrum activity and antibacterial efficacy in vivo. Antimicrobial agents and chemotherapy. 2011;55(3):1088-96. 249. Rose Y, Ciblat S, Reddy R, Belley AC, Dietrich E, Lehoux D, et al. Novel non-nucleobase inhibitors of Staphylococcus aureus DNA polymerase IIIC. Bioorganic & medicinal chemistry letters. 2006;16(4):891-6. 250. Tarantino PM, Jr., Zhi C, Wright GE, Brown NC. Inhibitors of DNA polymerase III as novel antimicrobial agents against gram-positive eubacteria. Antimicrobial agents and chemotherapy. 1999;43(8):1982-7. 251. Kling A, Lukat P, Almeida DV, Bauer A, Fontaine E, Sordello S, et al. Antibiotics. Targeting DnaN for tuberculosis therapy using novel griselimycins. Science. 2015;348(6239):1106-12. 252. Kjelstrup S, Hansen PMP, Thomsen LE, Hansen PR, Løbner-Olesen A. Cyclic Peptide Inhibitors of the β-Sliding Clamp in Staphylococcus aureus. PloS one. 2013;8(9):e72273. 253. Marceau AH, Bernstein DA, Walsh BW, Shapiro W, Simmons LA, Keck JL. Protein interactions in genome maintenance as novel antibacterial targets. PloS one. 2013;8(3):e58765. 254. Fossum S, De Pascale G, Weigel C, Messer W, Donadio S, Skarstad K. A robust screen for novel antibiotics: specific knockout of the initiator of bacterial DNA replication. FEMS Microbiol Lett. 2008;281(2):210-4. 255. Johnsen L, Weigel C, von Kries J, Moller M, Skarstad K. A novel DNA gyrase inhibitor rescues Escherichia coli dnaAcos mutant cells from lethal hyperinitiation. The Journal of antimicrobial chemotherapy. 2010;65(5):924-30. 256. Yamaichi Y, Duigou S, Shakhnovich EA, Waldor MK. Targeting the Replication Initiator of the Second Vibrio Chromosome: Towards Generation of Vibrionaceae-Specific Antimicrobial Agents. PLOS Pathogens. 2009;5(11):e1000663. 257. Weigel C, Schmidt A, Seitz H, Tungler D, Welzeck M, Messer W. The N-terminus promotes oligomerization of the Escherichia coli initiator protein DnaA. Molecular microbiology. 1999;34(1):53-66. 258. Renwick MJ, Simpkin V, Mossialos E. European Observatory Health Policy Series. Targeting innovation in antibiotic drug discovery and development: The need for a One Health - One Europe - One World Framework. Copenhagen (Denmark): European Observatory on Health Systems and Policies (c) World Health Organization 2016 (acting as the host organization for, and secretariat of, the European Observatory on Health Systems and Policies). 2016. 259. Beaber JW, Hochhut B, Waldor MK. SOS response promotes horizontal dissemination of antibiotic resistance genes. Nature. 2003;427:72. 260. Amundsen SK, Spicer T, Karabulut AC, Londoño LM, Eberhart C, Fernandez Vega V, et al. Small-Molecule Inhibitors of Bacterial AddAB and RecBCD Helicase-Nuclease DNA Repair Enzymes. ACS chemical biology. 2012;7(5):879-91. 261. Singh R, Swick MC, Ledesma KR, Yang Z, Hu M, Zechiedrich L, et al. Temporal Interplay between Efflux Pumps and Target Mutations in Development of Antibiotic Resistance in Escherichia coli. Antimicrobial agents and chemotherapy. 2012;56(4):1680-5.

121 262. Nishino K, Latifi T, Groisman EA. Virulence and drug resistance roles of multidrug efflux systems of Salmonella enterica serovar Typhimurium. Molecular microbiology. 2006;59(1):126-41. 263. Mahamoud A, Chevalier J, Alibert-Franco S, Kern WV, Pages JM. Antibiotic efflux pumps in Gram-negative bacteria: the inhibitor response strategy. The Journal of antimicrobial chemotherapy. 2007;59(6):1223-9. 264. Hannah YM, Shirin J, Sutton JM, Khondaker MR. Current Advances in Developing Inhibitors of Bacterial Multidrug Efflux Pumps. Current medicinal chemistry. 2016;23(10):1062-81. 265. Opperman TJ, Nguyen ST. Recent advances toward a molecular mechanism of efflux pump inhibition. Frontiers in Microbiology. 2015;6:421. 266. Lopez JS, Banerji U. Combine and conquer: challenges for targeted therapy combinations in early phase trials. Nature Reviews Clinical Oncology. 2016;14:57. 267. Cox MM, Goodman MF, Kreuzer KN, Sherratt DJ, Sandler SJ, Marians KJ. The importance of repairing stalled replication forks. Nature. 2000;404(6773):37-41. 268. Filutowicz M, Roll J. The requirement of IHF protein for extrachromosomal replication of the Escherichia coli oriC in a mutant deficient in DNA polymerase I activity. The New biologist. 1990;2(9):818- 27. 269. Ogawa T, Wada M, Kano Y, Imamoto F, Okazaki T. DNA replication in Escherichia coli mutants that lack protein HU. Journal of bacteriology. 1989;171(10):5672-9. 270. Katayama T, Takata M, Sekimizu K. The nucleoid protein H-NS facilitates chromosome DNA replication in Escherichia coli dnaA mutants. Journal of bacteriology. 1996;178(19):5790-2. 271. Charbon G, Riber L, Løbner-Olesen A. Countermeasures to survive excessive chromosome replication in Escherichia coli. Current Genetics. 2018;64(1):71-9. 272. Wargachuk R, Marczynski GT. The Caulobacter crescentus Homolog of DnaA (HdaA) Also Regulates the Proteolysis of the Replication Initiator Protein DnaA. Journal of bacteriology. 2015;197(22):3521-32. 273. Maffioli SI, Zhang Y, Degen D, Carzaniga T, Del Gatto G, Serina S, et al. Antibacterial Nucleoside-Analog Inhibitor of Bacterial RNA Polymerase. Cell. 2017;169(7):1240-8.e23. 274. Jabés D, Brunati C, Candiani G, Riva S, Romanó G, Donadio S. Efficacy of the New Lantibiotic NAI-107 in Experimental Infections Induced by Multidrug-Resistant Gram-Positive Pathogens. Antimicrobial agents and chemotherapy. 2011;55(4):1671-6.

122